PRODUCTION OF SINGLE CELL PROTEIN

BY CULTIVATION OF YEASTS
ON FATS AND FAT PRODUCTS
J. VAN DER VEEN
o
o o
o o
>0 Vf)
BIBLIOTHEEK TU Delft
P 1819 4231
555004
PRODUCTION OF SINGLE CELL PROTEIN
BY CULTIVATION OF YEASTS
ON FATS AND FAT PRODUCTS
PROEFSCHRIFT
TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN
DE TECHNISCHE WETENSCHAPPEN AAN DE TECHNI-
SCHE HOGESCHOOL DELFT, OP GEZAG VAN DE REC-
TOR MAGNIFICUS IR. H.B. BOEREMA, HOOGLERAAR
INDE AFDELING DER ELECTROTECHNIEK, VOOR EEN
COMMISSIE AANGEWEZEN DOOR HET COLLEGE VAN
DEKANENTE VERDEDIGEN OP WOENSDAG 20 NOVEM-
BER 1974 TE 16.00 UUR DOOR
JAN VAN DER VEEN /O/O ^2b/
CHEMISCH DOCTORANDUS
GEBOREN TE ROTTERDAM
.<:'^;f^-
DRUKKERIJ ELINKWIJK B.V. - UTRECHT
DIT PROEFSCHRIFT IS GOEDGEKEURD DOOR
DE PROMOTOR PROFESSOR DR. T.O. WIKEN
DIT PROEFSCHRIFT WERD BEWERKT OP HET
CENTRAAL INSTITUUT VOOR VOEDINGSONDERZOEK
CIVO TNO TE ZEIST
TO
THE FOOD AND AGRICULTURE ORGANIZATION
OF THE UNITED NATIONS (FAO)
CONTENTS
PREFACE 10
LIST OF ABBREVIATI ONS 11
Chapter I INTRODUCTION AND SCOPE 13
1. World protein shortage 13
2. General aspects of single cell protein (SCP) production 19
3. Specific problems of fat fermentation 22
4. Availability of fats and fat products 23
Chapter II VARIOUS ANALYTICAL METHODS 26
1. Fat analysis 26
1.1. Moisture, impurities and unsaponifiable matter (MIU) 26
1.2. Saponification value (SV) 26
1.3. Fatty acid content and fat content 26
1.4. Fatty acid profile of fat products 27
1.5. Fat content of emulsions 27
1.6. Emulsifier in fat products 28
1.7. Pesticides in fat products 28
2. Yeast analysis 28
2.1. Yeast concentration in culture liquid 28
2.2. Moisture content and ash content - 30
2.3. Crude protein content and lipid content 30
2.4. Polysaccharide content and crude fibre content 30
2.5. Amounts of emulsifier and pesticides 31
2.6. Elementary analysis and gross formula 31
2.7. Amino acid spectrum and fatty acid profile 31
2.8. Vitamin composition 31
2.9. Elements in ash 33
2.10. Content of ribo-nucleic acid (RNA) and purine base content 34
3. Sanitary analysis of yeast 34
3.1. Introduction 34
3.2. Total counts of aerobic bacteria and spores 35
3.3. Enterobacteriaceae count 35
3.4. Escherichia coli present/absent test 35
3.5. Salmonella present/absent test 36
3.6. Counts of Clostridium group and Clostridium perfringens 36
3.7. Lancefield group D Streptococci count 36
3.8. Staphylococcus aureus present/absent test 36
3.9. Moulds . 37
4. Nutritional testing of yeast 37
4.1. Protein calories per cent (PCP) 37
4.2. Chemical score and essential amino acid index (EAA-lndex) 37
4.3. Protein efficiency ratio (PER) 40
4.4. Net protein utilization (NPU), true digestibility (TD) and 41
biological value (BV)
5. Safety evaluation of yeast 43
5.1. Introduction 43
5.2. Screening test 44
5.3. Sub~chronic study 44
5.4. Chronic feeding study 44
5.5. Multigeneration study 45
5.6. Teratogenicity studies 45
5.7. Mutagenicity studies 46
5.8. Studies with farm animals 46
5.9. Derived tests 46
Chapter III STANDARD PROCEDURE FOR FAT FERMENTATION 48
1. Fermentation steps 48
1.1. Preparation of the slant cultures 48
1.2. Procedure for the shake flask cultures 50
1.3. Apparatus for 1 liter fermentation 54
1.4. Apparatus for 20 liter fermentation 61
1.5. Harvesting of the yeast cells 64
2. Selection of yeast strain, composition of the media etc. 65
2.1. Selection of the strain 65
2.2. Concentration of the fat products 70
2.3. Fermentation temperature 71
2.4. Fat emulsification 72
2.5. Composition of the salt mixtures 75
2.6. Suppletion of vitamins 81
\
Chapter IV APPLICATION OF STANDARD PROCEDURE 85
1. Batch fermentations of grease emulsions 85
2. Batch fermentations of fatty acids 90
3. Continuous fermentation of fatty acids 95
4. Batch-to-batch fermentations of fatty acids 98
5. Fermentations of tallow emulsions 102
6. Fermentations of technical fat 105
Chapter V PROPERTIES OF PRODUCED YEAST 108
5.1. Organoleptic examination 108
5.2. Chemical composition 108
5.3. Biological characteristics 111
5.4. Bacteriological examination , 117
5.5. Toxicological indications 117
Chapter VI DISCUSSION AND CONCLUSIONS 121
1. Discussion of the results 121
2. Economic considerations 121
3. Final conchision 123
LIST OF REFERENCES 125
SUMMARIES, 130
English 130
Dutch 133
Russian ^ 136
CURRICULUM VITAE 143
PREFACE
Ever since the production of single cell protein (SCP) on mineral oil products was
started in the early sixties, the Central Institute for Nutrition and Food Research
(CIVOTNO) has been involved in its research and development; especially as far as
related biological and toxicological aspects are concerned.
Besides, in 1969, the governmental supervision of the refining and distribution of
imported inedible fats was assigned to the CIVO-TNO. The director of this
Institute, Dr. CG.J.M. Engel, delegated this task to the author of the present thesis.
Small wonder that, with these two streams coming together in one Institute, the
idea of using the fats as substrates for the production of SCP gained momentum,
particularly since in previous years the idea had been suggested by some
CIVO-TNO research workers.
Research was started in the latter part of 1970, and from the very beginning there
has been close cooperation with Professor Dr. T.O. Wiken, head of the Laboratory
of Microbiology at the University of Technology, Delft. The many inspiring
discussions we had over the years have contributed considerably to the quality and
the presentation of the thesis; I have very much appreciated them.
As will be clear from the respective Chapters, several departments of CIVOTNO
have contributed to the work. An appreciable number of co-workers have devoted
time and keen interest to this research; I herewith thank each of them
whole-heartedly.
Finally, I would never have been able to carry out this research, beside my regular
work, without the continuous support of my wife. The frequent discussions with
her and our three children, about the social relevancy of the subject have been a
real stimulus to me. It is with their full consent that this thesis is dedicated to the
Food and Agriculture Organization of the United Nations (FAO).
We all share in the hope that the FAO, being the main platform on which world
food supplies are scrutinized, will be able to make the national governments aware
of the major interest of food research programs so that, in not too distant future,
there may be enough to meet the minimum requirements of food for every human
being.
LIST OF ABBREVIATIONS
AV Acid value
atm atmosphere
BV Biological value
BBL Baltimore Biological Laboratory
CBS Centraalbureau voor Schimmelcultures
C Degree Centigrade
CIVO Central Institute for Nutrition and Food Research
cm centimeter
DDT Dichlorodiphenyl Trichloroethane
DNA Desoxyribonucleic Acid
EAAIndex Essential Amino Acid Index
FAAslants Fatty Acid Agar slants
FAslants Fat Agar slants
FAD Flavin Adenine Dinucleotide
FAO Food and Agriculture Organization of the United Nations
FID Flame lonisation Detection
FMN Flavin Mononucleotide
G
g
g/c/d
GYA-slants
GRA-slants
h
HCB
H/L
HLB
lUPAC
Gravitative acceleration
gram
gram per capita per day
Glucose-Yeast-extract-Agar slants
GlucoseRenexAgar slants
hour
Hexachlorobenzene
Hydrophile to Lipophile ratio
Hydrophile Lipophile Balance
International Union of Pure and Applied Chemistry
Joule
kg
kJ
1/h
kilo gram
kilo Joule
liter
liter per hour
m meter
meg microgram
mg milligram
min minute
MIU Moisture, Impurities and Unsaponifiable Matter
MJ Mega Joule
ml milliliter
mm milUmeter
m/m mass to mass ratio
mMol milHgrammolecule
Mol grammolecule
mV milli-Volt
i ^ 10-' g
N Nitrogen
n normality or normal
NAD Nicotinamide Adenine Dinucleotide
NADP Nicotinamide Adenine Dinucleotide Phosphate
NPU Net Protein Utilization
PAG Protein Advisory Group of the United Nations System
PEGA Polyethylene Glycol Adipate
PER Protein Efficiency Ratio
p.m. pro memory
ppm parts per million
RAslants
RNA
rpm
Renex Agar slants
Ribonucleic Acid
rotations per miijute
SCP
SV
Single Cell Protein
Saponification value
TD True Digestibility
t/y tons per year
TNO Netherlands Organization for
AppUed Scientific Research
vol volume
v/v volume to volume ratio
v/v/h volume (air) per volume (medium) per hour
v/v/min volume (air) per volume (medium) per minute
Z centrifugal coefficient
CHAPTER I
INTRODUCTION AND SCOPE
1.1. World protein shortage
Tlie world protein problem has been discussed in so many pubUcations that it
would hardly be expected to find anyone who could add some truly new points of
view.
Therefore, it is rather amazing that from a number of recent publications the
impression emerges that there is no protein shortage at all or that, at least
statistically, the amoun of protein yearly available would seem to be more than the
minimum level for survival of man. In Table 11-1 this statement is uUustrated by
figures, collected by Autret (1969), Engel (1971) and Abbot (1967) about the
protein supply per capita per day.
When these figures are related to egg protein by multiplying the vegetable protein
figure by the factor of 0.55, and the animal protein figure by the factor of 0.8, the
amount available is found to be to 41-43 g/c/d. In FAO Report no. 52 (1973), a
theoretical safety level of intake of egg or milk protein calculated for human
populations gives an amount of 29 g/c/d. This means that the amount of protein
available would be 45% greater than the so-called 'safe level of intake'. The reason
for the fact that, in spite of this, a large part of the world population suffers from
undernourishment as regards protein is not only a wrong distribution system but
has also to do with the circumstance that, in some parts of the world, the protein
intake per capita is much higher than the safe level, leaving for people in other
regions far less than the required minimum. To illustrate this statement, the protein
consumption in India may be compared with that of the United States of North
America. As will be seen in Table I 1-2, the average daily intake of both areas
together is 47.5 g/c/d egg protein. However the consumption of the Indian people is
at the level of 29 g/c/d, whereas the inhabitant in the USA consumes on an average
66.3 g/c/d.
Again the point must be stressed that the fact that, statistically, the amount of
protein available for each inhabitant of India is just at the minimum level does not
mean that every inhabitant really gets the amount concerned. In the first place, this
is due to difficulties of transportation, distribution, etc. Secondly, the same finding
that is true for the world as a whole is true also for the area mentioned. The rich
people in India are able to buy more than the minimum amount, and will definitely
do so, leaving for the poor people an amount far less than the 'safe level' of 29 g/c/d.
The figures in Table I 12, related to egg protein, were calculated on the
13
Table 1 I 1 World supply of protein in grams per capita per day, related to egg protein based on figures collected by Autret (1969), Engel
(1971) and Abbot (1967)
NPU%
Cereals
55
Autret
Egg protein
Engel
Egg protein
Abbot
Egg protein
31 8
175
33 7
18 5
33 4
184
Starchy
roots
62
2 7
1 7
2 7
1 7
3 2
2 0
Pulses,
oil-
seeds,
and
nuts
44
8 0
3 5
9 0
4 0
9 0
4 0
Vege-
tables
and
fruit
80
2 6
2 1
2 7
2 2
2 4
19
Total
vege
table
protein
55
45 1
24 8
48 1
26 4
48 0
26 3
Meat
and-
poul-
try
70
9 8
6 9
8 4
5 9
8 8
6 2
Eggs
100
1 4
1 4
1 3
1 3
1 2
1 2
Fish
83
26
22
49
4 1
23
1 9
Milk
and
milk
products
Total
animal
pro-
tein
Total
protein
g/c/d
85 80
7 2
6 1
6 4
5 4
7 7
6 5
2 1 0
166
21 0
167
20 0
158
66 1
41 4
69 1
43 1
68 0
42 1
Vegetable, animal and total protein, related to egg protein,
expressed in grams per capita per day and in percentages
Vegetable
protein
Animal
protein
Total
protein
Egg protein
Total protein
g/c/d
%
g/c/d
%
g/c/d
Autret
Egg protein
Engel
Egg protein
Abbot
Egg protein
45 1
24 8
48 1
26 4
48 0
26 3
68 2
60 0
69 6
61 2
70 6
62 5
21 0
166
21 0
167
20 0
158
31 8
40 0
30 4
38 8
29 4
37 5
66 1
41 4
69 1
43 1
68 0
42 1
100 0
100 0
100 0
100 0
100 0
100 0
62 6%
62 4%
619%
Table I 1 2 Average protein supply, related to egg protein, in grams per capita
per day, in India and the United States of North America over the
period 1960-62
Cereals
Starchy
roots
Pulses, nuts
oilseeds etc.
Fruit and
vegetables
Total vege-
table protein
Meat
Egg
Fish
Milk and
milk products
Total animal
protein
Total protein
Food
protein
31.0
0.9
13.2
0.4
45.5
0.6
0.1
0.5
India
Egg
protein
17.1
0.6
5.8
0.3
23.8
0.4
0.1
0.4
United States of
North America
Food
protein
15.6
2.2
4.1
5.0
26.9
32.4
5.7
2.5
Egg
protein
8.6
1.4
1.8
4.1
15.8
22.7
5.7
2.1
4.8
6.0
51.5
4.1
5.0
28.8
23.5
64.1
91.0
20.0
50.5
66.3
Protein calories
Total calories
10.2% 11.7%
assumption that the various proteins serve as the only protein source available. With
mixed proteins, other conversion factors have to be applied and this might result in
higher values for the net protein utilization.
In this connection it must be emphasized that the net protein utilization of a
foodstuff can give a true picture of the nutritional situation only when enough
calories are suppHed by the diet concerned. In a well-balanced diet, the
contribution of the protein energy to the total energy available should lie between
10% and 12%. As is seen in Table I 1-2, it is 11.7% for North America and 10.2%
for India. But if the world population is considered, one finds that the calorie level
of 2380 cal/day needs a protein calorie contribution of 238 to 286 calories, which
15
means a protein intake of 60-72 g/c/d. As will be seen in tabel I 1-1, this value
is 66-69 g/c/d for the world population. In this respect, however, there is
undoubtedly a protein shortage in India (Table I 1-2). To express this specifically,
the term protein calorie malnutrition is used. Abbot (1967) has given figures
for particular areas which show that for the Near East, Africa, the Far East and
Latin America, (except Argentine, Uruguay and Paraguay), the protein supply,
expressed as egg protein, is 33.4 g/c/d and the protein calorie percentage 9.8%.
The total population living in these areas is about two thirds of the world
population. This means that although the total protein production could cover the
minimum requirements, two thirds of the world population are undernourished in
so far as they suffer from protein malnutrition.
On this point, a more detailed and exact picture is obtained when vegetable and
animal proteins are considered separately. According to Abbot (1967), the
production of animal foodstuff has increased in the developed countries since
World War II. In the developing areas, however, the total per capita protein supply
shows a decrease of about 6% in the same period, accompanied by an increased
dependence on cereals. From the data given in Table I 1-2, it is evident that the
conversion factor for vegetable protein to egg protein is 0.55 on an average, whereas
for animal protein this factor is 0.8. For the 'safe level' intake of 29 g/c/d, the
following equation can be given: 29 = V x 0.55 + A x 0.8, in which A denotes
animal protein and V vegetable protein. Theoretically this means that, if no animal
protein is available, 52.7 g/c/d of vegetable protein is the minimum level. For
exclusively animal protein this level is 36.3 g/c/d (V = 0).
Figure I 1 1 gives the said relation graphically:
36 3
^ \ Figure I 1-1 Graph of the equation:
\ . 29 = 0.55xV + 0.8xA
\ ^^ 29 = minimum protein level in g/c/d
'" \ ^ V = vegetable protein intake in g/c/d
\ , A = animal protein intake in g/c/d
25 \ ,
20 ^ v
10
0 5 10 IS 2(1 :; 3] 35 40 45 SO 52 7 V
16
From Figure I 1 1 it will be seen that, if the contribution of animal protein is 15
g/c/d, the amount of vegetable protein required to reach the minimum level is 31
g/c/d. It should be realized that the animal and vegetable proteins are not consumed
as such but incorporated in a foodstuff; the other components of this foodstuff
contribute to the t(jtal nutritional situation too. Moreover, in discussing the
difference in contribution of animal and vegetable proteins for covering the total
requirements it is essential to realize that, for the production of a definite amount
of animal protein, a far greater quantity of vegetable protein must be available as
fodder protein. To illustrate this aspect, the conversion of total feed to edible
carcass and of protein ingested to edible protein is given in Table I 1 3 for various
animals (Wilcke, 1973). In addition, the relation of protein consumed to edible
protein yield is listed (Wilson, 1971).
It should be taken into account that the fodder protein at least partly consists of
Table I 13 Conversion of total feed to edible carcass and of protein consumed
to protein formed for a number of animals according to Wilcke
(1973).
Class Edible carcass
Total feed
xl00%
Edible protein
Protein ingested
xl00%
Beef
Swine
Broiler
Hen
2.0
17.0
18.3
26.7
4.7
12.0
17.0
23.0
Protein efficiency according to Wilson (1971)
Class
Broilers
Porker
Lamb
Steer
Production
level
6 crops at
4 Ib/wt
21/2 litters
of 12 at
90 lb carcass
2 litters
of 3 at
35 lb carcass
650 lb carcass
Annual crude
protein yield
(lb)
2.5
270
24
75
Annual crude
protein consumed
(lb)
8
1850
275
1250
Protein
efficiency
%
31
15
9
6
17
Table I 1-4
Cereals
food protein
Potatoes
Food protein
Pulses, nuts
Food protein
Fruit and
vegetables
Food protein
Total vegetable
protein
Meat
Food protein
Egg
Protein
Fish
Food protein
Milk and
milk products
Food protein
Total animal
protein
Totals
18
Food balance sheet for the Netherlands expressed in thousand tons,
calculated over
Available
supply
5363
427
3537
56.6
102
11.4
2081
19.2
514.2
540
73.6
155
18.5
269
22.5
9550
421.1
535.7
1049.9
the period 1960
Waste,
seeds,
manufacture
259
21
1702
27.3
6
0.7
506
4.7
53.7
.
15.0
1.8
126
10.5
6637
292.7
305.2
358.9
-63
Animal
feed
3873
308
686
10.9
35
3.9
^
322.8
.
^
15
1.1
770
33.9
35.0
357.8
<
Gross
food
1231
98
1149
18.4
61
6.8
1575
14.5
137.7
540
73.6
140
16.7
128
10.7
2134
94.5
195.5
333.2
Food
protein
g/c/d
22.9
4.3
1.6
3.4
32.2
17.2
3.9
2.5
22.1
45.7
77.9
1
protein that cannot be, or is not, used for human consumption, such as grass
protein, protein from meat scraps, inedible fish meals, etc. However, to a very large
extent the protein used as fodder for animals could also serve as protein for human
consumption. Table 1 14 surveys the total supply of foodstuff, the amounts used
for animal feed, for wastes, seeds and manufacture and for human consumption,
related to protein; it is taken from the food balance sheets (1960-63) for the
Netherlands.
From Table I 1 4 the following conclusions may be drawn.
The total available amount of food protein in the Netherlands was in the period
mentioned 77.9 g/c/d, and from this amount 58.7% was of animal origin.
Of the supply of vegetable protein 62.8% was used for fodder, whereas for that
purpose only 6.5% of the animal protein supply was utilized.
Cereals were the main source for the supply of fodder protein (96.1%), followed by
protein from milk, skimmed milk and skimmed milk-powder (9.5%). On the other
hand, the percentage of cereals used for fodder was 72.1% and for milk protein 8%
only.
An interesting information that can be derived from Table I 14 is the total
amount of protein used as animal feed in relation to the total amount of meat
protein produced. This relation is (73.6 : 357.8) x 100% = 20.6%, which is much
higher than the figures given by Wilcke (1966). Even when the food wastes that are
used directly or indirectly as fodder are taken into consideration, the percentage is
far above the generally accepted average percentage of 9% to 11%.
Finally, something has to be said about the protein need in the near future. It is
expected (Abbot, 1973) that to meet the requirements the production of protein
should increase by 6% yearly. However, in order to increase the animal protein
production much more vegetable protein must be available as fodder. It would be a
very important contribution to the protein supply if at least a part of this fodder
protein could be provided by sources that are not contributing to human nutrition.
This very important condition is fulfilled by single cell protein (SCP) products.
For every 10% of the fodder protein in the Netherlands that would derive from this
source, which means a production of 60,000 to 80,000 tons of SCP yearly, the
gross food production would increase by 10%. Under these conditions, the
amount of milk protein available would increase by 60,000 tons per year. Apart
from any other application, it will be evident that SCP undoubtedly may be
considered an essential contribution to the world protein production.
1.2. General aspects of SCP production
The term 'Single Cell Protein' or SCP today includes a product more or less rich in
protein, obtained by cultivation of bacteria, yeasts and fungi. It has most probably
19
been chosen in order to enable the introduction of biosynthetic products of
microbial origin as animal fodder and/or human food, without mentioning yeasts,
bacteria or fungi by name and thus avoiding the prejudice people have in general
against these organisms
The advantages of SCP production are extensively emphasized by those interested
in promoting this biosynthetic protein production As compared with protein food
from animal and vegetable sources, the very high growth rate of microbes and the
possibility of producing large quantities of SCP of reproducible quality are the
most striking advantages For example, an annual production of SCP of 20,000 tons
with 9 6% N corresponds to the amount of protein produced by 100 000 cows
giving a milk production of 3900 1/y containing 30,5 g/1 protein (Frens 1961) It is
evident that milk production depends on climatological circumstances, whereas
SCP production does not Moreover, there is a very high economic use of land
space in the case of SCP production Worgan (1973) has calculated the land use of
yeast grown on sugar and that of beef cattle, he arrives at 2800 kg per hectare per
year for yeast protem and 42 kg per hectare per year for beef protein
As far as raw materials suitable for SCP production are concerned, a variety of waste
materials can be used, not only of agricultural but also of industrial origin, e g.
sulphite liquors, wood hydrolysates, etc
Very important sources for SCP production are or might be petrochemicals, e g
methanol and ethanol, and mineral oil products, e g gas oil and n-alkanes
In the case of agricultural wastes, synthetic symbiosis between various microbes,
e g yeasts, may be applied in production of SCP (The Symbaprocess, see Wiken,
1972)
Last, but not least, a minimal effluent, poor in organic substance, is often claimed
for SCP production and this is of great importance for environmental hygiene
(Wiken, 1972)
The rather impressive list of advantage of SCP production makes it understandable
that research in the field of microbial food production has grown tremendously
over the last few decades It is, therefore, rather astonishing that, in spite of all
these efforts, at this very moment of writing, Summer 1974, the production of SCP
on an economically profitable scale, i e 100,000 t/y, has not yet been effected In
search of the reasons for this phenomenon, a great number of factors emerge that
might be to a greater or lesser extent responsible
One of the most important reasons might be a difference in concept that exists
between technologists and microbiologists, as regards the way of tackling problems
connected with processes based on growth of microbial cells Although there has
been some intensive collaboration between the two disciplines in the last twenty
years, there is still a completely different approach as far as large scale production is
concerned The technologist often still prefers to overlook the fact that
micro-organisms are living cells and not simply enzyme bags that can adequately be
defined by their contents of fat, protein, carbohydrate and ash This aspect of
dealing with living cells gives the conditions and control of SCP production extra
20
dimensions. Contamination with undesirable micro-organisms, aeration problems and
prevention of excessive foaming are some of the factors that are obstacles for
scaling up to 100,000 t/y productions. But perhaps the most intriguing factor is the
susceptibility of yeasts, bacteria and fungi to even small changes in production
conditions. When emphasizing that a consistent quality is one of the advantages of
SCP production, we should add that this is true as far as the chemical composition
is concerned only, and as long as the operating conditions are strictly constant.
Because even minor changes in the process conditions may have important
consequences for the yield as well as for the nutritive value and the toxicological
safety of the product. Apart from technological reasons, this very fact might
indeed give an explanation for the present situation that as yet no 100,000 t/y
plant is in operation either in Japan, or in USSR, UK, France, etc.
A large number of toxicological investigations have been carried out with yeast
grown on gas oil and nalkanes, and the results published thus far do not give
any reason why this yeast should not be used as animal fodder. It does fulfil every
toxicological qualification and, when supplied with methionine, it has a very good
nutritive value (De Groot et al. 1970a, 1970b, 1971; Shacklady and Gatumel, 1972;
Van der Wal, 1968; Shacklady, 1969). One should, however, not overlook the fact
that the yeast concerned is an example of production on a much smaller scale than
the economically desirable 100,000 t/y plant. The technical disadvantages i.e. the low
concentration of the end product (2% max.), the expensive plant in terms of
materials and equipment, the high power consumption for aeration and
homogenizing, the necessity of great cooling capacity, etc., represent great
difficulties. But they can be solved as purely technical problems, though, admittedly,
these factors will have their economic consequences in terms of costs of the final
product.
As far as the sanitary, nutritive and toxicological aspects of the quality of the yeast
are concerned, much progress has meanwhile been made. The level of the aromatic
compounds present in yeast grown on gas oil is far below the safety level, thanks
to extensive extraction procedures. The amount of odd-numbered fatty acids,
present in yeast grown on n-alkanes and considered to cause undesirable effects, has
been reduced through a better procedure for the preparation of the n-alkanes.
Various processes have been developed to reduce the nucleic acid content of
microbial SCP, and its nutritional value has been raised by adding synthetic
methionine. For the rupture of the cell walls, which are indigestible for animals
because of the presence of 1 -4 linked polysaccharide derivatives, a number of
methods have been worked out, but so far they all result in lower protein yield and
higher production costs (Heydeman 1973). It must be admitted that, all in all, a
relatively large number of problems have still to be solved before SCP produced in
quantities of 100,000 t/y will be available for use as fodder.
21
1.3. Specific problems of fat fermentation
In addition to the more general problems of SCP production, briefly discussed in
the previous paragraph, some special remarks may now be made in relation to fat
fermentation. In the first place it be here emphasized that 'fat fermentation' is
incorrect terminologically, in spite of the fact that the term is generally accepted.
Actually the fats and fat products are not 'fermented' but used for anabolic or
biosynthetic and catabolic, mainly respiratory, processes, involving O2 uptake. A
better description, therefore, is 'the production of biomass, using fats and fat
products as sole sources of carbon and energy'. Wlien, in what follows, the term
'fermentation' is nevertheless applied, it denotes this concept.
Secondly, SCP production calls for a homogeneous reaction mixture. The nitrogen
source and other nutrients are water-soluble but the fats and fat products are not.
This gives rise to an extra problem in the case of fat fermentation. In addition to the
other substrate components, present in normal SCP production, the fat component
has to be homogenized too. Considering this problem, the raw materials used can be
divided in two groups: fats and fat products that have a melting-point at or below
30C, and those that have a higher melting-point. The first group of products can
be homogenized by mixing the fermentation liquid thoroughly, whereas for the
second group an emulsifier has to be applied. The application of an emulsifier has,
however, a number of disadvantages. The price of the emulsifier increases the
ultimate costs. Furthermore, the fact that the fat emulsion will be broken when
sterilized in the presence of mineral salts, makes it necessary to sterilize the
emulsion and salt solutions separately. This results in extra equipment costs.
Moreover, the presence of the emulsifier in the end product might influence the
nutritive value and may even have toxicological effects, in spite of the fact that the
emulsifier itself may well be guaranteed by the manufacturer to be nontoxic.
There are two other ways to utilize fats and fat products melting above 30 C as
raw materials for SCP production. The first would be to increase the fermentation
temperature. Actually, the Uterature describes processes in which SCP is produced
at higher temperature, even as high as 42C, but so far not with strains with
lipolytic activity.
Chapter IV will report some preliminary attempts with a selected strain and further
work to find a solution in this direction is in progress.
The second way would be to make mixtures of a liquid fat product and a solid one
that melt between 28C and 32C. This has been done in practice. However it
implies, of course, a restriction for the application of the animal fats, since the
procedure involves a certain amount of liquid products that might not always be
available.
22
1.4. Availibility of fats and fat products
The trade in animal fats in the Netherlands concerns edible fat, under control of the
Veterinary Inspection, and inedible fat. Under the supervision of the Ministry of
Health and Environmental Hygiene, the inedible fat is imported, distributed and
partly refined into products suitable for human consumption. In view of this
situation, the import of inedible animal fats and fat products is much larger than
the production of animal fat in the Netherlands.
This is illustrated in Table I 41
Table I 4-1 Production and import figures for animal fat in the Netherlands
(1968-1972) in 10* kg.
1968
1969
1970
1971
1972
Import
197.6
208.1
236.2
237.3
240.1
Production
99.0
94.1
106.8
117.9
123.4
Total
296.6
302.2
343.0
355.2
363.5
The refined animal fats are partly exported as such or as a product of fat, and
partly used in milk substitutes for calves. The quantity of imported fats has been
more or less constant over the last three years but that of refined fat is fluctuating;
it depends on export opportunity. Refining of inedible fats includes de-acidifica-
Table 1 4-2 Fatty acids, obtained in refining of various oils and fats in 1971 and
1972 in the Netherlands in 1000 kg.
1971 1972
Fatty acids from:
Coconut oil
Palmoil
Palmkernel oil
Soya oil
Marine oils
Various other vegetable oils
Animal fats
2794
7825
5317
2834
5556
3220
7367
5175
9920
6040
3632
5333
2270
5092
Totals 34913 37462
23
tion, which means that the transformation to edible fats results in a quantity of
fatty acids as by-products. However, the same fatty acids are also obtained when
vegetable oils and fats are refined. In fact, all vegetable oils are refined before they
are used for human consumption or used as raw materials in food products. To give
an impression of the quantities concerned, Table 1 42 shows the quantities of
various fatty acids, obtained in the refining, hardening, etc. of animal and vegetable
oils and fats in the Netherlands (1971 and 1972).
To avoid any misunderstanding, it be here pointed out that the totals for fatty acids
produced in 1971 and 1972 were much larger than the quantities listed in Table 1
42. Actually, considerable quantities of beef fat, mostly ranged as tallows, are also
used for the production of fatty acids. From the total thus available, a rather high
percentage is usually exported, as will be seen in Table 143.
Table I 4- 3
1969
1970
1971
1972
Production and
1972) in 10* kg.
Production
106
93
106
105
export of fatty
Export
87.7
84.3
79.4
81.2
acids in the Ne
Difference
18.3
8.7
26.6
23.8
In fact, the by-products of the fat refineries (see Table I 4-2) are not pure fatty
acids; they contain at least 25% fat or oil. Generally speaking, the composition of
these products is: MIU max 3%; FFA 60-70% and fat 30-40%.
In Table I 44 some characteristics are given for a number of fatty acids of
vegetable and animal origin.
As will be explained in Chapter II, the calculation of the percentages of fat and
fatty acids is based on the assumptions that the amount of mono- and diglycerides
can be neglected, and that the average molecular weight of the free fatty acids is the
same as that of the fatty acids bound in the triglycerides.
As has been discussed in item I 3, the most important property from a technical
point of view is the temperature at which the fats or fat products are in the liquid
phase. Table I 4- 4, therefore, shows the melting points, too. Table 1 4- 4,
specifically reveals that the melting point of the fatty acids obtained in refining
coconut oil is very low. This raw material is therefore suitable for mixing with fatty
acids of higher melting points. For example, a mixture of 67.5% animal fatty acids
24
Table 144 Characteristics of some fatty acids, obtained from two refineries in
the Netherlands
Fatty acids Coconut Palm Soya Arachis Animal
from: oil oil oil oil fats
Saponificat-
ion value 246.2 201.5 155.8 192.3 201.9
mg KOH/g
Acid value
mgKOH/g 180.5 150.7 133.9 116.3 123.7
Average mol.
weight 222.1 273.6 278.1 282.6 265.3
Fat content
g/lOOg 27.6 25.9 31.5 40.1 38.8
FFA content
g/lOOg 71.6 73.5 66.5 58.7 58.6
MIU content
g/lOOg 0.8 0.6 2.0 1.2 2.6
melting point
C 23.5 50.0 30.5 31.5 35.0
and 32.5% of the fatty acids from coconut oil melts at 30 C, and so does a mixture
of 20% fatty acids from palm oil and 80% of coconut fatty acids. These examples
illustrate the possibility, mentioned in 1.3., of utilizing mixtures of raw materials in
order to avoid the application of an emulsifier. They confirm, at the same time,
that the use of animal fat and palm oil fatty acids may be restricted by the
circumstance that sufficient amounts of coconut fatty acids will not be available for
mixing in the required proportions.
25
CHAPTER n
VARIOUS ANALYTICAL METHODS
1. Fat analysis
1.1. Moisture, impurities and unsaponifiable matter (MIU)
The determination of these quaUty characteristics was performed according to the
methods described in lUPAC Standard Methods (1966). The moisture percentage is
in fact the matter volatile at 105 C, and is expressed as % m/m; the impurities
constitute the percentage of the fat that is insoluble in petroleum ether (1:5) and
the unsaponifiable matter is determined by means of the lUPAC diethyl ether
method (1966).
1.2. Saponification value (SV)
The determination of this identity characteristic was carried out according to the
method described in lUPAC Standard Methods (1966). The result is expressed as
the number of mg KOH per g fat and will be referred to as SV.
1.3. Fatty acid content and fat content
These characteristics were derived from the acid value (AV), obtained by the
method described in lUPAC Standard Methods (1966), the saponification value
(SV) and MIU (see II 1.1. and 1.2.).
Supposing that a fat with an MIU value of m % contains a % fat and b % free fatty
acids we can express the saponification value and the acid value in these
percentages, assuming that the average molecular weight of the free and esterified
fatty acids has the same value (M) and that the amount of mono- and diglycerides
can be neglected.
The following formulas will be the result:
56.1 xa 56.1
^^ ~ M+12.67 "^ M (1)
AV = ^ (2)
m = 1 0 0 - a - b (3)
26
Eliminating M and rearrangement gives:
( SV- AV) x( l OO- m) ^ 12.7x AVx( SV- AV)
SV 561 xSV
^ ^ AVx( l OO- m) 12, 7xAVx( SV- AV)
SV 561 xSV
When the values of AV and b are known, the value of M can be calculated with
formula (2) and, next, compared with the value derived from the fatty acid
composition.
1.4. Fatty acid profile of fat products
The fatty acid profile of oils and fat products was estimated after conversion of the
fatty acids into methyl esters according to the method described in lUPAC
Standard Methods (1966).
These methyl esters were then determined by gas chromatography. The apparatus
used has the following characteristics:
Column: glass 200 cm; i.d. 0.3 cm
Stationary phase: 5% PEGA on gas chromosorb Q, 100-200 mesh;
Carrier gas: N2, 99,95% pure;
Temperatures: oven 170C, injector 210C and detector 210;
Detector: FID;
Measurement: electronic peak surface integrator.
From the fatty acids profile, the average molecular weight of the fatty acids,
present in the fat product, can be derived and this molecular weight can be
compared with the value obtained from the values of SV, AV and MIU.
1.5. Fat content of emulsions
For this determination, the method of Weibull (1892-1894) was used. The
emulsion was boiled with 1.5 n hydrochloric acid for 1 hour and the mixture then
filtered through a wet filter and washed until it was acid free (Congored paper as
indicator). After drying overnight, extraction with petroleum ether according to
Berntrop (1902) was performed for five hours. After evaporation of the solvent, the
residue was dried until constant weight was attained.
27
1 6 Emulsifier in fat products
Tlie amount of emulsifier could be estimated on the basis of its solubility in water
and in caibon tetrachloride 50 ml of the emulsion was boiled for 20 minutes with
3 35 grams of potassium hydroxide and 40 ml of ethanol, containing 0 4% v/v
amylalcohol After cooling, and adding 20 ml of a 25% m/v hydrochloric acid
solution, the mixture was extracted with diethyl ether After separation of the two
phases the water was evaporated and the residue extracted with carbon
tetrachloride The solution thus obtained was dried until constant weight was
reached
1 7 Pesticides in fat products
The methods used for determination of the chlorinated pesticides in fats and fat
products were based on the procedure described by Holden and Marsden (1969)
The sample was extracted with hexane, containing an internal standard, and, after
cleaning up on one or two alumina columns, the extract was analyzed by gas
chromatography with electron capture detection
2. Yeast analysis
2.1. Yeast concentration of culture liquid
2 1.1. Introduction
The purpose of this determination, carried out at mtervals, is to follow and control
the conversion of fat into yeast The simplest way is to count the viable yeast cells,
but this method has two disadvantages The first is that the results of the counts are
available only after three to five days The second difficulty is the formation of
clusters of cells, but this can be overcome by strongly shaking the culture liquid and
changing pipettes at least twice in making dilutions The determination of the cell
concentration by measuring the light extinction of the diluted culture hquid at 660
nm is no doubt the most rapid method, but it has the disadvantage that other
absorbing particles are interfering The problem that the fats are practically
insoluble in water was overcome by using a mixture of propanol-2 (IPA), hexane
and hydrochloric acid for making the dilutions
In view of the fact that, finally, it is the weight of the biomass that counts for the
calculation of the yield, it was sensible to relate these methods to one another In
doing so it was found that during the last hours of fermentation the number of cells
did not increase whereas the biomass did increase in weight Therefore in the
following the results obtained by the colorimetric method, related to counting,
28
will be given. The estimation of the biomass was performed by weighing the product
obtained after centrifuging the culture liquid, and washing and freeze-drying of the
wet yeast paste obtained.
2.1.2. Yeast count
Starting with 1 ml of the culture liquid, suitable decimal dilutions were made. In
most cases the 10'*, 10'* and 10'^ dilutions were used. From each of these
dilutions, 1 ml was plated in duplicate into an oxytetracycline glucose yeast extract
agar cooled to approximately 47C. The inoculated plates were then incubated for
three to five days at 22C to 24C before counting. From the six observations made
the average value of the yeast concentration, expressed as cells/ml, was derived.
0 7
0 6
OS
0 4
a3
0.2
ai
5 10 1^ 20 25 30 35 40 45 50 55 60 x 10* cells/ml
Figure II 2-1 Relation between extinction at 660 nm and cell concentration in cells/ml
Us/ml
10*
7
14
21
28
35
42
56
^660
0.14
0.26
0.38
0.50
0.60
0.68
0.80
cells/ml
x l 0 6
4
8
12
16
20
30
40
50
E660
0 06
0.14
0.22
0.30
0.37
0.52
0.66
0.77
cells/ml
x l 0 6
10
15
20
25
30
35
40
45
Eeeo
0.18
0.28
0.37
0.45
0.53
0.60
0.68
0.72
29
2 13 Colorimetric method
The test solution was prepared by diluting 1 ml of the culture hquid with 9 ml of a
mixture of 885 5 ml IPA, 111 ml hexane and 3 5 ml 4 n hydrochloric acid At very
high concentrations, the culture liquid was diluted 20 times
Tlie blank was prepared by diluting the fermentation liquid before inoculation with
the mixture of IPA, hexane and 4 n hydrochloric acid mentioned The
measurements were made at 660 nm with an Engel colorimeter, combined with a
hghtspot galvanometer AL 4 Figure 11 2 1 gives the relation between the light
extinction and the cell concentration, on the basis of the results of a number of
experiments in which the yeast concentration was estimated simultaneously by
means of both methods
As will be seen in Figure II 2- 1, the deviations of the various results are not very
great, especially not in the straight part of the curve For that reason most of the
determinations were made with dilutions of yeast suspensions in that part of the
curve (below Eeao =0 5)
2 2 Moisture content and ash content
The determination of the water content was carried out by drying with phosphor
pentoxide under vacuum at 70C according to the method described by Van der
Kamer (1949) The ash content was determined at 550C according to the AOAC
Method (1965) described in its item 13006
2 3 Crude protein content and lipid content
For the crude protein content (including nucleic acids, peptides free ammo acids
etc ), the method of Kjeldahl for the determination of nitrogen was followed in a
modification described by Bradshaat (1965) The crude protein was calculated from
the nitrogen content by multiplying with 6 25
The lipid content was estimated by means of a modification of the Weibull method
described in AOAC Methods (1965) After acid destruction of the sample with
hydrochloric acid, it was washed, dried and extracted with light petroleum After
evaporation of the light petroleum the residue was dried and weighed
2 4 Polysaccharide content and crude fibre content
The determmation of the polysaccharide, expressed as glucose, was based on the
method of Sachse (1878) After acid destruction of the sample, and clearing with
Carrez solutions I and 11 of the filtrate, the amount of reducing sugars was
30
estimated according to the Luff method (see Schoorl 1929). The filtrate was, in
addition, examined on pentoses by means of tfiin layer chromatography (TLG).
The crude fibre determination was performed by means of the method of Van der
Kamer and Van Ginkel (1952). Starch, protein and Hpid were made soluble with
the reagent of Scharrer and Kiirschner. After extraction with diethyl ether the
residue was dried and weighed.
2.5. Amounts of emulsifier and pesticides
The determination of emulsifier was carried out as is described for fat emulsion,
using a yeast suspension in water.
For the estimation of chlorinated pesticides, the dried yeast was ground to a fine
powder and extracted with hexane (with internal standard) in a conical flask by
shaking for one hour on a mechanical shaker. After cleaning up on a small alumina
column, the extract was analyzed by gas chromatography with electron capture
detection as described in II 1.7.
2.6. Elementary analysis and gross formula
The nitrogen content of the yeast was also estimated according to a modified
Dumas method, using a microcombustion apparatus, type Hosli. The carbon and
hydrogen were determined as is described by Buys and Schroder (1970) and the
oxygen according to the method of Romer (1972). These determinations were
carried out at the Institute for Organic Chemistry TNO. On the basis of the
results, the gross formula of the yeast was calculated.
2.7. Amino acid spectrum and fatty acid profile
In his doctoral thesis. Slump (1969) describes the methods of analysis for the
various amino acids and we used these methods, slightly modified, for determina-
tion of the amino acid spectrum.
For the determination of the fatty acid profile, performed as described in II 1.4.,
the Hpids used were obtained by means of the method described in II 2.3.
2.8. Vitamin composition
2.8.1. Introduction
The following will present a general survey of the procedures for the determination
31
of a number of vitamins of the B group. In the literature cited, the respective
microbiological methods are given in detail and in practice only slight modifications
of these methods were applied.
2.8.2. Thiamine
The determination of vitamin Bl was made using the ATTC strain 9338 of
Lactobacillus fermenti. The samples were heated with 85 ml 0.1 n sulphuric acid
for 30 minutes at 120C and after cooling treated with Taka diastase at 45C for
105 minutes (see Sarett and Cheldelin, 1944).
2.8.3. Riboflavin
The estimation of vitamin B2 was performed by means of the ATTC strain 7469 of
Lactobacillus easel after extraction of the samples with 0.1 n hydrochloric acid at
120C for 15 minutes (see Snell and Strong, 1939).
2.8.4. Pyridoxine
The vitamin B6 was determined with the ATTC strain 9080 of Saccharomyces
carlsbergensis after extraction of the samples with 0.4 n sulphuric acid at 120C for
1 hour (see Atkin et al., 1943).
2.8.5. Niacin
The determination of nicotinamide was carried out with Lactobacillus plantarum
strain ATTC 8014 after extraction of the samples with 1 n hydrochloric acid at
120C for 20 minutes (see Barton-Wright, 1944, 1945).
2.8.6. Panthotenic acid
For the determination of this vitamin Lactobacillus plantarum strain 8014 was used
after treatment of the samples with Taka diastase and papain in a buffer at pH 4.5
at 45 C for 4 hours (see Sheggs and Wright, 1944).
32
2.8.7. Folic acid
This vitamin was determined using the ATCC strain 7469 of Lactobacillus casei,
after extraction of the samples with phosphate buffer pH 6.1 at 120C for 10
minutes, followed by digestion with chicken pancreas at pH 7.2 and 37C for 24
hours (see Flynn et al. 1951).
2.8.8. Inositol
The determination of inositol was performed with Saccharomyces carlsbergensis
strain ATTC 9080, after extraction of the samples with 0.6 n sulphuric acid at
120C for 2 hours. In the medium used by Atkin et al. (1943) the inositol was
replaced by 50 meg pyridoxine in 100 ml of the basal medium.
2.9. Elements in ash
2.9.1. Introduction
The various elements were determined by means of atomic absorption spectro-
photometry (AAS), using a Perkin Elmer 303 spectrophotometer. The test
solutions were prepared with an internal standard after destruction of the samples.
For some elements this destruction was effected by ashing at 550C, using the
method described in AOAC Methods (1965), and for others it was carried out by
boiUng the sample with a mixture of concentrated sulphuric acid and nitric acid,
followed by boiling with distilled water to remove the excess of nitric acid. The
methods applied were modifications of the methods described in detail in the
literature cited below.
2.9.2. Calcium and magnesium
After dry destrucfion the method of Belcher and Brooks (1963) was applied, using
a solution of 5% lanthanum chloride in 25% hydrochloric acid for calcium but not
for magnesium.
2.9.3. Copper and iron
These elements were determined after wet destruction, using the method described
by Allen (1959).
33
2.9.4. Manganese and zinc
After wet destrucfion these elements were estimated according to the method
elaborated by Buchanan and Muraoka (1964).
2.9.5. Nickel
After dry destruction of the sample this element was determined by the method
described by Kinson and Belcher (1964).
2.10. Content of ribo-nucleic acid (RNA) and purine base content
The determination of yeast RNA was carried out by preparing a test solution, the
extinction of which at 260 nm and 300 nm was compared with that of a standard
solution of yeast RNA (Boehringer).
The test solution was obtained by the extraction procedure according to Schmidt
and Tannhauser (1945). However, the fat extraction was omitted.
For the preparation of the purine bases, the sample was hydrolyzed with 1 n
sulphuric acid in sealed tubes (Vischer and Chargaff, 1948).
The determination of adenine and purine was done spectrophotometrically after
separation, using an ion exchange column according to the procedure described by
Bonnelycke (1969).
3. Sanitary analysis of yeast
3.1. Introduction
The test described in the following are of two different types, namely a series of
counting tests and a series of present/absent tests.
It is a rather arbitrary decision whether a test should be done by the former or latter
method. For Enterobacteriaceae, for instance, a present/absent test is proposed in
Guideline No 5 of the Protein Advisory Group (PAG) (1969 - 1972) but, of course,
plate counting gives more information when bacteria of this type are present.
In Guideline No 5, the following specifications or conditions are suggested for the
present/absent tests:
Salmonella absent in 25 g
Arizona absent in 25 g
Shigella absent in 25 g
Escherichia coli absent in 10 g
Staphylococcus aureus absent in 1 g
Enterobacteriaceae absent in 0.1 g
34
For the tests concerned, the amounts of samples have to be chosen in agreement
with these specifications.
3.2. Total counts of aerobic bacteria and spores
Starting with 10 g/100 ml peptonsalt suspension (PES), suitable dilutions were
made and 1 ml thereof was plated into a tryptone pepton agar, according to the
method described by Wetzler et al. (1962), and incubated for 3 days at 3I 1C. In
these tests the total number of aerobic bacteria in so far as these are present in the
samples as single cells, and the number of spores of such bacteria in so far as the
spores germinate under the experimental conditions applied, are obtained.
In the case of spore counting, the tubes, containing the suspensions, were heated in
a waterbath until the thermometer in a parallel test tube, containing merely water,
reached a temperature of 79C. Then the tubes were transferred immediately to a
waterbath of 80C, in which they were kept for exactly 1 minute. The tubes were
chilled rapidly to approximately 10C in a beaker supplied with running tap water.
With the suspensions in the tubes, the same procedure was performed as described
for the aerobic bacterial count.
3.3. Enterobacteriaceae count
For this test we used the method described by Mossel (1970). A suspension
containing 10 g yeast in 100 ml tryptone soya pepton glucose broth was shaken at
20C for two hours to resuscitate sublethally impaired cells. Subsequently, a
number of decimal dilutions were made and 1 ml of these was plated on violet red
bile glucose agar, at a temperature of 47C. After soHdification the plates were
covered with a second layer of approximately 20 ml violet red bile glucose agar to
suppress strictly aerobic gram-negative rods. After incubation at 36C for 20 h, the
number of purple colonies surrounded by purple haloes of precipitated bile salts
was counted.
3.4. Escherichia coli present/absent test
This test was performed after resuscitation of sublethally impaired cells for 2 h at
20 C in a tryptone soya broth suspension of 10 g yeast/100 ml. The suspension
thus obtained was enriched in 100 ml of double concentrated brilliant green bile
lactose broth at 30C for 24 h. Cultures, that did not show any gas production were
discarded. The gas-positive enrichments were subcultured onto McGonkey agar no
3 and incubated overnight at 440.1C (Thomas 1971). Typical lactose-positive
colonies were examined by means of a modified Eijkman test (McKenzie et al.
1958).
35
3.5. Salmonella present/absent test
Following the method of Edel and Kampelmacher (1969), suitable aliquots of
resuscitated cell suspensions were enriched in MullerKaufmann broth, incubated
at 420.5C. After 24 h and 48 h the suspensions were streaked onto brilliant green
phenol red lactose sucrose agar. After incubation at 37C for 24 to 30 h, the
colonies surrounded by red haloes were examined for the following properties:
fermentation of glucose, absence of j3galactosidase and urease activities, presence
of lysine decarboxylase and oxidase and, finally, specific agglutination reactions.
3.6. Counts of Clostridium group and Clostridium perfringens
For this test, plastic pouches were used as described by De Waart and Smit (1962).
Suitable dilutions of the cell suspensions concerned were mixed with sulphite iron
polymyxin agar and poured in the plastic pouches. These were sealed and incubated
for24to48hat3rc.
For selecting Clostridium perfringens in principle the same method was applied.
This time, however, sulphite iron polymyxin neomycin agar was used and the
incubafion occurred at 460.2C for 24 to 48 h (Mossel and De Waart, 1968).
3.7. Lancefield group D Streptococci count
According to the method of Pike (1945), 0.1 ml of appropriate cell dilutions was
spread onto crystal violet azide agar (Streptosel BBL, Md USA). After 2 days of
incubation at 37C, a number of typical colonies (at least three) were examined for
the following characteristics: catalase negative cocci, usually in short chains,
capable of rapid growth at 45C and acid formation from glucose without gas.
3.8. Staphylococcus aureus present/absent test
1 ml of serial dilutions was added to 20 ml of tellurite glycine broth (Giolitti and
Cantoni, 1966) in tubes; these were sealed with sterile paraffin and incubated at
37C for 48 h. The presence of Staphylococcus aureus in blackened tubes was
confirmed by plating 0.1 ml onto Baird-Parker's tellurite glycine eggyolk agar,
followed by testing both eggyolk positive and negative isolates for coagulase
activity (De Waart et al., 1968).
36
3.9. Moulds
Suitable decimal dilutions were plated on oxytetracycline glucose yeast extract agar.
Incubation occurred for 35 days at 22C 2C before counting the mould
colonies appearing on the plates.
4. Nutritional testing of yeast
4.1. Protein calories per cent (PCP)
The energy available to the human body is the gross energy of the foodstuff minus
the pertinent losses in urine and faeces.
According to At water (Merill and Watt, 1955), these losses are allowed for by
applying 17 kJ per g for protein, 38 kJ per g for fat and 18 kJ per g for
carbohydrates, expressed as monosaccharides.
To calculate the PCP which truly should be called 'protein Joule per cent' the
exact figures have to be determined for protein content, fat content and
carbohydrate content expressed as monosaccharides.
When the carbohydrates are not determined by analysis, but calculated by
difference, the factor should be 16 kJ instead of 18 kJ (FAO/WHA report No 52,
1973).
Southgate and Durnin (1970) have compared the calculated energy value with that
obtained by bomb calorimetry analysis, and they found a quite good agreement
between both values. The PCP can be calculated as the percentage protein energy of
the total energy of the foodstuff.
4.2. Chemical score and EEAIndex
4.2.1. Introduction
The nutritional value of proteins is due to the amino acids they yield on digestion.
The relation between the chemical composition of the amino acid mixture found
after hydrolysis and the biological activity of the amino acids concerned may be
disturbed by the fact that the amino acids in dietary proteins may only be partly
available biologically. Particulariy after heat treatment this factor may be of great
influence.
The determination of the amino acid spectrum is mentioned in Chapter II 2,7. The
nutritional value based on this spectrum will be discussed below.
37
4.2.2. Chemical score
Mitchell and Block (1946) introduced the concept of a 'Chemical score' for
proteins on the basis of the amino acid composition of the proteins concerned.
They used whole egg protein as a standard, and expressed the concentration of each
essential amino acid (g per 16 g N) in the product under investigation as a
percentage of the concentration of the same amino acid in the whole egg protein.
The lowest value obtained was taken as the 'Chemical score' because the limiting
amino acid determines the nutritional value of the product.
On the basis of the determination of the Chemical score of a series of foods and the
determination of the net protein utilization of the same foods, Mitchell and Block
(1946) came to the following equation: y = 102 - 0,634X
in which y means the biological value, and X the percentage amino acid deficit
in comparison to the standard protein.
They also pointed out that there is no relation between amino acid deficits and
coefficients of digestibility. Miller et al. (1965) found that a heat treatment of dried
cod fillets reduced their Chemical score by 15%. However, the reduction of the
nutritional value for rats was nearly 40%. It is therefore evident that the Chemical
score has its limitations as a measure of the biological value of proteins. Since the
pubUcation of Mitchell and Block (1946), other investigators have suggested to
abandon whole egg protein as a standard and to replace it by a synthetic amino
acid mixture, which is more efficiently used by young rats (Bender, 1958, 1965).
Table II 41 Essential amino acid patterns (g/16 g N) considered as a standard for
calculating of the Chemical score
F A O - s t a n d a r d s
Provi-
sional
1957
4.3
4.9
4.3
Cow's
milk
1973
4.7
9.5
7.8
Hen's
egg
1973
5.4
8.6
7.0
Provi-
sional
1973
4.0
7.0
5.5
5.8
4.3
2.9
1.4
4.3
10.2
3.3
4.4
1.4
6.4
9.2
5.6
4.7
1.7
6.6
6.0
3.5
4.0
1.0
5.0
38
Isoleucine
Leucine
Lysine
Phenylalanine
Total aromatic
amino acids
Total sulphur
amino acids
Threonine
Tryptophan
Valine
Block and
Mitchell
1946
8.0
9.2
7.2
6.3
10.8
6.5
4.9
1.5
7.3
Bender
1965
4.3
7.8
5.2
4.9
- . -
4.7
4.1
1.0
5.0
Oser
1951
7.7
9.2
7.0
6.3
- . -
6.4
4.3
1.5
7.2
The FAO (1957, 1970, 1973) finally has suggested to adopt a provisional amino
acid pattern as a reference standard for calculation of the Chemical score. Table II
41 gives the various standards for the essential amino acids.
It is evident that the slope of the graph from which the equation BV = 102 - 0,634X is
derived will change considerably when, instead of the values of Mitchell and Block
(1946) the 1973 values of the FAO are used. In most cases the amino acid
isoleucine as a limiting amino acid is replaced by the sulphurcontaining amino
acids when the FAO standards are applied. The ehmination of isoleucine as Umiting
amino acid in human diets is in agreement with the biological evidence that the
sulphur-containing acids are the most commonly limiting amino acids in these diets.
The Chemical score of the yeast grown on fats and fat products will therefore be
calculated with reference to the provisional amino acid scoring pattern proposed by
the FAO (1973) as given in Table II 4 - 1 .
4.2.3. Essential amino acid index (EAA-Index)
Oser (1951) has introduced the Essential Amino Acid Index (EAA-Index); it is
derived from the formula:
I n 100 a
log EAA-Index = 2 log () n
e
in which n is the number of the essential amino acids taken into consideration and
a/a is the ratio of the individual amino acid concentrations, respectively, of the
food protein (a) and the egg protein (ae). The maximum value of the ratio taken
into account is 1.
The difference between the Chemical score and the EAA-Index is quite evident. For
example, the amino acids concerned are not the same. Histidine and arginine are
being considered in the case of the EAA-Index, whereas tyrosine is left out.
However, the values of the EAA-Index do not vary markedly whether histidine and
argnine are included or not.
An interesting equation is that giving the relation between the biological value (BV)
and the EAA-Index of a protein: EAA-Index = 0.877 BV + 7.38
Oser (1951) claims that statistical analysis of the relation between the EAA-Indices
and the biological values shows that the EAA-Index can be used to predict the BV-
value of a protein within 5 to +11 per cent.
In Table II 4- 1 the ae-values used by Oser (1951) are also given; it is evident that
these figures differ considerably from the values suggested by the FAO (1973).
Nevertheless, in addition to the Chemical score the EAA-Index can to a certain
extent be considered as a measure of the nutritional value of a food protein.
39
4.3. Protein efficiency ratio (PER)
4.3.1. Introduction
Eggum (1965) proved experimentally that different amino acids from the same
protein source can be digested in different ways. For that reason reliable figures for
the biological availability of individual amino acids cannot be obtained from the
digestibility estimated by analysis of total nitrogen. This makes a combination of
the chemical analysis with biological tests imperative.
The most simple biological screening test is that of the protein efficiency ratio
(PER), which uses the growth of laboratory animals as a parameter.
However, it is known that growth is not always a reliable measure of protein
synthesis. Moreover, the results of the efficiency tests vary with the level of protein
in the diet and, finally, the results differ with the food intake, which, in turn, is
dependent on palatability and other factors. These facts present some drawbacks in
the interpretation of the PER as a measure of the nutritional value of a foodstuff.
There are two more difficulties. Firstly, the casein in the reference diet can be
different in the various experiments performed all over the world. Secondly, the
response of the rats suppUed by different breeders is not the same througliout, even
if the animals are of the same species. For the reasons mentioned above, the
experimentally found PER-value is multiplied by a certain factor relating it to a
standard value for the casein diet (see II 4.3.2.).
4.3.2. The performance of the test
The determination of the PER-value is carried out as described by Derse (1958).
The dry sample is incorporated as the only source of protein into a complete diet at
a level providing 10% crude protein (N x 6.25). A diet which contains casein, also at
10% crude protein level, is used as a standard diet. Both diets are each fed to ten
newly weaned male rats kept individually in wire screen cages. Food and tap water
are provided ad libitum. Food consumption and body weight are recorded at
one-week intervals during a period of 4 weeks.
From the amount of food consumed, and the gain in bodyweight of each rat, the
PER is calculated. The value of the PER obtained with the casein diet is converted
to 2.50, and the experimental value for the other diet is multiplied with the same
conversion factor, thus relating the PER to the standardized value of 2.50 for
casein. Because of the drawbacks mentioned in II 4.3.1., and the lack of sufficient
amounts of yeast, the PER-value was not determined.
40
4.4. Net protein utilization (NPU), true digestibility (TD)
and biological value (BV)
4.4.1. Introduction
In addition to the chemical composition and the results of biological screening tests
more quantitative information is needed in order to characterize the nutritional
value of a protein foodstuff. In fact, the percentage of absorbed nitrogen used for
the synthesis of body protein is the true biological value (BV). This can be
measured when metabolic and endogenous losses are taken into account.
Thomas (1909) and Mitchell (1923) have expressed this in the formula:
_ I - ( F - F k ) - ( U - U k ) _ B- Bk
I - ( F - F k ) I - ( F - F k )
where:
I = nitrogen intake
F = faecal nitrogen
Fk = endogenous faecal nitrogen i.e. on nitrogenfree diet
U = urinary nitrogen
Uk = endogenous urinary nitrogen i.e. on nitrogenfree diet
B = body nitrogen after intake of testprotein diet
Bk = body nitrogen after intake of nitrogenfree diet
A modified method, known as the carcass nitrogen method, was introduced by
Bender and Miller (1953) in order to get a measure of the BV. According to this
method, the BV is not measured directly but derived from the percentage of
consumed nitrogen used for the synthesis of body protein, and the percentage of
consumed nitrogen not excreted in the faeces. Applying this principle they have
introduced two new formulas:
R Ri
(a) NPU= j^xlOO
I - (F-Fk)
(b) TD = ^ xlOO
where the designations are as given above.
The net protein utilizafion (NPU) value is obtained by comparing the nitrogen
content in the carcasses of rats, fed with the test diet, with the nitrogen content in
the carcasses of rats on a nitrogen-free diet.
In fact this value is a product of digestibility and biological value. After
determination of the digestibility, the biological value can be calculated according
NPU
to the formula: BV = - ^ x 100.
41
The digestibility is called true digestibiUty (TD) when the value of the faecal
nitrogen is corrected for endogenous nitrogen losses; this can be done by measuring
the amount of nitrogen excreted in the faeces of animals fed with a nitrogen-free
diet.
It is generally assumed that rats have a standard endogenous nitrogen value and that
their carcasses do not become oedematous within a test period often days.
4.4.2. Performance of the test
The determination of the NPU value is based on a publication of Miller and Bender
(1955).
The dry test sample was incorporated as the only source of protein into a complete
diet at a level providing 10% crude protein (N x 6.25). A diet containing no protein
was used in order to take into account the endogenous nitrogen losses. The 2729
days old rats were kept in groups of four in wire screen cages. During a pre-test
period of seven days, the rats received a stock diet and tap water ad libitum. The
body weights were recorded during these days and abnormally low weightgaining
animals were replaced.
Both diets were fed ad libitum to three groups of four preselected rats (usually
two males and two females). The food consumption was recorded and the total
amount of faeces was collected from each group of twelve. After an experimental
period of 10 days, the following determinations were made:
(a) amount of nitrogen consumed; (b) amount of nitrogen present in carcasses; (c)
amount of nitrogen excreted in the faeces.
We determined the nitrogen content of the carcasses indirectly from the total
content of bodywater (found as a loss of body weight after three days' drying at
105 C) because in previous experiments the relation between the contents of
nitrogen and body water had proved to be constant.
The amount of nitrogen in the faeces was estimated after dehydration, weighing and
grinding of the combined faeces pellets of the twelve rats on the same diet. The
carcass nitrogen was derived from the water content of the three groups of four rats
separately.
In order to obtain a value for endogenous nitrogen, which is necessary to calculate
the true instead of the apparent digestibility of a food stuff, the faeces of two groups
of rats, fed on different diets, have to be compared. It can be expected, however,
that the food intake as well as the amounts of faeces produced are not the same for
both groups. For that reason the nitrogen, present in the faeces of the group fed on a
nitrogen-free diet, must be corrected before it can be substracted from the nitrogen
present in the faeces of the test group. This correction is made by multiplication
with a factor equal to the ratio of food intake of both groups.
The method is based on the consideration that in most experiments there is a constant
relation between food intake and the amount of faeces, the latter being an
42
indication of the digestible part of the food. If this relation is different, however, it
might be more realistic to apply a correction factor equal to the ratio of the
amounts of faeces in the two groups.
5. Safety evaluation of yeast
5.1. Introduction
Although the yeast we produced is intended to be used only in animal feeds,
extensive toxicological testing is necessary. In the first place, the animals fed with
diets containing yeast should have a good performance, growth and food efficiency.
Otherwise, application of yeast as a common and normal component of animal
feeds cannot be advisable. Moreover, it is evident that any toxic or in other way
harmful substance present in fodder yeast, may be stored and accumulated in the
edible products derived from farm animals consuming the yeast (meat, fat, milk,
cheese, etc.). In such a case, the harmful substance implies a hazard to human
health. Therefore, in fact, the same precautions have to be taken as in the
case of products intended for consumption by animals as with additives
included directly in human food.
The safety evaluation programme that should be followed is extensively described
in the PAGguideline No 6 (1972). At the International Symposium on Single Cell
Protein (SCP)Standardization, Evaluation and Safety, De Groot (1972) has propo-
sed guidelines for minimum requirements to be fulfilled before SCP should be used in
animal feeds. These requirements are based not only on the PAG guidelines men-
tioned, and on a review by Oser (1970), but also on the standard procedures for the
toxicological examination of food additives and contaminants as applied by CIVO-
TNO to the safety evaluation of SCP (De Groot, 1971). In the next few paragraphs,
the procedures that should be applied to meet these requirements will be described
briefly irrespective of the fact that, with the yeast described in this thesis, only the
screening test was done. There are some reasons for doing so.
Firstly, this thesis covers only one part of the total study, namely the laboratory scale
production. For the pilot plant production, however, the same methods of analysis
will be applied but the safety evaluation should go as far as studies with farm
animals before a definitive judgement of the yeast can be given. The methods
concerned are, therefore, mentioned in view of further investigations to be done.
Secondly, SCP production processes on various raw materials have been developed
and some of them are already in the final stage. It may therefore be of some help
for other investigators when the methods for nutritional and toxicological tests are
published together. Thirdly, there is tendency to overestimate the importance and
the necessity of studies with farm animals. But it is evident that such studies
are useful only when it has been proved by extensive toxicological work performed
according to the methods described in this chapter, that the product concerned is
safe in every respect.
43
5 2. Screening test
This is a simple subacute feeding test with 10 male and 10 female weanling rats on
a diet containing the fodder yeast at the highest level that does not significantly
disturb the nutritional balance For yeast this level may vary between 40% and 60%
by weight giving a protein content of the diet of approcimately 30% m/m
Usually attention has to be paid to the ratio of Ca/P, because of the high P-level of
SCP In the test done with the yeast mixture, the diets and tap water were provided
ad libitum, and the following processes and properties were examined growth,
food intake, haemoglobin content, gross pathology and weight of liver and kidneys
In addition the organs mentioned were examined microscopically
Tlie test duration was 28 days and food intake and growth were measured weekly
In general the test duration is 3-6 weeks, and, when with the product concerned
the protein efficiency ratio (PER) test is also performed, that determination can be
considerated as a prescreening test, though the protein level in that case is only
10% m/m.
5.3. Subchronic study
This test IS done with groups of 10 male and 10 female weanhng rats on diets
containing the SCP at two or more graded levels for a period of 90 days
The observations made are food intake and growth, haematological data and
biochemical blood values i e some specific enzyme activities In the last week of
the experiment, samples of urine are examined The study is completed by careful
gross examination of all rats by autopsy
The major organs (6-10, e g livers, kidneys etc ) are weighed and a large number of
tissues microscopically examined.
5.4. Chronic feeding study
For this life-time study, groups of 30 males and females each are normally used,
but when there is any suspicion of carcinogenity, the number of animals of each sex
IS increased to 50 The study usually comprises four groups, namely three test
groups receiving different dietary levels of the SCP, and one control group
The performance of the study is comparable to that of the sub-chronic test The
same observations as those, made at the end of the subchronic test, are made in
the chronic experiment after 3, 6, 12 and 24 months The gross pathology involves
all rats in each group, but the histopathology is mostly restricted to 20 males and
20 females of the highest dose group and to those of the control group Visible
tumours, or lesions suspected of being tumours, are examined in all rats of each
group
44
5.5. Multigeneration study
In this test effects on fertility and lactation are examined. This is done with at least
two dosage groups and one control group, each consisting of 10 male and 20 female
rats each. After three months, the rats are mated within their diet group. The litters
are counted and weighed shortly after birth, and at weaning, and then discarded
(Fa). The parents are mated again, giving a second Utter that is counted and
weighed as the first litter. At weaning age, 10 males and 20 females are selected
from different litters in each group and maintained on the same diet for the
production of the next generation. These parents are treated as the first group and
give litters of a third generation. These litters are treated as their parents and at
weaning age of the second litters the experiment is completed by macroscopical and
microscopical examination of 10 males and 10 females from each diet group.
Figure II 5-1 gives a scheme of such a multigeneration study.
Figure II 5-1 Design of multigeneration study with 10 males and 20 females
5.6. Teratogenicity studies
To detect any embryonic malformations, a group of 10-20 pregnant rats are
treated with the SCP at two or more high levels during the time of organogenesis of
their embryos.
One day before term the mothers are killed. Total implantations are counted and
45
alive and dead embryos are recorded The foetuses are weighed, sexed and
examined macroscopically for malformations The soft tissues in 2/3 of the
embryos are examined, after fixation, by means of a dissecting microscope and so
are the skeletons of 1/3 of the embryos, after staining with alizarin red
5 7 Mutagenicity studies
The aim of these studies is to discover mutations indicating major genetic damage,
resulting in pre-implantation losses of non-viable zygotes, early foetal deaths and
sterility Groups of at least 5 adult male rats or mice are treated orally with the test
material of SCP at the tolerated maximum dose for 5 successive days A comparable
group of non-treated animals serves as control group
Subsequently, each male is caged together with 2 non treated virgin females at
weekly intervals for a total period of at least 4 weeks The femals are autopsied on
the 13th day after midweek of their presumable mating and scored for pregnancy,
total implants and early as well as late foetal deaths
5 8 Studies with farm animals
In addition to the tests with rats on SCP intended for use in animal feed, studies
with pigs and poultry are necessary, mainly to examine the nutritional properties
but also to discover deleterious effects
Groups of 20 pigs and 60 chickens are fed for 6 12 months with SCP at two levels
Tlie food intake and weight gain, haematology, blood biochemistry and urine
properties should be investigated, and by autopsy macroscopal examination of all
animals, the weight determination of liver and kidneys and the macroscopy of
1020 of the major organs should be performed However, it is quite clear that
studies with farm animals can never be a perfect substitute for chronic studies with
laboratory animals, because it is practically impossible to continue studeis with
farm animals for their whole life-time
5 9 Derived tests
Ferrando and Truhaut (1972) introduced a new approach to the toxicological
evaluation of SCP and called it "La toxicite de relais" Their suggestion is followed
by Russian scientists, according to the PAG Bulletin, Vol No 3 (1973) They did not
only perform extensive studies with the SCP per se but also investigated the
hygienic, physiochemical and organoleptic properties of meat, eggs and milk from
animals fed with SCP Subsequently, the meat was tested in longterm experiments
with rats and finally two feeding trials were carried out, each lasting more than six
46
months with groups of more than hundred persons, to investigate the influence of
the systematic consumption of products from animals fed on hydrocarbon yeast.
If people fed with certain particular products show unfavourable effects that can be
related to the same products, such so called derived tests can be useful but the
absence of any undesirable effects - as was the result of the Russian experiments -
does not give sufficient evidence that all the products containing SCP are harmless
under all conditions.
As a matter of fact, the products to be tested (milk, meat, organs, fat, eggs, cheese,
etc.) as well as the test conditions are chosen quite arbitrarity.
For these reasons it is questionable whether the derived test may be considered as a
method of sufficient value to become a routine method in safety evaluation of SCP.
47
CHAPTER HI
STANDARD PROCEDURE FOR FAT FERMENTATION
1. Fermentation steps
1.1. Preparation of the slant cultures
1.1.1. Introduction
The first point of investigation was to establish whether yeast grown on
GYA-slants would be an equally good inoculum for fat fermentation as yeast
grown on slants in which only fatty acid was present as a carbon and energy source.
1.1.2. GYA-slants without fat
The GYA-slants were prepared from a solution containing 20 g glucose, 5 g yeast
extract and 20 g agar dissolved in 1 1 distilled water. This solution was sterilized at
120C for 20 minutes.
1.1.3. FAslants with fat and minerals
Tlie FA-slants were prepared without glucose and yeast extract so that all the
minerals and growth factors had to be added. First 0.5 g fat was emulsified in 50 ml
water with 0.03 g Renex 650. To this emulsion was added 1.5 g agar and the
mixture obtained was sterilized at 120C for 20 minutes. Next were added 10 ml of
the traceelement solution (with phosphate), 10 ml of the salt solution (without
urea) and 1 ml of the vitamin solution as well as 29 ml sterile water. From the
emulsion thus obtained slants were prepared.
1.1.4. FAAslants with fatty acids and minerals
This third type of slants was prepared in the same way as described in 1.1.3., except
that 0.5 g animal fatty acids was used instead of 0.5 g fat and 0.03 g Renex 650.
1.1.5. GRAslants with glucose but without fat or yeast extract
The fourth type of slants was prepared by mixing 0.5 g glucose with 0.03 g Renex
650 in 50 ml water.
48
After adding 1.5 g agar, the solufion was sterilized at 120 C for 20 minutes and,
after cooling to 60C, were added 10 ml of the salt solution (without urea), 10 ml
of the solution containing the trace elements (with phosphate) and 1 ml of the
vitamin- -solution diluted with 29 ml sterile water.
1.1.6. RAslants without fat, glucose or yeast extract.
The fifth type of slants contanied Renex 650 in order to check whether this
emulsifying agent itself could be used as a carbon source for yeast. The slants were
prepared as described in 1.1.5., but without glucose.
1.1.7. Testing of the slants with three yeast strains
The various slants were inoculated with a pepton-salt solution of yeast cells that
were harvested from 4 GYA slants which had been inoculated for 75 hours at 30C.
The selected strain (1) and the CBS strains 599 (2) and 5611 (3) were used in this
experiment. The slants were observed after intervals of three, five and seven days.
The results are given in Table III 1 1.
From the results given in Table III 1 1 the conclusion may be drawn that the
yeasts are growing equally well in presence of the yeast extract (GYA) as in
presence of the mbcture of salts, trace elements and vitamins (GRA). The cultures
on the slants with fat, fatty acids and Renex 650 without glucose grew very slowly,
or not at all. This means that Renex 650 does not serve as a carbon source for the
yeast strains tested and that there is no reason to use yeast grown on fat or fatty
acids for inoculation.
1.1.8. Experiments with yeast from various slants
The yeast cells produced by the selected strain on the slants after 7 days were used
as an inoculum for shake flask experiments. With 10 ml pepton-salt solution the
yeast cells were transferred into a shake flask containing 15 ml fat emulsion, 15 ml
salt solution (with urea). 15 ml trace element solution (with phosphate), 1.5 ml
vitamin solution and 93.5 ml sterile water. After 82 hours of shaking, the
concentration of the biomass was measured colorimetrically. These are the results:
Slants: GYA FA FAA GRA
Cells/ml 5.5x10^ 2.4x10* 3.6x10* 4.8x10*
On the basis of these figures it was decided to use GYA-slants, made as described
in 1.1.2., for the preparation of an inoculum for the shake flask experiments.
49
Table III 1 1 Growth of yeast strains 1), 2) and 3) on various slants
Observation period:
GYA-slants
FA-slants
FAAslants
GRA-slants
RA-slants
1)
2)
3)
1)
.2)
3)
1)
2)
3)
1)
2)
3)
1)
2)
3)
2 days
+
-H-
+
(+)
(+)
-
(+)
(+)
-
+
++
+

(+)

3 days
+++
++++
++
+
+
(+)
+
(+)
-
+++
++++
++
(+)
(+)

5 days
-H-l-l-
-H-t-+
+++
++
+
(+)
+
(+)
(+)
-l-l-^-^
++++
+++
(+)
(+)
(+)
7 day
-^-l-^-^
-I-I-H+
-h-H-
++
++
(+)
++
++
+
++++
++++
+++
+
+
(+)
Explanation:
1) selected strain Endomycopsis lipolytica; 2) CBS strain 599 Saccharomyces
lipolytica; 3) CBS strain 5611 Candida maltosa;
GYA-slant: glucoseyeastagar-slant; FAslant: fatagarslant; FAAslant:
fatty acidagarslant;GRAslant: glucoseRenex-agarslant; RAslant: Renex
agar-slant;
no growth; (+) growth questionable; + minute growth; ++ clear growth; +++ good
growth; ++++ abundant growth
III 1.2. Procedure for the shake flask cultures
1.2.1. Introduction
The general aspects of every shake flask experiment that have to be considered are:
shape and volume of flask, quantity of culture liquid and its viscosity, the
circumstance that no ingredients can be added during shaking, and the fact that for
aerobic growth the oxygen supply will be a limiting factor. In the following each
of these aspects will be considered.
50
1.2.2. The incubator shaker
The available apparat us is a gyrot ory shaker t hat has a st roke of 2.5 cm and a speed
range from 50 t o 40 0 rot at i ons per mi nut e. The t emper at ur e in this shaker can be
regulated bet ween 0 C and 60 C wi t h a cont r ol t ol erance of 0 . 5C. Aft er some
preliminary exper i ment s, a st andard speed of 160 rpm was chosen because at this
speed no foami ng occurred when t he liquid vol ume was not more t han 20 % of t he
flask vol ume.
The t emperat ure chosen for all the shake flask exper i ment s was 30 C.
1.2.3. pH Regulation
For fat f er ment at i on it is essential t o keep t he pH value bet ween 4 and 5 because
no pr oduct i on of bi omass occurs at pH values bel ow 4. Therefore, t he probl em of
how to keep t he pH above a value of 4 had t o be solved.
Actually, t he initial pH is 5 because, as will be expl ai ned in III 2. 5, t he salt sol ut i on
and the sol ut i on cont ai ni ng t he trace el ement s were acidified wi t h sul phuri c acid t o
this pH val ue in or der t o prevent hydrol ysi s of t he salts dissolved. As soon as t he
Table III 1 - 2 Influence of ur ea on pH vari at i on during ferment at i on of fat emulsions
Shake flask number I II III IV
Volume in ml of:
urea sol ut i on
grease emul si on
salt solution
trace el ement
solution
vitamin sol ut i on
distilled wat er
yeast suspensi on
pH after
0 hours
21 hours
69 hours
75 hours
85 hours
116 hours
140 hours
1.5
15.0
15.0
15.0
1.5
92.0
10.0
5
4+
IVL
\Vi

2
2+
3.0
15.0
15.0
15.0
1.5
90.5
10.0
5
4>/4
4 -
3
3 -
3+
3'/4
4.5
15. 0
15.0
15.0
1.5
89.0
10.0
5
A^h
4+
4
4 -
4
4+
6.0
15.0
15.0
15.0
1.5
87.5
10.0
5
5
4i4
AVi
4+
A^h.
5
51
consumpt i on of ammoni a st art s, t he sulphuric acid l i berat ed will make t he sol ut i on
acidic For t hat reason, somet i mes calcium carbonat e is added, which will give a
preci pi t at e of CaS0 4 and CO2 product i on This met hod was not succesful in our
experi ment s because the fat t y acid, produced by hydrol ysi s, was preci pi t at ed by
t he Ca ions and the ferment at i on st opped
The idea of using ammoni um carbonat e or bi carbonat e as an alternative ni t rogen
source was considered but had t o be rejected for t he following reasons
The solubility of ammoni um bi carbonat e is 1 25 Mol/1 and the dissociation
const ant of the equdi bri um H2CO3 = H"*" + HCO3 is 3 5 x 10 ' ^at 25C To reach a
pH of 5, t he concent rat i on of CO2, dissolved as H2CO3, mus t be 36 Mol/1 and this
value IS t housand t i mes its sat urat i on concent r at i on at 25 C In view of t he
solubility of CO2, t he concent rat i on of NH4HCO3 t hat will give a pH value of 6 is
1 15x10 ^ Mol/1 At this very l ow concent rat i on t he buffer capacity is t oo small t o
be of any use This was confirmed by shake flask exper i ment s wi t h various amount s
of NH4HCO3 Tlie ferment at i ons were st art ed at pH 7 After t wo days pH 5 was
reached and after t wo more days t he pH value was bel ow 4.
The next idea was t o use urea as a second ammoni a source and t o see whet her the
Table III 13 Influence of urea on pH variation duri ng ferment at i on of f at t y acids
with t wo different amount s of i nocul um
Shake flask number I II III
Volume in ml of
urea sol ut i on
fatty acids
salt sol ut i on
trace el ement
solution
vitamin sol ut i on
distilled wat er
yeast suspension
pH after
0 hours
15 hours
48 hours
66 hours
90 hours
114 hours
45
40
150
150
1 5
100 0
10 0
5
41/2
45d
31/4
3
2%
60
40
150
150
1 5
98 5
100
5
4'/2
4+
4+
4+
4+
60
40
150
150
1 5
88 5
20 0
5
4y2
A'A
4'/i
4+
4+
52
selected strain would also be able to grow on a mixture of this compound and
ammonium sulphate. For that purpose, shake flasks were prepared with different
amounts of a solution that contained 25 g/1 ofurea.
After sterihzation at 120C for 20 minutes, the pH was adjusted to a value of 5 by
addition of sulphuric acid. The composition of the fermentation liquid in the shake
flasks, and the variation of the pH values, are given in Table III 12.
The cell concentrations in experiments III and IV estimated by plate counting, were
2.5x10* cells/ml and 3.5x10* cells/ml respectively. These experiments were
repeated with fatty acids and two inoculum concentrations. The results are given
in Table III 1-3.
The cell concentrations in experiments II and III, measured by plate counting, were
1.8x10* cells/ml and 3.5x10* cells/ml. As seen in Table 111 1-3, the pH in the
culture with the higher inoculum concentration was still above 4.
In view of these results, the shake flask experiments were carried out with a
fermentation liquid that had an initial pH of 5 and contained 1 g/1 of urea in
addition to the other components.
1.2.4. Oxygen absorption
As stated in the introduction, the absorption of oxygen depends on the various
conditions under which the experiments are performed. In this case there is one
invariable property, namely the viscosity of the fermentation liquid. In the standard
procedure, the speed of the incubator shaker is kept constant at 160 rpm and the
ampHtude of the shaker is 2.5 cm; moreover the clamps of the rotator were most
convenient for a conical flask of 750 ml.
As is shown by Auro et al. (1957), there is no statistically significant difference
between oxygen uptake values at a constant liquidto flaskvolume ratio for flask
volumes varying between 250 to 2000 ml. Hence conical 750 ml flasks were used.
Tlie experiments of Auro et al (1957) show also that a practical upper Hmit is
reached of 1.5 to 2 mMol oxygen absorption per liter and per minute, when the
ratio of liquid volume to flask volume is 1 to 5. For the 750 ml shake flask this
means a Hquid volume of 150 ml.
According to the experiments of Olson and Johnson (1949), the oxygen absorption
of 2 mMol/1/min during flask experiments with aerobically growing cells means that
there is not a limiting supply of oxygen for glucose dissimilation. However, this
might not be true for cells growing on fat as a carbon and energy source.
The closure of the shake flask also plays an important role. The influence of the
cotton plug is shown by Gaden (1962), who finds an absorption of 0.26
mMol/1/min for an open flask and 0.20 mMol/1/min for a flask with a cotton plug.
Last but not least, the baffles placed in the shake flask, or the indentations in the
53
flask bottom, will increase the oxygen absorption markedly. In the experiments of
Gaden (1962) the oxygen absorption rate increased from 0.20 mMol/1/min to 0.50
mMol/1/min when small baffles were applied and to 1.8 mMol/1/min when large
baffles were used. In all experiments, 20% of the flask volume was filled with
liquid. Corman (1957), who used a Fernbach flask with 72 indentations, found for a
cotton plug closure an oxygen absorption rate of 1.6 mMol/1/min. Apphcation of a
threefilter-disc gave 3.5 mMol/1/min. However, it is evident that the influence of
the closure is much less if the oxygen absorption rate is lower. Without baffles,
Corman (1957) found an oxygen absorption rate of 0.2 mMol/1/min but he did not
perform any experiments with different closures at that value of the oxygen
absorption rate.
In conical flasks of 750 ml, we made three indentations, and experiments were done
with 150 ml liquid and closures with cotton plugs and three layer pads. It turned
out that the latter method was not safe, as far as infections were concerned, and
therefore the cotton plug was preferred. The experiments with the cotton plug and
a rotation speed of 160 rpm, resulted in wet plugs and for that reason shake flasks
without indentations were used. In the standard procedure, we used 750 ml conical
flasks filled with 150 ml liquid and closed with a cotton plug.
1.2.5. Standard procedure
In the experiments with fat emulsions, which had been sterilized separately, the
conical flask of 750 ml was filled with 90 to 95 ml water and then sterilized at
120C for 20 minutes.
After cooling the fat emulsion and the solutions with salts, trace elements and
vitamins were added just before inoculation. For fatty acid fermentations, a
mixture of 105 to 110 ml water and 3 ml fatty acids was sterilized in the conical flask
of 750 ml at 120C for 20 minutes. After cooling the solutions with salts, trace
elements and vitamins were added before inoculation. As mentioned in the
introduction the temperature in the incubator shaker was kept at 30C and the
rotation speed was 160 rpm at an amplitude of 2.5 cm.
1.3. Apparatus for 1 liter fermentation
1.3.1. Introduction
In addition to the 1 liter fermentor itself, the equipment for the growth of cells in
liquid culture consists of a preparation vessel, a pH electrode with recorder, an
air flow meter with recorder, stirrers, a temperature regulator, a dosing pump for
ammonia and an oxygen analyzer with recorder. Figure III 1-1 presents a
schematic flow sheet of the equipment.
54
Figure III 1-1 Schematic of 1 liter culturing apparatus
Explanation: 1. Fermentor vessel; 2. Preparation vessel; 3. pH electrode; 4. pH
recorder and regulator; 5. Ammonia dosing pump; 6. Air flow meter; 7. Meter
and recorder of effluent gas; 8. Plastic bearing system; 9. Thermostatically
controled waterbath with water circulating pump; 10. Oxygen analyzer and
recorder.
In the following sections, those parts of the equipment that consisted of glassware
not normally used for fermentation will be described in detail.
The auxiliaries will be mentioned briefly.
1.3.2. The 1 liter fermentor
Following the decision to build up equipment made of glassware, and to use a
fermentor vessel with a jacket from the various models of available glassware, the
Sovirelsystem was chosen because in this system the closing of the vessel, the tubes
etc. looked rather promising as far as sterility was concerned. Though Sovirel could
not guarantee the steriUty, we found that, if the necessary precautions are taken,
infections during fermentation do not occur. Moreover, the boronsilicate
composition of the glass allowed an operating temperature up to 250C.
The sealing rings, as well as the ethoxypropylene clamps, were temperature-
resistant; this means that the entire fermentor can be sterilized.
The fact that no steel was used for any part of the fermentor was furthermore of
major advantage, because the fatty acids can be rather aggressive at a high
55
Figure III 1-2 1 Liter culturing vessel with Totion needle valve in the bottom
sterilization temperature, causing an increase of the Fe content of the fermentafion
liquid. Figure III 12 shows the fermentor vessel in greater detail.
The 1.2 liter vessel is provided with a special Torion needle valve particularly made to
enable the taking of samples during fermentation. This needle valve has an opening of
8 mm. The vessel of 1000 ml working capacity has a cylindrical body with a full
aperture of 100 mm with a flat flange.
The lid on this vessel can be chosen from a variety of types, all, however, connected
with the vessel in the same special way. The flat flange adaptors are fitted to the
vessel by special flange like rings, clamped together with three clips of the same
material, which can be adjusted by a screw. Between lid and vessel a sealing ring of
silicon rubber, packed with teflon, completes the closing system that is not only
tight enough to prevent any bacterial infection but can, moreover, withstand an
internal or external pressure difference of at least 1 atm.
These closing flanges can also be connected to a system of two steel rods used to
support the fermentation unit.
56
Figure III 1-3 Flange-like ring to clamp lid and 1 liter culturing vessel together, shown in
perspective
^^^1^=^
Figure III 1-4 Lid of 1 liter vessel with extra nylon bearings (6) and openings for:
1. Air-inlet and -outlet; 2. pH electrode; 3. Bearing system for stirrer; 4. Am-
monia dosing or thermometer; 5. Culture liquid and phosphoric acid addition
57
In figure III 1-3 this is shown in perspective. The chosen lid has maximally 5
openings; two are vertical and three under an angle of 20. The latter fact is rather a
disadvantage because special inlet tubes must be provided whenever a vertical inlet
was needed, e.g. one for air.
As will be seen in Figure 111 14 the lid has a central opening of 30 mm and on
this opening a special flange was connected, consisting of two hollow caps. One of
these caps is connected with the Hd and the other with the bearing of the shaft of
the stirrer. This bearing system is shown in Figure III 15. It consists of an
adjustable screw, a spring and three bearing rings all fitted in the glass bearing house
of the shaft.
This system can be made tight enough to prevent any bacterial infection but, at
high stirrer speed, the bearing rings become considerably corroded. This problem was
tackled by constructing an extra nylon bearing of 35 mm just above the bearing
system of the 1 1 fermentor, and these two bearings combined with a flexible drive
properly solved the corrosion problem for all further experiments, even at stirring
speeds of 800 to 900 rpm.
MTTW;
J) -^air inlet
Figure III 1-5 Bearing system for stirrer in 1 liter culturing vessel
1. Adjustable screw; 2. Pressure ring; 3. Spring;
bearing house
4. Bearing rings; 5. Glass
Figure III 1-6 Glas tube construction for air-inlet and -outlet
1. Inlet tube of 10 mm outer diameter; 2. Elongation tube with 22 mm outer
diameter; 3. Outlet tube with cotton plug
58
As will be seen in Figure III 1 4, the four openings in the lid are used as air inlet
and outlet, for ammonia dosing, for the pH electrode and for the addition of
culture liquid from the preparation vessel.
The openings have a screw-neck and can be covered with a hollow cap and a sealing
ring that has a desired inner opening. The construction of this teflon seaHng ring is
such that screwing of the cap will tighten the rod or tube that is fitted through the
opening in the sealing ring. This construction makes it possible that, for a given
opening of 15 mm diameter, a connection can be made with a variety of other
diameters.
The small opening meant for the thermometer was used only in the beginning. When
after a number of experiments we found that the difference between the
temperature of the fermentation liquid and the temperature of the water circulating
through the jacket was always less than 0.2C, this opening was used for adding
ammonia automatically.
For adding phosphoric acid, a 70 mm Sovirelelongationtube was mounted in
one opening and this tube was covered with a closed sealing ring in an open cap.
Through this sealing ring (opening 5 in Figure III 14), the phosphoric acid could
be injected with a needle.
A side tube, opening into this elongation tube, makes it possible to add culture
liquid to the fermentor.
For the inlet and outlet of air, in principle the same idea was applied (see Figure III
16). This time the sidetube served as an outlet for the air and the elongation
tube, with a diameter of 22 mm, was closed with a cap that had a sealing ring with
an opening of 10 mm. The inlet tube had the same diameter and was bent to
follow the shape of the fermentor. At the end, a sintered disc was connected with
an opening just under the impeller of the stirrer (see Figure III 14). The impeller
was of a Vortex type, enabling us to keep the foaming under control.
For the pH-electrode, a special glass housing was constructed and the electrode was
placed in the fermentor just before inoculation. The electrode was sterilized by
placing it in a solution containing 70% m/m of ethanol for at least 24 hours. This
method turned out to be quite succesful. The preparation vessel was attached to the
fermentor by a special ground ball-cup joint connection. After addition of the
culture liquid, the air outlet opening (Figure III 1-6) was closed, thus forcing the
air to escape through the preparation vessel (Figure III 1 1). The advantage of
this system is that if excessive foaming suddenly occurs, as it sometimes did at the
end of the fermentation period, the foam has space enough in the preparation vessel
and does not disturb the experiment.
1.3.3. The preparation vessel
The vessel has no jacket but is provided with a Torion needle valve with a vertical
outlet. By means of a special steriHzable teflon tubing this outlet is connected with
59
one part of the spherical baHcup joint. These baH and cup are connected by a
special stainless steel clip so that it is possible to sterilize vessel and fermentor
separately.
The lid of the vessel has two openings. One is used for the stirrer in case in situ
emulsions have to be prepared; the other serves as an inlet for the ingredients of the
fermentation liquid and the inoculum, and as an outlet for the air during the
fermentation process.
The closing of this vessel, which has a working volume of 1000 ml, is the same as
that described for the fermentor. Since both vessels are standard glassware just as
the additional.tubes, closing-rings, caps, etc. spare parts can be obtained very
easily if replacing is necessary.
Because no baffles were placed in the fermentor, the stirring rod and the pH
electrode sometimes had fat deposits but normally the fermentation liquid was
homogeneous and the foaming could be kept under control.
1.3.4. Auxiliaries
The pH measurement, recording and regulation was done with a Sargent Welch pH
recorder and a pH combination electrode. For the addition and distribution of the
ammonia, which was stored in a burette, a simple peristaltic pump was used; it was
put into action when the pH of the medium reached the pH of the set point of the
recorder.
The air flow was measured by a simple flowmeter. The air was sterilized through a
cotton plug filter. The effluent gas flow was registered by an ordinary gas flow
meter.
The stirrer motor was a Janke and Kunkel model, provided with a thyriston
regulator to vary the speed from 40 to 1000 rpm. The Sovirel has a choice of screw
type propellers and, because no baffles could be provided in the fermentation
vessel, an impeller was chosen with six blades of 60 mm diameter. It proved to be
capable of keeping the foaming under control.
The temperature regulation was effected by means of circulating demineralized
water through the jacket of the fermentor, which was kept at a constant
temperature by a combination of a heater and a cooler. The cooler consists of a
cooling coil fed with tap water. The heating unit was connected to a contact
thermometer that has an accuracy of 0.05C. The continuous flow of tap water was
adjusted so that the heating coil was not working constantly. The heating capacity
was 300 watts and the effective flask volume 1.6 1, so that the whole content could
be warmed up 2.7C in 1 minute. The total capacity of the water circuit was
850 ml and the pump capacity 7000 ml/min, which means that the water
could circulate 8 times per minute through the jacket. The temperature difference
between the fermentation liquid and the circulating water was checked several
times. There was never a difference greater than 0.2C. Hence the desired
60
temperature for fermentation was achieved by setting the contact thermometer of
the heating unit.
The oxygen content of the effluent gas was measured with a Servomex oxygen
analyzer type O.A. 137 connected to a 100 mV recorder. The maximum capacity
of this instrument is 18 1/min; from this quantity 17.9 1 is bypassed because the
maximum flow through the analyzer is 100 ml/min.
The internal oxygen percentage meter-indicator has a range from 0 to 25% v/v and
from 0 to 100% v/v oxygen and the temperature can be varied between -IOC and
-l-50C.
1.4. Apparatus for 20 liter fermentation
1.4.1. Introduction
Just as the 1 liter fermentation apparatus, the equipment for the 20 liters
fermentation consists of a number of vessels, pumps, instruments, etc. In Figure III
17 a flow sheet of the continuous fermentation is given. For the yeast
production no special apparatus had to be constructed. The available standard
apparatus could be used and combined with various instruments. Therefore the
description of the equipment used will be rather brief.
1.4.2. The Biolafitte fermentor
This standard type glass fermentor has a total volume of 20 liters. The air inlet goes
through the hollow stirring axis and its flow meter has a range from 3 to 26.5 l/min.
The temperature can be kept constant between 20 C and 50 C by a tap water
circulation system, with to within 0.3 C accuracy. This tap water can be warmed up
by an electric heater, and cooled by some more tap water.
Agitation of the fermentation liquid is effected by a variable stirring motor with an
oil seal and three Vortex type impellers. In addition a mechanical foam separator was
used, when necessary. The rotation speed has a maximum of 800 rpm and for good
agitation the fermentor was provided with adjustable baffles. The fermentor can be
steriHzed in situ, forcing open steam through the overflow pipe and, also, by steam
that passes through the circulation system which is normally used for temperature
regulation.
1.4.3. The pH regulation
For pH regulation we used an Ingold Elektrode No. 465-35. It cannot be sterilized
in situ. This electrode was connected to a Knick instrumentation panel that can
61
0\ Figure III 1-7 Schematic of 20 hter culturing apparatus
Explanation 1. Fermentor vessel, 2 Medium storage vessel, 3. Conical flasks
for stock medium, 4 Verder peristaltic pump, 5. Knick mstrumentation panel,
6 Ammonia dosmg pump, 7. Temperature and atr regulation panel, 8. Servomex
analyzer and recorder, 9. Meter and recorder of effluent gas
8
^
^
^ ^

o A o
^ ^ 7
11
10
register the pH, and regulate maximum and minimum values if necessary. The
special advantage of this instrumentation panel is, moreover, that the addition of
acid or alkali can be interrupted after time intervals varying between 0.6 and 6
minutes or between 0.1 and 1 minute and be started again after another desirable
time interval, which, too, may be varied between the limits mentioned. In this way
an overdose can be prevented.
In practice the impulse was normally interrupted after 3 seconds and the interval
time was 1 minute. The peristaltic tube pump was connected to two burettes and to
the inlet of the fermentor used for inoculation.
1.4.4. The stock solutions
The stock solutions were partly sterOized in conical flasks of 4 liter, completed
with trace elements and vitamins and then transferred to a Terzano fermentor that
was used as a medium storage vessel (Figure III 17). This glass fermentor has a
total volume of 30 liters and can be sterilized by autoclaving only. The stirring velo-'
city can be adjusted to 188 rpm, 286 rpm, 428 rpm and 650 rpm. The average speed
of 428 rpm was sufficient for preparing a homogeneous emulsion. The temperature
was regulated by a contact thermometer connected to a plunger. The medium was
transported from the Terzano fermentor into the Biolafitte fermentor with a
"Verder" peristaltic pump; it can transfer simultaneously the medium into the
fermentor and the fermented substrate into the conical flasks that served as
recipients for the finished product.
1.4.5. The auxiliaries
The air was supplied by a special Atlas compressor, filtered by a series of PALL
filter units and reduced to 2.5 atm by a Degusta reduction unit.
Tlie steam was prepared in a Henschel steam generator that can produce 200 kg/h
of steam at a pressure of 12 atm. During fermentation the working pressure was
1 atm.; a safety valve had been set at 1.3 atm.
The effluent gas was analyzed by the same Servomex analyzer as that described in
Figure III 11 and the total size of this flow was recorded by means of an
ordinary flow meter and recorder.
For sampling a special device was used, that had been developed at Central Institute
for Nutrition and Food Research TNO; its details are given in Figure III 18.
63
\- v
/
v_y
Figure III 1-8 Device for taking sterile samples
1.5. Harvesting of the yeast
1.5.1. Introduction
The procedure for harvesting the yeast cells was the same throughout the
investigation, irrespective of any raw materials used for fermentation. After its
separation from the culture Hquid by centrifuging, the yeast was washed with a
solution containing 8.9 g/1 of sodium chloride. Mostly one washing was sufficient
but sometimes, after centrifuging the culture liquid, some unfermented fat or fatty
acid was found at the surface of the supernatant. In such cases a second washing
was performed. If, after one washing, the taste of the yeast was fatty, the
washing was repeated. The yeast paste, after centrifuging, normally had a dry
matter content of 20-25%. It was dried in a freeze drying apparatus.
Neither the centrifuging nor the drying were done under sterile conditions, and the
sodium chloride solution used for washing was not sterilized. Accordingly, the yeast
thus produced had to be tested for bacterial infections.
1.5.2. Various centrifuges
For the liquid from the Biolafitte fermentor a refrigeratorultracentrifuge type
64
WKFG 50 K, with automatic acceleration was used. The conical rotor of this appa-
ratus can house six steel tubes of 500 ml each. The rotation speed, which can be
varied from 200 to 20,000 rpm, is normally chosen at 400 rpm. This results in a
centrifugal coefficient of Z=2,500x6, according to the nomograph given by the
WKF-company.
For the liquid from the 1 1 fermentor, a small centrifuge was available. The rotor of
this centrifuge has a diameter of 0.42 m and can house six glass tubes of 250 ml
each.
The maximum rotation speed is 3,000 rpm. For the bottom of the tubes, the
centrifugal coefficient is Z=2,100x6. For that reason, centrifuging was done at
maximum rotation speed.
1.5.3. Freeze drying apparatus
The drying of the yeast paste was done by means of a Virtistype automatic freeze
dryer with a capacity of 5 liter. The paste was frozen at 20C for at least two
hours and placed in the freeze dryer overnight.
2. Selection of the yeast strain, composition of the media, etc.
2.1. Selection of the strain
For this selection two properties of the yeast strains had to be tested. Firstly the
Upolytic activity and, secondly, the capacity to use the glycerol and fatty acid
produced as a carbon and energy source. Of the various methods described for
measuring lipolytic activity, the following were tried. Firstly, the method of
Eykmann (1901), according to which the fatlayer is separated from the medium
agar layer and the Ringer solution is used for detection; secondly, the method of
Berry (1933), in which the fat is dispersed into the medium with agar and the fatty
acid produced is demonstrated with copper sulphate and, thirdly, the method of
Alford et al (1964), in which a double-layered plating procedure is applied and
Victoria Blue, which is incorporated into the medium, serves as an indicator for
lipolytic activity.
The following strains present in the collection of Dr. J. de Waart of the Central
Institute for Nutrition and Food Research TNO were tested:
a, b and c : three Candida lipolyfica strains;
d : an Endomycopsis lipolytica strain;
e : the Candida tropicalis strain CBS 94;
f : the Candida lipolytica strain CBS 599.
65
With each of these strains, three to four experiments were performed. The average
results are given in Table III 21.
Table III 2-1 Lipolytic activities of various Candida and Endomycopsis strains
Strains
Methods
Eijkman (1901)
Berry (1933)
Alford (1964)
- (
-H+-I-
++
-^+
+
-i-
++
+
+
-H-t-
++-H
-h++
-H-
+
+
-^+
-l-H-
-H-
Explanation:
a. b and c: three Candida lipolytica strains; d: Endomycopsis lipolytica strain;
e: Candida tropicalis CBS 94 strain; f: Candida lipolytica CBS 599 strain;
no growth; (+) growth questionable; + minute growth; ++ clear growth; +++ good
growth; +-i~i-+ abundant growth
In addition, the growth of the three strains with the highest lipolytic activity (a, d
and f) was examined by means of plate tests on a number of media of the following
compositions:
TYGP
T Y G
TYP
trypton 5
yeast extract 2.5 g,
glucose 1 gj
agar 25 g,
KH2PO4 3
K2HPO4 8
trypton 5
yeast extract 2.5 g,
glucose 1 g,
agar 25
trypton 5 g,
yeast extract 2.5 g,
agar 25
KH2PO4 3 g
K2HPO4 8
TY
GP
trypton 5
yeast extract 2.5 g,
agar
glucose
agar
KH2PO4
K2HPO4
glucose
agar
agar
KH2PO4
K2HPO4
25 g
1
25
3
1
25
25
3
8 g,
In table III 22 the results of these plate tests are summarized
66
Table 111 2- 2 Growth of yeast strains a), d) and f) on various media
Strains
Media
T Y G P
T Y G
T Y P
TY
GP
G
P
a
+
H-
+
+++



d
-H-^-^
-I-++-I-
-i-f++
+-H-
++
+
+
f
-l-l-H-
++++
++++
+
+
-

Explanation:
T = trypton; Y = yeast; G = glucose; P = phosphate; a = Candida lipolytica; d =
Endomycopsis lipolytica; f = Candida lipolytica CBS 599;
no growth; (+) growth questionable; + minute growth; ++ clear growth; +++ good
growth; ++++ abundant growth
In view of these results, the Endomycopsis strain d was chosen for further
investigation. A control determination of this strain, performed at the Yeast Division
of Centraal bureau voor Schimmelcultures at the Laboratory of Microbiology, Delft
University of Technology, showed that it belongs to the species Endomycopsis
lipolytica Wickerham (Kurtham et Herman) von Arx (1972).
After a year of experimental work with the strain mentioned, favourable medium
had been developed and investigation then proceded to discover whether perhaps
other strains were better in terms of Hpolytic or growing capacity, when growing in
this medium. A great number of yeasts had meanwhile been tested for their ability
to grow on hydrocarbons as a sole source of carbon. Though there is no direct
relation between the assimilation of hydrocarbons and the assimilation of fats, the
strains concerned have some properties in common, i.e. the oxidation products of
hydrocarbons are acids and so are the lipolytic products from fats.
It seemed, therefore, likely that some of the yeast strains attacking hydrocarbons
had the ability to dissimilate and assimilate fats, too.
This also seemed possible in view of the fact that, according to Bos and de Bruyn
(1973), all strains of a species either possess the ability t o assimilate hydrocarbons
or fail to do so. A number of the most promising strains were tested as follows.
An emulsion containing 5 g/1 refined lard, 0.3 g/1 emulsifier Renex 650 and 15 g/1
agar was autoclaved at 120C for 20 minutes. After cooling, and before
soHdification, were added the mineral mixture, the solution of trace elements and a
67
Table III 23 Growth on lard(L), fatty acids (FA), glucose (G) and no carbon
source (B) by various strains observed for a period of 28 days
Capacity with
FA B
Strains
Candida cacoi
CBS 2020
++++
Candida dendronema
CBS 6270
++
Candida edax
CBS 5657
++
Candida cloacae
CBS 5612
+++
Candida maltosa
CBS 5611
++
+++
(+)
Candida parapsilosis
CBS 2211
+++
Candida tropicahs
CBS 94
++ ++
-l-t-H
Pichia vini
var melibiosi
CBS 5254
-H++
Saccharomycopsis
hpolytica
CBS 599
+++
-(-+-(
++++
Torulopsis
haemulonii
CBS 5149
+++
Selected strain
+++ ++++
Explanation:
- no growth, (+) growth questionable, + minute growth; ++ clear growth; +++ good
growth, -H+-H- abundant growth
68
vitamin solution according to Van der Walt and Van Kerken (1961). The same was
done with a medium in which 5 g fatty acids replaced the lard and the emulsifier.
The emulsions were used to make slants that were inoculated with a suspension of
the various strains of yeast. The growth was compared with slants containing the
same ingredients except that the lard was replaced by 2.5 g/1 of glucose; and also
with slants that had no carbon source but did contain only 0.3 g/1 of the emulsifier
Renex 650. The slants were inoculated at 30C during four weeks. The growth after
time intervals of 7, 14, 21 and 28 days appear from Table III 23.
From these results it will be seen that there is a certain analogy between
hydrocarbons and fatty acids as far as the use of the compounds as sole carbon
source for growth of the yeast strains tested is concerned.
In order to have more quantitative data about this growth, a second series of experi-
ments was performed, this time based upon the work of Markovetz and Kallio
(1964).
The most promising strains of our research were inoculated on GYA slants for
three days. These slants were shaken with 10 ml pepton salt solution, and from the
ceU suspensions obtained 1 ml was transferred into a 50 ml shake flask containing
10 ml sterile solution of glucose 200 mg, pepton 100 mg and yeast extract 50 mg.
After 24 hours' shaking at 30C, 0.25 ml of the culture liquid obtained was
transferred into a 100 ml shake flask with 25 ml sterile solution, containing the
following ingredients in 100 ml water:
animal fatty acids
NH4H2PO4
(NH4)2S04
KCl
MgS04.7aq
ZnS04 7 aq
MnS04.4aq
CuS04.5aq
Kl
Na2Mo04
2 g
200 mg
50 mg
50 mg
30 mg
15 mg
1.5 mg
O.I mg
0.1 mg
0.1 mg
To the above solution was added 0.2 ml of the separately sterilized vitamin
solution of Van der Walt and Van Kerken (1961). The same was done with a lard
emulsion containing 2 g lard, emulsified with 50 mg of Renex 650, instead of 2 g of
animal fatty acids.
The flasks were shaken for 80 hours at 30C, and the cell concentration was
measured by the colorimetric method. The results are given in Table III 2-4.
From the results listed in Table III 24 it will be seen that there is no reason to
replace the strain selected by another one.
69
Table III 2-4 Growth of various strains after 80 h incubation at 30 C given in cells
per ml culture liquid
Strains
Selected
strain
C B S 599
CBS 5611
CBS94
CBS 5254
CBS 2020
CBS 5149
CBS5612
CBS2211
CBS 5657
Endomycopsis
Hpolytica
Saccharomycopsis
Hpolytica
Candida maltosa
Candida tropicalis
Pichia vini
Candida cacoi
Torulopsis
haemulonii
Candida cloacae
Candida parapsilosis
Candida edax
Fatty acid
substrate
123x10' '
115x 10''
86x10 '
94 X 10''
84x10 ' '
92 X lO''
80 X 10''
23x10 ' '
14 X 10'
_
Lard
substrate
138X10'
134 X lO''
102 X 10''
98x10 ' '
94 X 10''
93 X 10''
76 X 10''
60 X 10''
52x10 '
24 X 10'
2.2. Concentration of the fat products
In any culture the maximum density of cells, i.e. the maximum yield of biomass, is
limited by inherent cellular factors of a so far partly unknown nature, and this
situation cannot be changed by increasing the concentration of the carbon source
above a certain level.
It is, therefore, not surprising that the experiments with alkanes are always
performed at low concentration of the carbon source. Munk, Volvova et al. (1969)
tested the chain length specificity at a concentration of 1% m/m; in other experiments,
carried out by Munk et al. (1969), 1.5% m/m was used. Wagner et al. (1960)
describe a maximum concentration of 13.5 g of dry biomass per liter, which is in
good agreement with estimates by Humprey (1968) and Wang (1969); they take
15 g of dry biomass as a base for calculating fermentation costs.
70
In the original patents, Champagnat (1964) used 10% gas oil, containing 12%
nparaffins, which means an alkane concentrafion of 1.2%; in later patents,
however, a new strain, CBS 6331, was introduced giving as much as 23.6 g of dry
yeast/I (B.P. 1973). In that case an air pressure of 2.5 kg/cm^ is used. Assuming
a final production of 15 g/1 of dry biomass, which is a good estimation at 1 atm.,
the fat concentration chosen in relation to a yield of 60-80% would amount to
20 g/1 which seems quite reasonable. For the shake flask experiments, and in the
1 liter fermentor, this quantity was, therefore, chosen as a standard amount.
2.3. Fermentation temperature
In the shake flask experiments, a temperature of 30 C was chosen because for most
of the yeast strains examined a temperature between 28C and 32C is
recommended. One disadvantage of this temperature is that the efficiency of the
cooling, necessary in bulk production, is very poor. Another disadvantage is that
the melting point of fats is between 42C and 47C. Only the fatty acids obtained
from some vegetable oils have a melting point below 30C, and can be used as such.
All fats and the other fatty acids have to be emulsified before they can serve as
subtrate for fermentation.
In Table I 44 some characteristics of various fatty acids are given.
With the 1 Uter fermentor a number of experiments were performed in order to
establish whether the strain selected was perhaps able to produce biomass at a
higher temperature than 30C. Meyrath and Matt (1973) reported that strains
closely related to Candida tropicalis were able to produce biomass from molasses at
a temperature of 44C, though the growth at 40C was much better. With Candida
guilliermondii growing on hydrocarbons, Gradova et al. (1973) were able to apply
temperatures up to 37C, but the maximum productivity was reached already at
temperatures between 32C and 36C. In both publications it is stressed that the
strains selected are characterized by a high ability to prevent the cultures from
being contaminated by other microorganisms.
Therefore, it was possible to use these strains in production processes under
non-sterile conditions.
Unfortunately, experiments done with our selected strain of Endomycopsis
Hpolytica in the 1 Hter fermentor at higher temperatures (35C and 40C) yielded no
production at all; this means that this strain was neither thermophile nor
thermotolerant.
For that reason the fermentations in the 20 liter fermentor were carried out at a
temperature of 30 C.
71
2.4. Fat emulsification
2.4.1. Introduction
According to Erdtsieck (1967), and other investigators, the growth of yeast on
waterinsoluble substances as a carbon source is only possible on the surface of
these substances. Consequently, fat and fatty acids have to be emulsified in very
small particles to give the yeast an opportunity to grow. If the fat product has
a low melting point, i.e. lower than 30C, it can be dispersed by agitation resulting in
droplets with .a diameter of 515x10"* m; but the fat products with a higher
melting point have to be emulsified. The problem of emulsifying fat products in
particles with a diamter of 2xl0'*m is extensively described by Geyeret al. (1948)
in their studies of the intravenous nutrition of fat emulsions. On the basis of their
method it is possible to prepare fat emulsions resistant to autoclaving.
2.4.2. Theoretical considerations
It may be assumed that oil in a water emulsion has the emulsifier on the surface of
the oil droplets, the lipophilic part of the emulsifier being dissolved in these
droplets. This amount of emulsifier is capable of giving a stable emulsion only when
the intersurface between oil droplets and water is occupied with the emulsifier. For
an emulsion of oil in water used in experiments with yeast, the emulsifying agent
must have a high HLB number and it should have no such unfavourable properties
as foaming action or toxicity. It should furthermore not be attacked by the yeast.
In accordance with these principles, a number of emulsifiers were chosen and with
two of them elaborate experiments were done. These nonionic emulsifiers were Brij
35, a polyoxyethylene (23) layrylether and Renex 650, a polyoxyethylene (30) alkyl
arylether. Although it was not specified in the product information bulletin, the
Renex 650 was a nonyl phenylether.
Both substances have a lipophilic part (L) of about the same length, which was
estimated to be 20xl0' "' m. However, in Renex 650 there are 30 polyoxyethylene
groups and in Brij 35 only 23. Therefore, the molecular volume of the hydrophilic
part (H) is supposed to be 80% and 75% of the total molecular volume of the
respective emulsifiers. On the basis of the molecular weights and the densities, the
following calculations can be made:
Mol. weight
Density
Mol. volume
Volume of one molecule
H/L rafio
Volume of the lipophilic part
Length of lipophilic part
Renex 650
1540
1.15
1285x10- * m'
2100x10-''' m'
420x10-' m'
20x10-' m
Brij 35
1200
1.05
1143x10" ^ m'
1900x10-' m'
475x10"' m'
20x10-' m
72
These figures allow an estimation of the amount of each emulsifier that is necessary
to form droplets of a desired diameter. The results of the estimations are given in
Table III 2-5 for Renex 650, and m Table III 2- 6 for Brij 35;
diameter ot tat droplet and amount ot Renex 650 that can give a stable
I 2 3 4 5 x 10"' m
1000 1000 1000 1000 1000 x l O^ m
19 23 0 7 0 3 0 15 xlO"*
420 420 420 420 420 x 10 ^m^
0 6 25 54 101 157 x l O" m^
45 23 15 12 0 9 X 10^
7.6 3.9 2.5 2.0 I 6 % m/m
diameter ot lat droplet and amount ot Renex 650
1 2 3 4
1000 1000 1000 1000
19 23 07 03
420 420 420 420
0 6 2 5 5 4 10 1
45 23 15 12
7.6 3.9 2.5 2.0
diameter of fat droplets and amount of Bnj 35
Table III 2 5 Relation between
emulsion
Droplet diameter 0 5
Volume I Mol fat 1000
Number of droplets 150
per Mol fat
Volume lipophilic part 420
Volume surface layer 0 15
Mol Renex 650 8 8
per Mol fat
g Renex 650 15.8
per lOOgtat
Table 111 2-6 Relation between
emulsion
Droplet diameter 0 5
Volume 1 Mol fat 1000
Number of droplets 150
per Mol fat
Volume hpophilic part 4 75
Volume surface layer 0 15
Mol Brij 35 7.8
per Mol fat
g Bri) 35 10.4
per 100 g fat
1 2 3
1000 1000 1000
19 23 0 7
4 75 4 75 4 75
0 6 25 54
3.9 2.0 1.3
5.3 2.7 1.8
that can give a stable
4 5 X 10" * m
1000 1000 X 10 "* m'
0 3 0 15 xl O' "
4 75 4 75 xl O" ' *m'
10 1 15 7 xlO-^^ii^
1.0 0 8 xl O^
1.4 1.1 %m/m
73
From these calculations it may be concluded that the amount of emulsifier is
inversely proportional to the diameter of the fat droplets. Furthermore an estimate
can be made of the amount of emulsifier needed to obtain an emulsion of a
desired droplet diameter.
2.4.3. Preparation of the fat emulsion
Geyer et al. (1948) stress the point that, for the stability of an emulsion, the size of
the droplets is- of primary importance because only the droplets with a diameter
greater than SxlO'^m will easily coalesce and thus cause creaming out of the
emulsion. For that reason their method is a two-step emulsification; the first step
being effected in a Wearing blender and the second in a high pressure homogenizer,
provided with a temperature regulation and a recycling system. The specific
conditions of the preparation were of course influenced by the kind of fat, the
emulsifier, the temperature, the time of emulsifying and the presence or absence of
electrolytes.
Geyer et al. (1948) report on a number of emulsions prepared in the presence of 6
to 12 grams of disodium hydrogen phosphate in 1 liter of emulsion. In these
emulsions, the amount of emulsifier used was as high as 25% m/m of the amount of
fat. In our experiments with 4%6% emulsifier the emulsions were broken after
sterilization at 120 for 20 minutes, if the mineral nutrients were added before
autoclaving. Tlierefore, in the final preparation demineralized water was used and
the mineral nutrients were sterilized separately. Avery important principle, worked
out by Geyer et al. (1948), was to make emulsions with a tenfold concentration
and dilute sterilized emulsions just before use, with demineralized water. This
principle was applied in the present work, and it proved successful.
First the various amounts of emulsifier were mixed with a small amount of water
and heated to 70C. At this temperature, the fat product, also heated to 70C, was
added under constant stirring with a Thurrax blender. Next the rest of the water
was added in order to bring the fat emulsion to its desired concentration. The
preliminary emulsification was continued for about 3 minutes, giving an emulsion
that was not oiling out immediately. The emulsion was transferred to a high
pressure homogenizer type Renny, with a capacity of 75 1/h circulation fluid. It
proved to work successfuHy at 175-200 atm for 5 minutes and to finalize at 30
atm. for 3 minutes; this last step in order to get rid of the air dissolved in the emul-
sion during the first 5 min. The temperature, which increased in the first 5 minutes
from 47 C to 52 C, decreased during the second period to 50C. A longer period of
emulsifying, as used by Geyer et al. (1948), yielded no improvements. On the
contrary, the amount of particles of a diameter larger than 2xlO'*m increased when
the time of emulsifying at high pressure extended.
We used the modified Geyer method for preparing a number of emulsions of
inedible animal fatty acids and various amounts of emulsifying agents.
74
The emulsions prepared were tested as regards their particle diameter and their
stability before and after autoclaving. The sterile emulsions were diluted and
subsequently aerated (20 1/1/h) to see whether excessive foaming occurred, and
whether the foaming could be kept under control by means of stirring at a speed of
600 rpm. The emulsifiers gave a foam that could easily be kept under control at the
stirring rate mentioned.
The various results are summarized in Table III 27.
Table III 2-7 Emulsions of 200 g fat product per liter with various amounts of emulsifiers
Emulsions
Renex 650
Brij 35
Diameter droplets
StabiHty,
g/1
g/1
10'* m
immediately
after 24 hours
after autoclaving
after aeration
1
10
2.5
C
C
B
_
2
15
10
1.4
L
C
B
_
3
15
1.3
N
L
B
_
4
15
5
< 1
N
N
L
B
5
20
< 1
N
N
N
N
6
20
5
<
N
L
L
C
Explanation:
C creaming; L large particles formed; B broke; N no changes
On the basis of these results, it was decided to use Renex 650 as the emulsifying
agent and to prepare emulsions of the foUowing composition:
filtered fat product 800 g
Renex 650 80 g
demineralized water 3600 g
The emulsions thus obtained were autoclaved in 1 liter infusion flasks and stored
for some time before they were used for fermentation. A small amount was taken
out, diluted and used for shake flasks experiments.
As raw materials, three fat products were chosen namely grease, fancy taUow and
fat used for deep fat frying. The properties of these products will be described in
Chapter IV.
2.5. Composition of the salt mktures
2.5.1. Introduction
It is known that a number of elements are involved in the metabolism of yeast cells.
This may be illustrated by Table 111 2- 8 (see Wiken 1954).
75
Table III 2- 8 Enzymes and groups of enzymes activated by one or more of the
ions of Mg, Ca, Zn, Mn, Co, Ni, Fe, Cd, V, K, CI, 1 and P
Arginase Carboxylase
Dipeptidase Oxalacetic acid carboxylase
Polypeptidase Enolase
Nucleotidase Fumarase
Phosphoglucomutase Pyruvic acid dehydrase
Phosphoglyceromutase Isocitric acid dehydrase
Hexokinase Aldolase
Phosphatase Amylase
Phospoenolpurivic acid transphosphorylase
Phosphoguanidine transphosphorylase
However, the available information about the necessary quantities of ions
does not display any consensus of opinion. For example, it is known that
Mn ions stimulate arginase activity in a concentration of 3x10-* Mol/1
(Edlbacher and Baur 1938). The same ions play an important role for
isocitric acid dehydrogenase (Adler et al. 1939) but in this case their function can
partly be taken over by Mg ions. Mg ions are essential for the activity of many
enzymes, e.g. enolases, carboxylases, transferases, etc and they can act in very low
concentrations (Morris 1958; Bowen 1966).
Carboxylase is also stimulated by Zn ions in concentrations from 2 to 5x10"* Mol/1
and this activity is intensified in the presence of Mn ions (Leuthardt 1941). The
growth of yeast is promoted when Zn ions are added to the medium (McHargue and
Calfee 1931).
The ions of Fe and Cu induce increased activity in the citric acid cycle (Elvehjem
1931). In spite of the fact that these ions act in very low concentration, they are
added to media for microbes in rather large quantities.
2.5.2. Mineral content of media and bacteria
From the survey given by Wiken (1954, Tables 14 and 15) it is evident that the
media used by various investigators, when growing different microbes, show very
strong variations as regards the amounts of mineral and trace elements added.
Dostalek et al. (1972) have given a survey of the averages of the quantities of
minerals used in media for gram positive and gram negative bacteria, all related to
the amount of nitrogen. For P, S, K, Mg, Mn and Zn, the figures are:
N = 100 K = 201 (65)
P = 176 (66) Mg = 15 (74)
S = 59 (70) Mn =0 . 15 (26)
Zn =0 . 13 (26)
76
The figures between brackets indicate the number of references quoted by Dostalek
etal. (1972).
Tlie British Petroleum Company (quoted below as B.P.) has a great number of
patents in which media are described used for the fermentation of hydrocarbons.
Very likely B.P. had done much research - partly not pubHshed before the media
were chosen, and perhaps the minimum requirements of the various elements are
still under investigation. It is, therefore, important to compare the composition
averages of a great number of media given by Dostalek et al. (1972) with the data
for the media described by B.P. (1963, 1968, 1971). This is done in Table III
2-9, which also presents the number of references on which the figures of
Dostalek are based.
Table III 2-9 Compositions of B.P. media compared with those of Dostalek
(1972) aH given in grams per liter
Dostalek Number of
B.P. 1963 B.P. 1968 B.P. 1971 1972 references
(NH4)2HP04
KCl
MgS04.7aq
ZnS04
MnS04.1aq
H2SO4
H3PO4
2.0
1.15
0.65
0.17
0.045
.
.
2.0
1.15
0.65
0.17
0.045
.
._
- . -
0.916
0.521
0.153
(7aq)
0.035
(4aq)
0.172
1.594
2.0
1.03
0.406
0.0015
(7aq)
0.0016
(4aq)
0.436
.
65
74
26
26
70
Table III 2-10 Relative amounts of minerals present in bacteria, various media
and the chosen medium
Johnson
media
Dostalek
bacteria
B.P. Patents
1968 1971
Chosen
medium
N
P
S
K
M g
M n
Zn
100
22
3.4
15
2.6
0.03
0.02
100
23
8.9
14
4.9
0.05
0.14
(25)
( 7)
(24)
(27)
(20)
( 6)
23
6.1
29.5
3.3
0.7
3.4
23
6.7
21.9
2.3
0.4
1.6
23
8.7
11.2
1.3
0.16
1.45
77
From the figures in Table III 2-9 it is evident that the composition of the B.P.
media is based on a concept that, in certain respects, is quite different from that
underlying the composition of the other media. Dostalek (1972) had also
published the average amounts of elements present in gram positive and gram
negative bacteria grown on a number of media based on bacteria containing 10%
nitrogen, based on the dry weight. Johnson (1972) has summarized the approximate
quantities of mineral elements required for microbial growth. In Table III 210
these figures are compared with those mentioned in the B.P.-patents. The number
of references are given in brackets.
Johnson (1972) explains that these figures, actuaHy given as amounts required for
the growth of 1 gram of dry cellmaterial, wiH vary with the organism used.
Dostalek (1972), however, claims that there is a more or less constant relation
between the amounts of the various elements necessary for the growth of bacteria.
2.5.3. Composition of the salt media
On the basis of the above mentioned concept of Dostalek et al. (1972), we chosen
the mineral composition of the media prepared for fat fermentation.
There were two reasons for making two separate solutions of the mineral macro
and micro-nutrients, viz. one solution with rather high amounts of salts and
another one with diammonium hydrogen phosphate and low amounts of salts. The
first reason was that the solubility product of MgNH4P04 is only 2.5x10-' ' , which
means that in a solution of this salt the magnesium ion concentration is bound to
be very low. The diammonium hydrogen phosphate is, therefore, dissolved in the
micro-nutrient solution which for that reason is called solution with trace elements.
The second reason was that, for the 20 liter fermentations, a solution of a hundred-
fold concentration of the trace elements, without phosphate, could be made saving
much time needed for preparation of the media. In practice the following
components were dissolved in 1 1 of water and acidified with dilute sulphuric acid
until a pH value of 5 (solution I).
Solution I containing per liter:
Urea 10.0 g
(NH4)2S04 5.0 g
KCl 5.0 g
MgS04,7aq 3.0 g
ZnS04.7aq 1.5 g
MnS04.4aq 0.15g
This solution was used for shake flask experiments. The same solution without
urea was applied in the 1 Hter fermentation experiments. In both cases, the
solutions were added in an amount of 10% v/v of the final media.
The solutions were sterilized by filtration through a G5 glass filter.
78
It should be noted that the actual relative concentrations in the 20 liter
fermentation differed from those mentioned because for these fermentations tap
water was used. According to Vewin (1974) the mineral composition of tap water
was as follows:
mg/1 min.
mg/1 max.
ion
N
1.6
3.4
NOj-
P
0.007
0.016
PO4'"
S
7.7
10.0
S04^'
K
1.1
1.3
K+
Mg
7.5
9.6
Mg^ +
Mn
0.01
0.04
Mn' +
Because no figures of the Zn ion concentrations were given, the tap water was
analyzed according to II 2.9.4, yielding 0.1 ppm Zn.
Solution II contained, in addition to NH4H2PO4, small amounts of the foUowing
compounds: CUSO4, NaMo04 and KI.
The pH of the solution was adjusted to 5 by means of addition of sulphuric acid and
the solution was sterilized by passing through a G5 glass filter. This solution II
contained the following amounts of salts in 1 Hter of double-distilled water:
Na2Mo04: 10 mg; CuS04.5aq: 10 mg; KI 10 mg; NH4H2PO4: 20.0 g The choice
and the quantities of the trace elements were based on the foUowing considerations.
Kurdiva et al. (1972) found, for the growth of Candida tropicalis, an optimum level
of 1 mg Mo/1 in their medium. Dostalek (1972) gives a relative Mo concentration of
0.002 in bacteria and 0.09 in media (17 references, see Table III 2-10). In the
Johnsonmedium (1972), the relative amount of Mo is 0.014 and in the MiUer
medium (1964) it is 0.008. Wickerham (1948) dissolved 0.01 mg Mo/1. The
concentration of Mo in solution II is 0.5 mg/1, which corresponds to a relative
amount of 0.01 (see Table III 2-10).
The relative copper concentration in solution II is also 0.01. Johnson (1972) has
suggested 0.14, whereas a value of 0.03 is given as an average of 21 references by
Dostalek (1972, see Table III 2-10).
The fat products used for fermentation contained a rather high amount of iron. For
soya fatty acids it was 200 ppm and for animal fats 300 ppm, which means that in
the fermentation liquid iron was present in a relative amount of 0.2 to 0.3, values
which are of the same order of magnitude as those mentioned by Dostalek (1972)
and Johnson (1972) viz. 0.3 and 0.4 respectively.
Greaves et al. (1928) found that the growth of some commercial yeast strains
increased upon addition of 1 ppm of iodide or iodine, whereas Olson and Johnson
(1949) did not find any stimulation of growth of Candida utilis in the presence of
iodine in the substrate. Since Wiken and Richard (1951), Wiken et al. (1954) and
MiUer (1964) are using media with iodine for the cultivation of yeast we dissolved
10 mg/1 of KI in solution II. This amount corresponds to 1 mg/1 in the final
medium i.e. the same amount as used by Wiken and Richard (1951), Wiken et al.
(1954) and MiUer (1964). Wickerham (1946) used 0.1 mg/1 KI in his medium.
79
The amount of phosphate chosen is based upon the B.P. patents, where 470 mg/1
(1968) and 504 mg/1 (1971) are used. When 2 g NH4H2PO4 is dissolved in the 1
Hter solution, the amount of P is 540 mg/1.
For the fermentations in the 20 Uter fermentor, the phosphate was added directly
to the medium that was sterilized in situ, and only the three micro-nutrients
mentioned above were added afterwards. Accordingly a solution was prepared that
contained 100 mg of each of the three salts dissolved in 1 liter distilled water
acidified with sulphuric acid to a pH of 5. It was sterilized by filtration through a
G5 glass filter.
2.5.4. Uptake of some elements by the growing yeasts
The amounts of the ions of the elements Mg, Mn and Zn added to the culture
media, were subjected to ample variations, and mostly tested in shake flask experi-
ments.
The final concentrations were checked in batch fermentations. After harvesting
the culture, samples were taken from the yeast paste and the supernatant. From the
analytical results obtained it could be concluded that from the original amount of
Mg, Mn and Zn present 65% - 85% had been taken up by the yeast. From these
results it seems probable that none of the elements mentioned limited growth. This
conclusion was confirmed by the analyses performed during the continuous
Table III 2-11 Results of the analysis of Mg, Mn and Zn ions present in yeast paste,
culture medium supernatant and stock medium during continuous
fermentation
2nd
day
3rd
day
Supernatant
Yeast paste
Total
mg/kg
Supernatant
Yeast paste
Total
mg/kg
g
848.5
41.5
890.0
655.0
36.5
691.5
Stock medium
mg/kg
Mg
ppm
26
182
27
198
mg
22.1
7.5
29.6
33.2
17.7
7.2
24.9
36.0
33.5
Mn
ppm
5.1
30.8
4.9
32.4
mg
4.3
1.3
5.6
6.3
3.2
1.2
4.4
6.3
6.1
Zn
ppm
34
214
34
235
mg
28.8
8.9
37.7
42.4
22.3
8.6
30.9
44.7
45.0
80
fermentation of the soya fatty acids. After two and three days, samples were taken
from the yeast, the stock medium and the supernatant of the culture medium and
analyzed. The results are given in Table III 211.
The figures of Table III 2-11 confirm that the three metals did not limit the growth.
The ratio of the concentrations present in the yeast and the concentrations in the
supernatant was calculated as 88% for Mg, 86% for Mn and 87% for Zn, which is
higher than the values obtained for the batch experiments. These figures also
corroborate the hypothesis that the amounts of these elements should be chosen in
a more or less fixed interrelationship.
2.6. Suppletion of vitamins
2.6.1. Introduction
The role of vitamins in the growth of yeast is extensively described by Suomalainen
and Oura (1971). Some vitamins like riboflavin and folic acid are synthesized by aU
yeasts, others are not. Odintsova (1973) has shown that, under anaerobic
conditions, a great many unpigmented yeasts and yeastlike organisms can
synthesize mesoinositol, biotin, pantothenic acid, thiamin, pyridoxine, niacin and
para-aminobenzoic acid. She also found that mesoinositol, thiamin and niacin are
only very sparingly excreted by yeast ceUs into the medium, whereas the other
vitamins are excreted to 4060% depending on the stage of multipHcation.
As is well known, one or more vitamins, either in pure form or as components of
yeast extract, are often included in media used for growing of yeasts. The way
in which some vitamins are involved as agents in yeast metabolism, or as
constituents of the biomass, will therefore be briefly illustrated below.
2.6.2. Role of the vitamins as agents in yeast metabolism
Biotin participates in yeast anabolism in a number of ways: e.g. in the
carboxylation of pyruvic acid, in the synthesis of pyridine nucleotides, nucleic
acids, purine and pyrimidine bases, and in the synthesis of polysaccharides and
fatty acids. Biotin deficiency leads to damage of the plasma membrane.
Pantothenic acid is involved in the transfer of the acylgroups and in the
carbohydrate and fatty acid metabolism. It is a component of coenzyme A.
Inositol may affect the growth of some yeasts, and lack of inositol leads to less
efficient cell division. Attached to Hpids it acts as a structural component.
Thiamin diphosphate is the coenzyme for the oxidative decarboxylation of
alpha-keto acids and as such it is often referred to as cocarboxylase.
Niacin or nicotinic acid promotes the production of nicotinamide adenine
81
CX)
Table III 2-12 Vitamin contents of vanous yeasts in ppm
A B C D E F G
Thiamin HCl Bl 3.2 3 0 7.0 5.3 125 4.0 3.1
Riboflavin B 2
Capantothenaat B 5
Pyridoxine B 6
Biotin H
Niacin
pAminobenzoic acid
Folic acid
Inositol 5000 3265 1430
Explanation
A Difco yeast extract, information from the manufacturer
B Bacto yeast extract, information from the manufacturer
C Pekilo yeast. Process Biochemistry 1973, page 19
D Torula yeast. Single Cell Protein, MIT press 1968, page 92
E Brewers' yeast" The Yeasts III, Academic Press 1970, page 434
F and G B P yeast grown on nparaffins and gas od respectively,
Outlook on Agriculture 1970, page 102
19
20
1.4
279
20
20
1 0
280
55
62
25
1 8
440
15
45
37.2
33.4
2.3
417.3
17
21.5
35
120
50
1 0
500
49
180
125
25
430
40
6.4
183
10
57
125
35
70
dinucleotide (NAD) in yeasts under aerobic conditions and its amide is part of the
hydrogen transferring coenzymes NAD and NADP.
Of the constituents of the vitamin B 6 group pyridoxal phosphate acts as a
coenzyme of those enzymes that participate in amino acid metabolism such as
amonitranferases and amino acid decarboxylases
Riboflavin is a constituent of several enzyme systems that are involved in
intermediary metabolism Its phosphate, the flavin mononucleotide (FMN), is a
constituent of the Warburg yeUow enzyme, cytochrome reductase and the amino
acid dehydrogenases Combined with phosphate, adenine and ribose it gives flavin
adenine dinucleotide (FAD), the prosthetic group of some oxydases and dehydroge-
nases The participation of the folic coenzymes in reactions leading to synthesis of
the purines and thymine required for DNA synthesis emphasizes the fundamental
role of fohc acid in the growth and the reproduction of ceUs
There are nine isomeric inositols but only one, the socalled myo-inositol,
mesomositol or iinositol, is biologically effective Especially the lipid content of
yeasts is strongly influenced by inositol
pAmmobenzoic acid forms a part of the folic acid molecule and is a growth
factor for some Rhodotorula yeasts
2 6 3 Vitamin concentration
Table III 2-12 gives the vitamin contents in a number of yeasts and in two yeast
extracts
From the figures listed in Table III 2-12 it is evident that no constant relations exist
between the various amounts of vitamins For that reason, the final composition of
Table III 213 Vitamin concentrations in various final media in ppm
Thiamin HCl B 1
Riboflavin B 2
Capatothenaat B 5
Pyridoxine B 6
Biotin H
Niacin
p- Aminobenzoic acid
Folic acid
Inositol
Wickerham
1946
400
200
400
400
2
400
200
2
2000
Van der Walt
Van Kerken
1961
400
200
400
400
2
400
200
2
2000
This
thesis
1974
1000
200
400
400
2
400
200
2
2000
83
the vitamin solution used for fat fermentation was based on the publication of
Wickerham (1946), and on those of Van der Walt and Van Kerken (1961), except
for the amount of thiamin
According to the B P.patents, originally 25 mg/1 of yeast extract was added as a
source of vitamins (1963, 1968) but a later patent (1971) mentions only the supply
of thiamin hydrochloride Of this 220 mg is added to 1 liter of the medium
If the yeast extract contained 3 ppm thiamin this means that instead of
75x10'*mg/1 the amount is 220 mg/1
Whatever might be true, this was a reason for us to raise the thiamin amount to 1
mg/1 The vitaipm concentrations of the various media are given m Table III 2 13
In fact a solution was thus prepared with demineralized water containing a
hundredfold concentration, and this solution was stenlized by filtration through a
G5 glass filter
84
CHAPTER IV
APPLICATION OF STANDARD PROCEDURE
1. Batch fermentations of grease emulsions
1.1. Introduction
For these fermentations, a sample of an inedible grease was chosen. In analyzing
this sample according to the methods described in Chapter II, the foUowing
chemical and physical characteristics were obtained:
MeUing point
Moisture
Impurities
Unsaponifiable matter
Saponification value (SV)
Acid value (AV)
Fat content
Fatty acid content
Molecular weight of
fatty acids
Acid profile: <
Cl 6
Cl 6
Ci 6 : i
C, 7
Cl 8
Ci 8 : i
Ci 8: 2
Ci 8: 3
47.0C
0.1%
0.2%
0.2%
190.6
1.1
99.3%
0.5%
272.8
1.5%
25.4%
3.1%
0.5%
15.1%
46.4%
7.8%
0.3%
m/m
m/m
m/m
mg KOH/g
mg KOH/g
m/m
m/m
m/m
m/m
m/m
m/m
m/m
m/m
m/m
m/m
The calculated molecular weight based on the fatty acid profile is 273.8, which is in
good agreement with the value of 272.8 calculated from the SV and AV figures. When
the molecular weight of these fatty acids is taken as 273.6, the fat weight must be
859. The fatty acid formula can approximately be formulated asCn 3H34O2 and
the corresponding formula for the fat wiU become: C55Hio406.
In line with the procedure described in Chapter II 2.4, this fat was used to make
five portions of a sterile emulsion, starting with 200 g of fat, 20 g of Renex 650 and
900 ml of water.
After autoclaving and cooling, a sample was taken and analyzed. The emulsion
contained 18.1% m/v fat and 1.9% m/v of the emulsifier Renex 650.
This sterile grease emulsion was next used for a number of fermentations.
85
1 2 Preparation of the inoculum
For the fermentations in the 20 1 fermentor, an inoculum of about 1 liter had to be
prepared This occurred in three steps preparation of the slant, shake flask fermenta-
tion and fermentation m the Sovirel apparatus A slant containing GYA was moculated
and incubated for 71 hours at 32"C By that time the yeast cells were, according to
our experience, m their logarithmic phase and could be used for moculation of the
shake flask
For this purpose a 750 ml conical flask, containing 92 ml distilled water, was
sterdized After cooling, we added 16 5 ml of the sterile grease emulsion To the
emulsion were added 15 ml of the salt mixture with urea, 15 ml of the solution of
the trace elements with phosphate and 1 5 ml of the vitamin solution The final
emulsion of 140 ml in the shake flask contained the foUowing ingredients
grease 2 98 g
urea 150 mg
(NH4)2S04 75 mg
(NH4)H2P04 300 mg
MgS04 7 aq 45 mg
ZnS04 7 aq 22 5 mg
K CI 75 mg
MnS04 4 aq 2 25 mg
CuS04 5aq 150 fig
Na2Mo04 150 fxg
KI 150 fig
Renex 650 315 mg
In addition the vitamins
The yeast cells were washed from the slant into the shake flask by means of 10 ml
salt solution, and the shake flask was placed in the rotating machine for 65 hours
The Sovirel fermentor was meanwhile prepared A conical flask of 1 liter was filled
with 610 distilled water and stenhzed at 120C for 20 min After cooling to 30C,
90 ml of the grease emulsion, 90 ml of the solution containing salts without urea
and 90 ml of the solution of trace elements and phosphate were added together
with 9 ml of the vitamin solution
The mixture thus obtained was transferred into the sterile Sovirel apparatus, and
the emulsion stirred and warmed up to 30C by means of the external water
circulation system Subsequently, 100 ml of the inoculum in the shake flask was
transferred into the fermentor Next, the pH-electrode was positioned, the
ammonia distribution system connected and the pH registration started with the
setpoint at 4 3 The air flow was set at 20 1/h and the stirrer at 600 rpm Ammonia
consumption started within an hour and fermentation continued until this
consumption became very low There were no problems of excessive foaming
during this penod which lasted 32 hours At the beginning and at the end of the
fermentation, samples were taken to check the sterdity The total amount
transferred to the 20 1 fermentor was 950 grams
86
1.3. Preparation of the Biolafitte fermentor
After the Biolafitte fermentor had been steriHzed for at least 30 min at 120C and
cooled down, 13.2 liters of a solution was added; it contained the foUowing
amounts of ingredients, dissolved in tap water:
(NH4)2S04 7.5 g
NH4H2PO4 30.0 g
MgS04 7aq 4.5 g
ZnS04 7aq 2.25 g
KCl 7.5 g
MnS04 4aq 225 mg
This solution was sterilized in situ at 120C for 20 min. After the solution had
cooled down, the stirrer was set at 200 rpm and the thermostat at 30C.
In a sterile conical flask of 4 1, which had an opening at the bottom, 1490 ml of the
grease emulsion was mixed with 150 ml of the solution of the trace elements
without phosphate and 150 ml of the vitamin solution.
After the fermentation in the Sovirel apparatus was completed, the content was
transferred into the conical flask mentioned, except for a small sample which was
used for a plate count. The sample had a ceU concentration of 1.5 x lO' cells/ml.
1.4. First fermentation of the grease emulsion
After inoculation the ammonia distribution system was connected and the pH
regulation started with the setpoint at 4.3. The two burettes contained 6.1 n
ammonia solution. On the basis of results of the Sovirel experiments, the air flow
was adjusted to 200 1/h and the stirrer speed to 600 rpm. Within 1 hour the
ammonia consumption started and, in a sample taken, the concentration of the cells
was 8 X 10'' ceUs/ml. After 4 hours the fermentation liquid showed excessive
foaming and neither lowering of the stirrer speed nor addition of 10 ml rape seed
oil as an antifoaming agent could reduce the tendency to overfoaming. Even when
the air was reduced to 165 1/h, excessive foaming continued. Only after reduction
of the air flow to 105 1/h, the foam formation could be kept under control for
several hours. The stirrer speed was then kept at 600 rpm and I ml of the rape seed
oU was injected regularly at intervals of about one hour.
After 28 hours the air flow had to be reduced to 85 I/h and after 30 hours the
ammonia consumption was only 1.7 ml/h. Fermentation was then stopped.
The fermentation Hquid was transferred into 4 1 conical flasks and centrifuged. In
the supernatant a small amount of fat, viz. about 10 grams, was found. The yeast
paste was washed twice and dried as described in Chapter III 1.5 giving a yield of
200.7 g dried yeast.
87
1.5. Discussion of results of first fermentation of grease emulsion
The yeast produced had a good colour and smeU. Its weight was 200.7 g and the
total amount of fat used was 287.5, including the fat in the supernatant. Hence the
yield amounted to 69.8%.
The nitrogen content of the yeast was 6.3% m/m, from which the crude protein was
calculated as 39.4% m/m.
The total amount of ammonia used was 185 ml 6.1 n solution; this means a
nitrogen consumption of 15.8 g N. Since in 200.7 g yeast 12.6 g N was present, the
nitrogen uptake had an efficiency of (12.6 : 15.8) x 100% = 80%.
It is evident that the available amount of air in the fermentation Hquid had been
much lower than the generally assumed amount of 1 v/v/min because that would
have corresponded to an air flow of 9001/h for 151 fermentation liquid.
According to Erdtsieck (1967, p. 109), an air flow of 0.34 v/v/min gives the same
production as an air flow of 0.65 v/v/min. The corresponding partial pressures of
oxygen are 36 mm and 70 mm Hg (30C).
Damaschke (1958) has found that, in a suspension of 1% of yeast, a lower oxygen
concentration than 0.3 mg/1, which is about 4% of the saturation concentration of
8.2 mg/1 (20C and 76 cm Hg), Hmits respiration. Winzler (1941) found that the
respiration of Bakers' yeast was limited by oxygen at concentrations below that
corresponding to a partial pressure of 6 mm Hg (20C) in the gas phase of the
cultures. Because the partial pressure of oxygen in air at 20C is about 160 mm Hg,
this value corresponds to 3.8% of the saturation concentration.
When the figures given by Erdtsieck (1967, p. 109) are extrapolated to 6 mm Hg
partial pressure, the corresponding air flow wiU become 0.057 v/v/min or 3.4 v/v/h.
For a volume of 15 1 fermentation liquid at least 51 1/h of air are needed to prevent
limitation of respiration.
Mateles (1971) has presented a formula to estimate the quantity of oxygen required
for SCP production. If ammonia is used as the nitrogen source, and the only
products of metabolism are the ceUs themselves, carbondioxide and water, the
following equation can be used:
goxygen ^ 32C + 8 H- 1 6 0 ^ ^ ^^ ^^ ^ ^ ^ ^ ^ ^ ^ , _^^^57 c* -0. 08H*
g ceUs Y x M
where: C, H and 0 represent the number of atoms of carbon, hydrogen and oxygen
present in 1 molecule of the carbon source; C*, H*, 0* and N* represent the
percentage of carbon, hydrogen, oxygen and nitrogen of the yeast cells; Y
represents the yield i.e. grams of cells per gram carbon source used and M represents
the molecular weight of the carbon source.
In item IV 1.1. it was stated that the carbon source was a mixture of 99.3% m/m of
fat and 0.5% m/m of fatty acids. The formula of the carbon source can therefore be
represented by C54 sHioaO; 9, and the molecular weight will become 851. The
88
yield has been 0.7 and, when for yeast the composition is chosen as 52.1% C, 8,5%
H, 28.7% 0 and 6.0% N (see V 2), substitution wUl give:
g o x p n ^ 32x54.5 + 8x103-16x5. 9 ^ ^^^^^8.7 + 0.01714x6 - 0.0267x52.1
g cells 0.7x851
-0 . 0 8x8. 5=4. 152-1. 681 =2.471
This results in the equation: g oxygen = 2.47 x g ceUs.
The total oxygen uptake during 30 hours of fermentation had been 176 1 of
O2 at 30C and 80 cm Hg, corresponding to approximately 240 g O2, which means
that the amount used has been about half the value calculated according to the
formula of Mateles, since for a production of 200.7 g of ceUs would have been
needed 200.7 x 2.47 g O2 = 496 g O2. It can be expected, therefore, that a greater
air flow would have given an increase of yeast production. The construction of the
fermentor, however, did not permit a much higher air flow. Hence we decided to
repeat this fermentation, starting with an air flow of at least 80 1/h.
It seemed desirable to apply an antifoaming agent, i.e. cetyl alcohol, in order to get
the foaming under control.
1.6. Second fermentation of grease emulsion
The fermentation in the Sovirel apparatus and the preparation of the 201 fermentor
were the same as described in IV 1.2 and IV 1.3.
This time, 983 g from the Sovirel apparatus was used as inoculum and the yeast
concentration in this suspension was 4x10 * ceUs/ml. The volume of the medium
prepared was 15 liters. After inoculation, air flow was set at 901/h and stirrer speed
at 500 rpm. To keep foaming under control 1 ml of antifoam was added every
hour. After 16 hours the consumption of ammonia was rather high; air flow was
then raised to 135 1/h. By that time the yeast concentration was 6x 10* cells/ml.
After 29 hours, foaming suddenly intensified and some overfoaming occurred.
The air flow therefore was reduced to 105 1/h and some extra antifoam was added.
After 30 hours fermentation was stopped and 14.5 1 was withdrawn from the
fermentor. Centrifuging, washing and drying yielded 149.5 g of dried yeast.
1.7. Discussion of results of second fermentation of grease emulsion
The yeast produced had a neutral taste and smell, and a light yellowish colour. The
calculated yield referred to fat consumed was (149.5 : 14.5 x 18) x 100%= 57.3%.
The N content was 6.5% m/m; this means a crude protein content of 40.6% m/m.
The total ammonia consumption had been 147 ml 6.2 n solution which corresponds
89
to a nitrogen consumption of 12.7 g. The yeast contained (6.5 x 149.5 : 100) g N =
9.7 g N. So the efficiency of the N-uptake had been (9.7 : 12.7 x 100% = 76.4%.
The total volume of oxygen uptake amounted to 204 1 at 30C and 80 cm Hg; this
means approximately 280 grams of oxygen. Table IV 1-1 summarizes the results of
the two fermentations of grease emulsion:
Table IV 1 1 Results of two grease emulsion fermentations
Time Volume Average Production
h 1 air flow 1/h g total g/h g/1
1st
2nd
30
30
16.0
14.5
121
102
200.7
149.5
6.7
5.0
12.3
10.4
The results show that a lower air flow gave a lower production, not only per hour
but also per Hter. As excessive foaming of the fermentation liquid did not allow a
higher rate of air flow, it was decided to change the volume of the fermentation
Hquid.
Moreover, the stirring rod was now provided with a mechanical foam seperator and
an extra impeUor. It was hoped that this would contribute to a better oxygen
transfer, which in turn would result in an improvement of the yield. In order to
examine whether the foaming was partially caused by the emulsifier, subsequent
fermentations were performed with fatty acids.
2. Batch fermentations of fatty acids
2.1. Introduction
For these fermentations the raw material was a sample from a refinery miU. The
fatty acids were byproducts of the soya refining process. Their properties were
estimated according to the methods described in Chapter II.
Melting point
Moisture
Impurities
Unsaponifiable matter
Fat content
Saponification value (SV)
Acid value (AV)
Fatty acid content
Molecular weight of
fatty acids
<
<
29.5C
0.1%
0.1%
1 %
33.2%
199.1
134.3
65.8%
274.9
m/m
m/m
m/m
m/m
mg KOH/g
mg KOH/g
m/m
90
Acid profile <
C. 6
Cl 6
Ci 6 : i
Cl 7
C, 8
Ci 8 : i
Ci 8: 2
Ci 8: 3
0.1%
10.3%
0.1%
0.1%
4.1%
23.7%
55.9%
5.6%
m/m
m/m
m/m
m/m
m/m
m/m
m/m
m/m
From this profile it will be seen that the ratio polyunsaturated to saturated fatty
acids is 4.2 to 1. The average molecular weigjit based on the fatty acid profile is
277.9, which is in good agreement with the value of 274.9 calculated from the SV
and AV figures.
Based on these figures, the formula for the fatty acids can be given by
Ci7 7H33O2. The corresponding formula for the fat wiU become CsgHiosOe
with a molecular weight of 869.
The "gross formula" of the mixture is C23H42O2.S with a calculated molecular
weight of 358. This mixture of fats and fatty acids wiU below be called soya fatty
acids, although it is clear from the data just given that for one third it consists of
fats and for two thirds of fatty acids.
2.2. Preparation of the'inoculum
As stated in IV 1.2, the preparation of the inoculum involved three steps. The GYA
slants were made as described in IV 1.2, and this time the incubation temperature
was 32C and the time of incubation 70 hours. The shake flask of 750 ml total
volume, containing 104.5 ml of water and 3 g of the soya fatty acids, was
autoclaved at 120C for 20 min. After cooling 15 ml of the salt solution con-
taining urea, 15 ml of the solution of the trace elements containing phosphate and
1.5 ml of the vitamin solution were added. Apart from the fatty acids the
solution contained the same ingredients as described in IV 1.2, except for the
emulsifler. The yeast from the slant was next washed into the shake flask, using 10
ml of salt solution, and the flask was placed in the shaking machine.
The Sovirel fermentor was sterilized, as was a separate flask of II total volume
containing700mlof water and 18 grams of fatty acids. After cooling to 30C 90 ml
of the salt solution 90 ml of the solution containing the trace elements and 9 ml
of the vitamin solution were added. Next, the contents were transferred into
the Sovirel apparatus. This mixture was kept at a temperature of 30C until, after
shaking for 64.5 hours, 100 ml of the shake flask culture Hquid could be used for
inoculation. The medium in the Sovirel apparatus contained, before inoculation,
the following ingredients, in addition to the vitamins.
Fatty acids 18 g
(NH4)2S04 450 mg
91
NH4H2P04
MgS04 7 aq
ZnS04 7 aq
KCl
MnSo4 4 aq
CUSO4 5 aq
Na2 M0O4
KI
1 8
270
135
450
13 5
0 9
0 9
0 9
g
mg
mg
mg
mg
mg
mg
mg
After addition of 100 ml of inoculum, fermentation was started with an air flow of
20 1/h and a stirring speed of 67? rpm Ammonia consumption began after 45
rmnutes at pH 4 3 and contmued for 26 hours, then the uptake of 1 n ammonia
solution was only 0 5 ml/h So fermentation had nearly finished and the
fermentation liquid could be used as moculum for the Biolafitte fermentor
The analysis of a sample taken at that time showed that there were no viable
bacteria present, and that the yeast concentration was 2x10 * cells/ml
2.3 Preparation of the Biolafitte fermentor
The first fermentation was carried out with a total volume of 16 1 of fermentation
hquid The fermentor was prepared as described m Chapter IV 1 3 The water
content on this occasion was 14 7 1 and 300 grams of fatty acids was also present
in the mixture This was steriHzed in situ
In a 2 1 sterile conical flask with a bottom opening, 150 ml of the solution
containing trace elements and 150 ml of the vitamin solution were mixed with 870
grams of the completely fermented Sovirel culture liquid This suspension was then
used as an moculum for the Biolafitte fermentor
2 4 First batch fermentation of fatty acids
After inoculation, fermentation was started with an air flow of 801/h and a stirrer
speed of 800 rpm The ammonia suppletion was connected to keep the pH at 4 3
The burettes were filled with 6 1 n ammonia solution
When, after 7 hours, the consumption of the ammonia solution had increased to 7
ml/h, the air flow was doubled Three hours later the ammonia consumption was 15
ml/h and now the air flow was set at 280 1/h After 18 h fermentation, suddenly
overfoaming occurred, the air flow had to be reduced to 103 1/h and some rape seed
od was added Three hours later, ammonia consumption stopped and the pH
became 4 85 The fermentation was then termmated The fermentation liquid was
transferred into 4 1 conical flasks and centrifuged After washing, 466 grams of
yeast paste were obtamed
92
Because of excessive foaming at the baffles and the bottom of the lid of the
fermentor, rather large quantities of yeast paste and fatty acid had precipitated.
After washing and centrifuging an additional amount of 399.4 grams of yeast paste
was collected. Both portions of yeast paste were dried, giving respectively 110.1 g
and 90.1 g of yeast dry substance.
2.5. Discussion of results of first fatty acids fermentation
The yeast produced had a good smeU, a light yellow colour and a neutral taste.
Tlie total amount of fatty acid consumed was 320 g and the yeast produced
weighed 200.2 g. Hence the yield was: (200.2 : 320) x 100% = 63%. This rather low
yield can be accounted for by considering the fact that part of the yeast and fatty
acids had precipitated on the baffles and on the bottom of the lid of the fermentor
and, therefore, could not participate in the fermentation.
The nitrogen consumption had been 250.2 ml of a 6.1 n ammonia solution, which
corresponds to 21.4 g N. Since the yeast had a nitrogen content of 6.8% N, the
efficiency of the nitrogen uptake had been: (13.6 : 21.4) x 100%) = 63%.
From this experiment the conclusion was drawn that 16 Hters of fermentation liquid
was too much for this fermentor, when fat emulsions or fatty acids served as
substrates. Therefore, the next fermentation was performed with a volume of 12
liters culture liquid.
2.6. Second batch fermentation of fatty acids
Preparation of the inoculum by means of slant culture, shake flask culture and
Sovirel fermentation was performed as described in sections 2.2 and 2.3 of this
Chapter. Preparation of the Biolafitte fermentor was done as described in 2.4. This
time, 10.8 liters of water, 200 g of fatty acids and the following amounts of salts
were steriHzed in situ for 20 min at 120C.
(NH4)2S04 5.5 g
NH4H2PO4 22.0 g
MgS04.7aq 3.3 g
ZnS04.7aq 1.65 g
K CI 5.5 g
MnS04.4aq 165 mg
Tlie mixture was cooled and kept at 30C under constant stirring at a rate of 200
rpm in the Biolafitte fermentor.
The Sovirel culture Hquid, this time weighing 938 g, was supplemented with 120 ml
of the solution containing trace elements and 120 ml of the vitamin solution. This
mixture was transferred into the Biolafitte fermentor. Fermentation was started
with an air flow of 120 1/h and a stirrer speed of 600 rpm.
93
The foaming could be easily kept under control for the first 20 hours Addition of
cetyl alcohol, which was tested as an antifoam in this particular case, was necessary
only at the end of fermentation, le after 24 hours By that time the yeast
concentration was 9 x 10* cells/ml and the ammonia consumption tended to
decrease Therefore, the fermentation was stopped already after 26 hours After
centnfuging and washing wath salt solution, 650 4 g of yeast paste was obtained and
that gave 151 0 g dry ceU substance
2 7 Discussion and results of second fatty acid fermentation
As in the first fermentation, the yeast produced had a good smeU, a neutral taste
and was light yellow The total amount of fatty acids consumed was 220 g and
151 Og of yeast were produced This corresponds to a yield of (151 0 220) x 100%
= 68 6%
The nitrogen content of the yeast was 6 7% The nitrogen consumption had been
165 2 ml 6 1 n ammonia solution which means 14 1 g N Hence the uptake of
nitrogen this time had an efficiency of (6 7 x 1 51 14 1) x 100% = 71 7% The
total volume of the oxygen taken up, calculated from the flow of effluent gas and
the oxygen content in this gas, was 195 1 O2 at 30C and 80 cm Hg, this
corresponds to approximately 265 g O2.
The air flow as well as the stirrer speed were kept constant during the whole
fermentation and it was decided to start continuous fermentation under the same
conditions The results of the two fermentations with fatty acids are summarized m
Table IV 2-1
Table IV 2 1 Results of two fermentations with soya fatty acids
Time Volume Average Production
air flow
h 1 1/h g total g/h g/1
1st 21 16 170 200 2 9 5 12 6
2nd 26 12 119 1510 5 9 12 6
From these figures it wdl be seen that because of a lower rate of air flow the growth
rate was slower m the second experiment In the medium used m the first
fermentation (see IV 2 5) the fatty acid concentration was 20 g/1, whereas m the
second fermentation it was 18 g/1 Because of this difference the yield in the second
fermentation was higher (68%) than in the first (63%), althougli the total
production was smaller in the latter case
94
3. Continuous fermentation of fatty acids
3.1. Introduction
The fatty acids used for this experiment had the same properties as those described
in IV 2.1
From the previous experiments it was clear that, in the logarithmic phase, the
amount of fat in the fermentation Hquid was too high and the yeast concentration
too low to start the dUution at that point. We, therefore, decided to proceed with the
batch fermentation until 10 ml of fermentation liquid in the pearshaped test tube
gave at least a volume of 0.5 ml yeast paste after centrifugation. Hence a high
dilution factor could not be chosen. From a theoretical point of view, however,
which is extensively described by Herbert et al. (1956), it is not too important
which dilution rate is chosen because at practically any rate a stable system will
result in which the substrate concentration will be such that the specific growth
rate becomes equal to the dilution rate.
In order to estimate the dilution rate to be chosen, the following calculation was
made.
In the more or less 'steady state' period from 18 22 hours, the amount of
ammonia solution used was 9 ml/h; this means about 9 x 6.1 x 14 mg N/h = 0.77 g
N/h. If the yeast produced contains 7% nitrogen, the production of yeast in that
period has been: (0.77 : 7) x 100 g yeast/h = 11 g yeast per hour.
The final yeast concentration in the fermentation Hquid, in both cases, was 12.6 g/1.
Hence the flow of fermentation liquid should be approximately (11 : 12.6) 1/h =
0.87 1/h. At a higher ammonia consumption, the flow could be higher but care has
to be taken to prevent washing out of the fermentation liquid. For this purpose the
fat concentration in the supernatant after centrifugation was used as a control.
3.2. Preparation of the inoculum
The procedure for the slant culture, the shake flask culture and the Sovirel
fermentation were the same as described in Chapter IV 2.2. After 24 hours the
content of the Sovirel apparatus was ready for use as inoculum.
3.3. Preparation of stock solution
Figure III 1-7 shows the scheme of the apparatus applied for the continuous
fermentation.
The conical flasks of 4 1 were used to keep the stock solution; each contained 3.8 1
water in which were dissolved:
95
fatty acids
(NH4)2S04
NH4H2PO4
MgS04 7 aq
ZnS04 7 aq
KCl
MnS04 4 aq
72
2.0
8.0
1.2
0.6
2.0
60
mg
This solution was sterilized at 120C for 20 min. The fatty acids floated on the
water layer but, when the content was transferred into the Terzano fermentor,
which was used as a buffer tank, it was mixed by manual shaking to ensure that
no fatty acid would remain behind. Just before transferring the mixture into the
Terzano vessel, 40 ml of the vitamin solution and 40 ml of the solution containing
the trace elements were added to the solution in the conical flasks. In the Terzano
vessel, the mixture was kept at 30C.
Continuous stirring at 200 rpm provided a homogeneous emulsion.
3.4. Preparation of the Biolafitte fermentor
The Biolafitte fermentor was prepared as usual, and this time extra connections
were made to transfer the mixture of fatty acids, minerals, trace elements and
vitamins from the Terzano vessel to the Biolafitte fermentor by means of the
Verder pump (see Figure III 1-7).
This fermentor was filled with 10.8 1 of water, 180 g of fatty acids and the salts
mentioned in IV 2.6. After sterUization in situ at 120C for at least 20 min, the
emulsion was cooled to 30C. Then 120 ml of the vitamin solution and 120 ml of
the solution containing the trace elements were added. After the fermentation in
the Sovirel apparatus was completed, 975 ml of its content was transferred into
the Biolafitte fermentor. In the Sovirel liquid, the cell concentration was 6 x lO'
cells/ml and no viable bacteria were present.
3.5. Continuous fermentation
Fermentation was started with an air flow of 120 1/h and a stirrer speed of 600
rpm. After 19 hours, consumption of the 6.1 n ammonia solution was as high as
12.75 ml/h which means a nitrogen intake of 12.75 x 6.1 x 14 mg N/h = 1.09 g
N/h. In a sample the amount of used fat was very low in the supernatant after
centrifugation. The colorimetric measurement gave 2 x 10* ceUs/ml. Supposing the
yeast concentration in the fermentation liquid amounts to 12.5 g/1, and the N
concentration of this yeast to 7.0%, the nitrogen balance is reached when (1.09 :
0.875) 1/h = 1.2 1/h is replaced by fresh medium. This only holds true as long as the
ammonia consumption is constant; this, in turn, means that the yeast ceU
concentration is constant.
96
Hence the pump was started at 1.2 I/h for both: addition of the stock emulsion and
transport of the fermentation Hquid into the 4 1 receivers.
The air flow was increased to 160 1/h because the oxygen content in the effluent
gas had become as low as 9% v/v and it was quite possible that, because of lack of
oxygen, the yeast production would diminish. For the rest of the fermentation
procedure, the air flow was kept at this level and no excessive foaming occurred.
The oxygen content of the effluent gas varied between 12.1 and 13.5% v/v oxygen.
Since the fermentation liquid did contain foam, but the stock solution did not, the
volume in the fermentor rose, and after 40 hours 1 Hter extra was taken away. In
the meantime the analysis of the samples taken showed an increase in the amount
of fat in the supernatant after centrifuging. Therefore, the pump was set at 1 1/h.
The cell concentration measured with the colorimeter was 4x10 * cells/ml.
After 30 h of continuous fermentation, the consumption of ammonia solution had
gradually decreased to 9.5 ml/h and the outflow was accordingly reduced.
Surprisingly, the cell concentration of the fermentation Hquid, measured by the
colorimetric method, showed constant values whereas the amount of fat found in
the supernatant after centrifuging slightly increased. As the volume in the Biolafitte
fermentor had increased, again 1 I of the fermentation liquid was taken away.
After 50 hours the consumption of the ammonia solution was 7 ml/h, and, when
after 65 hours it was as low as 6.5, the fermentation was terminated and the
fermentor emptied. The cell concentration was then 3 x 10* cells/ml and the
amount of fat in the supernatant had increased only slightly.
3.6. Results and discussion
The total yield, obtained after centrifugation of the fermentation liquid, and
washing and drying of the wet paste, was 920 g dry sunstance. This yeast had a
good smeU and taste, and was light yellow. Because of the foam present in the
fermentation Hquid harvested it was difficult to estimate the total amount of this
Hquid.
However, from the Terzano vessel i.e. the conical flasks containing 4 1 stock
solution, in total 56 1 had been added. Together with the 12 1 initially present in the
Biolafitte fermentor this makes 68 Hters of fermentation Hquid, from which 920 g
dried yeast were obtained. Thus the production was 13.5 g/1, which means that the
production of the batch fermentation that preceded the continuous fermentation,
might have been 12 x 13.5 g = 162 g dried yeast. Hence the average production had
been: (920 - 162): 65 g/h = 11.7 g/h.
From the figures above it can be calculated that the average Hquid flow should have
been: (11.7 : 13.5) 1/h = 0.86 1/h. As is mentioned above 56 Hters were added in 65
hours; this means a Hquid flow of (56 : 65)l/h = 0.86 1/h. The amount of fatty acid
was 18 g/1 and hence the yield was (13.5 : 18) x 100% = 75%.
97
On the whole, the continuous fermentation described gave no troubles at all. There
was no excessive foaming, and no antifoaming agent had to be used. The gradual
reduction of the ammonia consumption might follow from the fact that, when the
volume in the fermentor had increased, 1 liter of the fermentation liquid was taken
away, which reduced the amount of yeast rather abruptly.
Nitrogen consumption over the 65 hours of continuous fermentation had been 662
ml of 6.1 n ammonia solution, which corresponds to 0.662 x 6.1 x 14 g N = 56.5 g
N. The yeast weighed 758 g and had a nitrogen content of 6.3%. Hence the yeast
produced contained 47.88 g N. From these figures it follows that the efficiency of
nitrogen uptake had been (47.88 : 56.5) x 100% = 84.7%.
The protein content was 6.25 x 6.3% m/m = 39.4% m/m crude protein. The
amount of oxygen used in the 65 hours of continuous fermentation, calculated
from the air flow and the oxygen content in the effluent gas, was 832 Hter of O2,
measured at 30 and 78 cm Hg; this corresponds to approximately 1100 g O2. In
IV 1.5 the formula of Mateles (1971) has been applied to calculate the relation
between oxygen needed and ceUs produced. If the same is done for the 65 hours of
continuous fermentation, the foUowing values have to be used in the formula:
Y = 0.75;C = 23;H = 42;O = 2.5 and M = 358 (zee IV 2.1); C* = 52.1; H* = 8.5;
0 * = 28.7 and N*= 6.0 (see IV 1.5).
This results in the equation: g oxygen = 2.16 x g ceUs. Hence according to the
formula proposed by Mateles (1971), 1640 g of oxygen would have been needed to
produce 758 g of yeast whereas actuaHy only 1100 g has been used. This might be
an indication that the formula of Mateles cannot be generally applied.
As mentioned in IV 1.5, the construction of the fermentor did not allow a higher air
flow and, for this reason, the specific growth as weU as the dilution rate had to be
kept very low.
A working volume of 12 liter and a Hquid flow of 0.86 1/h means a dilution rate of
(0.86: 12)h-' =0 . 0 7h-' .
The reciprocal value, the retention time, had thus been very long, namely 14 hours,
while normally 5 or 6 hours should be enough. For these reasons it was decided not
to improve the conditions of the continuous fermentation but to try out batch-to
batch techniques.
4. Batchtobacth fermentations of fatty acids
4.1. Introduction
After it had been found that the continuous fermentation of soya fatty acids did
not constitute any particular difficulties, it seemed useful, for the reasons
given in IV 3.6, to examine whether a batch-tobatch fermentation would give
the same results.
98
4.2. Preparation of the inoculum
After the cultures on the GYA slants had grown for 72 hours, and those in the
shake flask for 80 hours, the substrate in the Sovirel fermentor was inoculated.
After 25.5 hours, 990 grams of the contents was used for inoculation. A small
sample was taken for plate counting. It gave 1.2 x 10* yeast cells/ml and less than
10 viable bacteria per ml.
4.3. Preparation of the Biolafitte fermentor
The preparation for the first batch was the same as that described in IV 3.3.
The Terzano vessel was this time prepared to feed the 20 1 fermentor as soon as the
first batch was finished. A special Multifix pump with a capacity of 54 1/h was
installed instead of the Verder pump used for the continuous fermentation.
In the Terzano vessel, which was kept at 30C, the same ingredients were present as
with the experiment on continuous fermentation; the volume was 25 1 and this was
maintained constant, to ensure good homogenicity and constant temperature,
except for the two last batches. Although a smaU amount of insoluble salts was
formed, these did not precipitate and the mixture could be kept homogeneous.
4.4. The fermentation processes
The 11 liter medium was inoculated with 990 g from the Sovirel fermentor. The air
flow was set at 120 1/h and the stirrer at 600 rpm.
The increase in cell concentration was followed by colorimetric measurements and
plate counts. When the cell concentration had reached 5x10 * ceUs/ml, and 10 ml
of fermentation liquid gave hardly any fatty upper layer after centrifugation, the
contents were harvested but for a volume of 1 liter. The fermentation liquid was
then centrifuged and the wet paste washed twice and dried.
The amount of dried yeast obtained from the first 11 Uter was 136.3. Colour and
taste of the yeast were good.
From the Terzano vessel, 13 1 was pumped into the Biolafitte fermentor and the
fermentation was continued, the air flow and stirrer rotation speed being the same
as in the first fermentation.
After 22 hours the colorimetric measurements showed a ceU concentration of 5.2 x
10* cells/ml and on centrifugation of a 10 ml sample no upper fatty layer was
observed. The contents of the fermentor were transferred into the conical flasks
except for 1 liter that was left as an inoculum for the third fermentation. After
centrifugation, washing and drying, 160.3 g of dried yeast was obtained.
The Biolafitte fermentor was filled again, this time with 13 1 of the medium from
99
the Terzano vessel by means of the Multifix pump.
In this fermentation, the oxygen uptake was less than in the foregoing
fermentation, namely 7.5% instead of 9.5% v/v, and apparently the growth was
slower. After 24 hours there were again 5x10 * cells/ml and no upper fatty layer
was found when 10 ml of the fermentation liquid was centrifuged. The
fermentation was therefore stopped; 12.8 liters were taken away, while approximately
1.2 liters was left for the next fermentation.
After washing and centrifugation, the wet paste was dried; this yielded 138.9 gof
yeast dry substance.
In these first three fermentations, all in all 36.7 1 of medium was fermented, giving
435.5 gof dried yeast. Hence, 11.87 grams of yeast per Hter of medium was produced.
The fat concentration was 18.2 g/1. Thus, the yield in these first three fermentation
was (11.87: 18.2)x 100%= 65.3%. The 1.2 liter left was fiUed up to about 141 with
fresh medium from the Terzano vessel. The air flow was again set to 120 1/h but
this time the rotation speed of the stirrer was kept at 800 rpm. This resulted in
a higher uptake of oxygen. The effluent gas contained 12.2% v/v oxygen, whereas
in the 3rd batch the concentration was 13.5% v/v on an average. On the other hand,
it seemed that the fermentation liquid now had a stronger tendency to foam than
in the previous experiments.
After 26 hours, 13 liters of the fermentation liquid was taken away and the rest was
fiUed up to 14 liters with the stock medium.
The yeast produced seemed to contain more fat than in the previous experiments,
according to the result of an organoleptic test, and had to be washed with an extra
amount of salt solution. After drying, the weight of the yeast was 148.3 g dry
substance.
The fifth fermentation was continued for 26 hours under the same conditions as
the previous one, i.e. with an air flow of 120 1/h and stirrer rotation speed of 800
rpm. The average oxygen content in the effluent gas was in this experiment 12.2%.
After washing, centrifuging and drying of the wet yeast, 156.9 g of dry substance was
obtained. Compared with the first batch it seemed that the colour was more
yeUow. Due to an extra washing the taste was still good.
The same sequence of five fatty acid fermentations was repeated. The conditions
were kept as constant as possible. In the last part of the fourth and throughout the
fifth fermentation, the stirrer rotation speed was set at 800 rpm.
The results obtained in the previous experiments were confirmed. By the end of the
series, the foaming had increased; this made it necessary either to reduce the air
flow or to increase the rotation speed of the stirrer, or to do both.
The yeast produced was of a good quaHty as far as taste and colour were concerned.
4.5. Results and discussion
Table IV 4-1 summarizes the results of the two sequences of five fermentations.
100
Table IV 4 1 Results of two sequences of fermentations with soya fatty acids
1st Week
Total
Average
2nd Week
Total
Average
Time
h
21
22
24
26
26
119
23
21
25
26
25
120
1st and 2nd
Week
Volume
1
10.9
13.0
12.8
13.0
13.8
63.5
11.6
13.6
12.0
11.5
13.6
62.3
125.8
Average
airflow
1/h
121
121
123
123
121
122
116
124
123
112
135
122
g total
136.3
160.3
138.9
148.3
156.9
740.7
148.8
163.2
139.1
123.9
150.1
725.1
1465.8
Production
g/h
6.5
7.3
5.8
5.7
6.0
6.2
6.5
7.8
5.6
4.8
6.0
6.1
g/1
12.5
12.3
10.9
11.4
11.4
11.7
12.8
12.0
11.6
10.8
II.O
11.6
11.6
From the figures listed in Table IV 4 1 it will be seen that the yeast concentration
varied from 10.9 g/1 to 12.8 g/1 and that the average yeast concentration for both
weeks was 11.65 g/1. Since the concentration of the fatty acids was always 18.2 g/1,
the yield was 64%.
Some additional remarks wiU now be made on the production per hour.
The fact that sufficient time was not always available had as a consequence the result
that the fermentations were not interrupted at the most favourable time, as far as
production was concerned. Because of this, the average amount of yeast produced per
hour amounted to 6.15 g only, which, as compared with the yield of the
continuous process, is very low. The differences in production time also resulted
from the different amounts of inoculum used. Because of foam formation, the
amount of Hquid that was used as an inoculum for the next fermentation could
only be estimated. For the same reason it was, furthermore, rather difficult to
determine the volume of fermented liquid taken away. This volume, however, was
101
checkedonthebasisof the amount of supernatant and the volume of the yeast paste,
on the one hand, and the total amount of medium prepared on the other. This was
125 liters. Combined with the volume of the first inoculum from the Sovirel vessel,
this is in good agreement with the total volume of 125.8 liters (see Table fV 4-1).
The portions of yeast, obtained in 10 fermentations, were mixed and the mixture
analyzed. The nitrogen content was 5.97o m/m, which corresponds with a crude
protein content of 36.9% m/m.
5. Fermentations of tallow emulsions
5.1. Introduction
For the experiments concerned, a sample of an inedible fat product was chosen.
Tlie chemical characteristics of this tallow, estimated by the methods described in
Chapter II, were the foUowing:
Melting point
Moisture
Impurities
Unsaponifiable matter
Saponification value (SV)
Acid value (AV)
Fatty acid content
Fat content
Molecular weight
of fatty acids
Fatty acid profile < C, 6
42.1C
0.1%
0.2%
0.7%
199.3
14.9
7.1
92.2
267.6
4.4%
m/m
m/m
m/m
mg KOH/g
mg KOH/g
m/m
C. 6
C, 6
Ci 6: i
Cl 7
C, 8
C18: 1
Ci 8: 2
4.4%
24.4%
3.9%
2.3%
17.0%
43.6%
4.4%
m/m
m/m
m/m
m/m
m/m
m/m
m/m
The molecular weight based on the fatty acid profile is 272.3, which is in good
agreement with the value of 267.6 calculated from SV and AV figures.
If a molecular weight for the fatty acids of 269 is assumed, the formula of the
fatty acid wUl approximately be Ci7_3H3 302. The corresponding fat formula
would become CssHjoiOe, with a molecular weight of 857.
According to the procedure described in Chapter II 2.4, five portions of 1 liter of
an emulsion were prepared with this fat, starting with 200 g of fat, 20 g of Renex
650 and 900 ml of water each.
After autoclaving, a sample was taken that was analyzed for fat and emulsifier. The
102
fat content was 17.5% m/v and the amount of Renex 650 found was 1.8% m/v.
These sterile emulsions were used for a number of fermentations.
5.2. Preparation of the inoculum
The inoculum was prepared according to the procedure described in IV 1.2. After
the fermentation in the Sovirel fermentor was completed, an amount of 950 ml of
its content was used as inoculum for the Biolafitte fermentor. A small sample taken
from the culture Hquid contained no viable bacteria. Yeast was present at the rather
low concentration of 4 x 10* cells/ml, estimated with the colorimeter.
5.3. First fermentation of tallow emulsion
In the Biolafitte fermentor, 9.7 1 of a salt solution was sterilized as described in IV
1.3. The conical flask this time contained 1.1 liter of taUow emulsion, 110 ml of
the solution with trace elements and 100 ml of the vitamin solution. After addition
of the fermentation Hquid from the Sovirel apparatus, the mixture in the conical
flask was transferred into the Biolafitte fermentor, and fermentation started.
Tlie air flow was 120 1/h and the stirrer speed 600 rpm. It seemed that this
emulsion had a tendency to foaming; the air flow was, therefore, reduced to 1101/h
which was appropriate to prevent overfoaming.
After 24 hours, however, when the consumption of the ammonia solution showed
that the fermentation was near its end, excessive foaming suddenly occurred and
fermentation had to be stopped. It was decided to keep 1 liter of the contents for
the next fermentation.
Tlie total amount of fermentation liquid that had to be centrifuged was 9.5 1. The
foam contained fat in a rather high percentage, which made it necessary to wash the
yeast three times with salt solution before drying. Consequently, the amount of
yeast after drying was only 78.7 g dry substance. The nitrogen content was 6.75%
m/m, which corresponds to a crude protein content of 42.2%.
5.4. Second fermentation of taUow emulsion
An amount of 12 1 of medium was added to the residue (1 liter) of the first
fermentation. The air flow chosen was 75 1/h and the stirrer speed 800 rpm.
Fermentation proceeded without any difficulty. The lower air flow caused a lower
oxygen concentration in the effluent gas, i.e. an average of 7.3% v/v.
After 21 hours, foaming suddenly occurred and, to prevent excessive foaming, the air
flow had to be stopped for 10 minutes; after that time it was not possible to
continue with an air flow of more than 40 1/h and, for that reason, fermentation
103
was terminated. Again 1 Hter of the fermentation liquid was left in the fermentor,
to be used as inoculum for a third fermentation. In centrifuging the culture liquid, a
rather fatty yeast paste was obtained; it had to be washed three times with a warm
salt solution before drying.
Tlie amount of dried yeast was 128.9 grams. The nitrogen content was 5.85% m/m,
corresponding to a crude protein content of 36.4%. The taste of the yeast was
nearly neutral.
5.5. Third fermentation of tallow emulsion
The fermentor was again filled with 12.5 1 of medium, and fermentation started.
Tills time the air flow was reduced to 40 1/h for the first 16 hours. After that
period, the oxygen in the effluent gas had become as low as 2.3% v/v and,
therefore, the air flow was increased to 65 I/h for the last 10 hours. Hence the
average air flow was 501/h. The oxygen content in the effluent gas became 3.3% v/v
and increased to 4.4% v/v over the last two hours.
After 27 hours of fermentation, the yeast concentration, measured by the
colorimetric method, was 5 x 10* ceUs/ml, and the fermentation liquid suddenly
foamed over; fermentation had then to be halted. The liquid, which amounted to
13.5 1, contained less fat than in the previous tallow fermentations, and after
centrifugation the yeast paste had to be washed only twice before drying. The
weight of the dried yeast was 140.2 grams. The nitrogen content of this yeast was
6.30% m/m, which meant a crude protein content of 39.4% m/m.
5.6. Results and discussion
The mixture of the yeast of the three batches had a protein content of 38.9% m/m
and had still a rather fatty taste. The colour and smeU of the yeast mixture was
good.
Table IV 5.1 summarizes the results of the second and third fermentations.
Table IV 5-1 Results of two fermentations with taUow emulsions
Time Volume Average Production
air flow
h I 1/h g total g/h g/1
2nd
3rd
21
27
12.5
13.5
75
50
128.9
140.2
6.1
5.2
10.3
10.4
104
Tlie concentration of the taUow in the fermentation Hquid being 17.5 g/1, the yield
obtained was (10.35 : 17.5) x 100% = 59%, which is low. It seems highly probable
that the yeast in these experiments had been cultivated with a Umiting amount of
oxygen.
6. Fermentation of technical fats
6.1. Introduction
Regularly found among the fats which are inedible and can be used for technical
purposes only, are the fats that have been used for deep frying.
We used a sample of such fat, obtained from a cafetaria, as raw material for
fermentation.
The analysis of this sample according to the methods mentioned in Chapter II gave
the following results:
Melting point
Moisture
Impurities
Unsaponifiable matter
Saponification Value (SV)
Acid Value (AV)
Molecular weight
of fatty acids
Fatty acid profile < C , 6
Cl 6
Ci 6 : i
C, 7
C, 8
Ci 8 : i
Ci 8: 2
> C, 8
44.0 C
0.1%
0.3%
0.4%
199.6
4.1
268.2
5.4%
40.6%
2.9%
0.2%
9.3%
36.9%
2.4%
2.5%
m/m
m/m
m/m
mg KOH/g
mg KOH/g
m/m
m/m
m/m
m/m
m/m
m/m
m/m
m/m
The molecular weight based on the fatty acid profile is 268.5, which is in good
agreement with the value of 268.2 calculated from the SV and AV figures. The
amount of chlorinated pesticides was below the detectable limits. If a value of
269 is assumed for the molecular weight of the fatty acids, the molecular weight of
the corresponding fat becomes 845. The formula of the fatty acid can be
represented by C17H33O2 and that of the fat by C54H101O6. This fat was
emulsified according to the procedure described in Chapter II 2.4 and steriHzed at
120C for 20 min, starting with 200 grams of fat, 20 g of Renex 650 and 900 ml of
water. A number of emulsions were prepared. The final product was analyzed and
found to contain 17.8% m/v of fat and 1.8% m/v of Renex 650.
The emulsions were used in the following two fermentations.
105
6.2. Preparation of the inoculum
In order to avoid the fermentation in the Sovirel fermentor, on this occasion 6 slants
were inoculated. After incubation for 3 days, the yeast cells produced were washed
with 10 ml saU solution into 6 shake flasks of 750 ml total volume, each containing
the appropriate salts, trace elements and vitamins. After 72 hours' growth the
contents of these shake flasks were used for inoculation of the medium in the
Biolafitte fermentor.
6.3. Fermentation of fat emulsions
In the Biolafitte fermentor 10.5 1 solution, containing
(NH4)2S04 6.0 g
NH4H2PO4 2.4 g
MgS04 7aq 3.6 g
ZnS04 7aq 1.8 g
K CI 6.0 g
MnS04 4aq 0.18 g
was sterUized at 120C for 20 min. After cooling to 30C, 1200 ml of the fat
emulsion and 120 ml of each of the vitamin and trace element solutions were added.
The medium mentioned above was inoculated with 900 ml fermentation liquid
from the shake flasks and the fermentation started.
The concentration of the yeast ceUs was, after one hour of fermentation, 5x10 *
cells/ml and therefore the air flow was set as high as the foaming permitted, i.e.
between 115 and 120 1/h. After about 10 hours, this flow was decreased to 48 1/h.
Tlie rotation speed of the stirrer was 800 rpm tiiroughout the experiment.
After 24 hours the cell concentration was 8x10 ^ cells/ml, and fermentation was
continued for another 5 hours. When after 29 hours the fermentation liquid started
foaming excessively, the air flow was reduced, and 10.9 1 was taken away from the
fermentor and centrifuged. The yeast paste obtained was washed and dried, giving
111.2 g of yeast dry substance. To the residual culture liquid in the fermentor was
added 11.5 liter of medium, having the same composition as mentioned above.
This time the ceU concentration after 1 hour was 3x10 ^ cells/ml. The airflow was
then set at 55 1/h and the stirring speed at 800 rpm.
After 27.5 hours, fermentation could be halted without any noticeable foaming.
The amount of fermentation liquid obtained was 13.3 1 and this gave, after
centrifugation, washing and drying, 137.2 g of dried yeast.
6.4. Results and discussion
Tlie results of these fermentations, performed with a rather low air flow are
summarized in Table IV 6- 1.
106
Table IV 6-1 Results of two fermentations with technical fat emulsions
Time Volume Average Production
air flow
h 1 1/h g total g/h g/1
1st
2nd
29
27.5
10.9
13.3
63
55
111.2
136.0
3.8
5.0
10.2
10.3
From these figures it is evident that growth of the yeast with a reduced air flow
results in a low production and an unsatisfactory yield of only (10.2 : 17.8) x 100%
= 57.3%. The mixture of the yeast obtained with these two batch fermentations
had a neutral taste, a light yellow colour and a good smeU.
The nitrogen content was 6.88% m/m, corresponding to a crude protein content of
43.0% m/m.
107
CHAPTER V
5. PROPERTIES OF PRODUCED YEAST
5.1. Organoleptic examination
It has been suggested in Chapter 1 that the organoleptic properties and the
palatability of SCP is important, even when it is to be used in animal feed only. For
that reason, a batch of every fermentation has been tested after each of the
foUowing processes: centrifuging; washing; drying; and storage in dried form.
Mostly the odour was neutral. Sometimes, however, an odour was apparent
reminisent of the odour of cheese. The colour of the dried yeast was light to
dark yeUowish, depending on the raw material used.
The taste of the yeast paste was usuaHy neutral. Sometimes, however, it was fatty;
the yeast was then washed twice, reducing its fat content enough to eliminate the
fatty taste of the paste. The dried product was tasteless. It has to be stored at 5C
to prevent oxidation of the fat present in the yeast.
To summarize the examination of the yeast does not reveal any organoleptic
properties that make special treatment of it necessary before use in animal feed.
5.2. Chemical composition
Table V 2-1 gives some characteristics of the dried yeast.
m/m
m/m
m/m
m/m
m/m
m/m
m/m
From these figures the overall composition and the gross formula of the yeast can
be calculated if the formulas of the components are known. For carbohydrate, the
formula C6H12O6 can be taken; for fat the formula CsgHiosOg is derived from
the fatty acid composition, and for the protein the formula C4HgN0i.. is chosen
on account of the nitrogen content in the crude protein.
On the basis of these presumptions, the foUowing calculation scheme was used:
3.3 g water contains: (3.3:18)x 2 gH = 0.37 g H
and: (3.3:18)xl6 gO = 2.93 gO
Table V 2-1
Moisture
Crude protein
Fat
Carbohydrate
Crude fibre
Emulsifier
Ash
Composition of t
3.3%
37.7%
31.8%
19.6%
1.0%
1.0%
4.7%
108
37.7 g protein contains:
and:
and:
and:
31.8 g fat contains:
and:
and:
20.6 g carbohydrate contains:
and:
and:
(37.7x87.6)x 8
(37.7:87.6)xl7,6
(37.7:87.6)x48
(37.7x87.6)xl4
(31.8:873)xl05
(31.8x873)x 96
(31.8x873)x672
(20.6
(20.6
(20.6
180)xl2
180)x96
180)x72
gH
gO
gC
gN
gH
gO
gC
gH
gO
gC
' 3.44 gH
7.57 gO
20.66 g C
6.03 gN
3.82 gH
3.50 gO
24.48 g C
= 1.37 gH
= 10.99 gO
= 8.24 gC
If the above figures are totaUed, we arrive at: 9.0 g H; 25.0 g 0 ; 53.4 g C and
6.03 g N in 100 g yeast. The elementary analysis, carried out by the Institute for
Organic Chemistry TNO gave: 8.5 g H; 28.7 g 0 ; 52.1 g C and 6.0 g N.
In view of these results the overall formula of the yeast, apart from ash, might with
good approximation be simplified to Ci i H21NO4.
The crude protein of the yeast was analyzed for the amino acid composition, the
content of RNA and the purine bases. In Table V 2- 2, the amounts of amino acids
present in the mixture obtained on hydrolysis of the yeast are compared with the
FAO figures for hen's egg (1973) and the figures given by Lunven (1973).
Table V 22 Essential amino acids present in mg/gN in yeast and hen's egg
according to the FAO (1973) and Lunven (1973).
Isoleucine
Leucine
Lysine
Total aromatic
amino acids
Total sulphur
amino acids
Tlireonine
Tryptophan
Valine
Yeast mixture
I
335
469
489
521
192
366
84
374
Hen's egg
FAO 1973
II
338
535
439
578
353
296
108
413
Ratio
I/II
0.99
0.88
1.11
0.90
0.54
1.24
0.78
0.91
Hen's egg
Lunven
1973
348
548
477
514
338
313
121
378
The guanine content of the yeast was 0.54% m/m, whereas for the adenine content
0.60% m/m was found. The amount of RNA found by analysis was 3.9%.
109
Calculated from the purine bases, it was 5 1'% m'm In both cases the amount of
DNA is neglected. For the calculation, the formula RNA = 4 5 (guanine + adenine) is
used. This is based on the assumption that guanine, adenine, cytosine and uracil are
present m equimolecular amounts.
After extraction, the fat in the yeast was also analysed It contained 1.2% m/m
unsaponifiable matter. The composition of the fatty acids is given in Table V 2-3
Table V 2-3
< C, 6
Cl 6
C, 6 :
C, 7
total
Fatty acid
= .
0 =1 5 6,
1 = 27,
= 0 5,
saturated
mono-unsaturated
poly-unsaturated
profile of
ClB' O
C| 8 1
Ci 8: 2
Ci 8: 3
32.3 or
38.1
27.7
yeast lipids in g per I (
= 16 3 , C2o:0
= 35 0 , C20 . 1
= 2 7 4 , C20-2
= 0 2 , > C20
100%
118%
86%
)0g
=
=
=
=
0.4,
0.4,
0 2,
1.5,
These results show that the odd numbered fatty acids were present only in small
quantities. The ratio between polyunsaturated and saturated fatty acids was 0.86,
whereas nearly 40% mono-unsaturated fatty acids were present
The average molecular weight of the fatty acids, calculated from the figures in Table
V 2- 3, IS 278.3 This means that the corresponding fat has a weight of 873 For the
fat concerned a gross formula of CseHiosOe is obtained and this formula is used
for the determination of the gross formula of the yeast (see above).
Analysis of the ash gave the following results, referred to dry yeast:
Ca = 145 ppm, Mg = 710 ppm, Zn = 1175 ppm, Mn = 175 ppm, Cu = 16 pp, Fe =
109 ppm and Ni = 4 ppm.
From these figures can be derived the Mol-percentages, that give a better
impression of the relative quantities, namely
Mg = 52 Mol%, Zn = 32 Mol%, Mn = 6 Uo\7c, Ca = 6 Mol%, Fe = 4 Mol% and the
sum of Cu and Ni is < 1 Mol %
Calcium originates from the tap water used and its concentration is very low as
compared with the concentration of phosphorus, which was found to be 2.2% m/m,
expressed as P2OS For that reason, our toxicity study was done with an extra
amount of calcium carbonate in the diet
The iron originates from the fats used because no particular iron compounds were
added to the fermentation media. The vitamins of the Bgroup were present in the
yeast, in the following amounts, referred to dned yeast
110
Thiamin 18.9 ppm; riboflavin 225 ppm; pyridoxine 6.7 ppm: niacin 200 ppm;
pantothenic acid 85 ppm; folic acid 8.8 ppm and inositol 3000 ppm.
Since this yeast was a mixture of products, obtained after fermentation of fat as
well as fatty acids, the amount of emulsifier originating from the fat emulsion had
to be determined; it was 1% m/m.
The raw fats used as fermentation substrates sometimes contained pesticides. The
analysis of the yeast produced for DDT, lindane and HCB showed that the last
pesticide was present in a concentration of 0.14 ppm only, whereas the others were
absent, or present in concentrations lower than 0.1 ppm.
The average values for the yeast mixture mentioned above are, as a whole, also
representative for the various batches. For example the protein content of the
mixture is 37.7% m/m and the values of the separate batches show a variation
between 36.4 and 43.0% m/m (see Chapter IV).
5.3. Biological characteristics
5.3.1. Protein calories per cent
On the basis of the composition of the dried yeast given in Table V 21, the yeast
can be considered as a foodstuff containing 37.7% protein, 19.6% carbohydrate and
31.8% fat. The total energy obtained in metaboHsm of this yeast, expressed in
kJ/lOO g dry cell substance can be calculated from its composition, using the
factors 17 kJ, 18 kJ and 38 kJ for protein, carbohydrate and fat. The result is:
37.7 X 17kJ + 19.6x 18 kJ + 31.8 x 38 kJ = 2202.1 kJ/lOO g or 2.2 MJ/lOOg.
The protein calories per cent has the same value as the protein Joule per cent and is
found by dividing the protein energy by the total energy of the foodstuff and
then multiplying by hundred. Hence protein calories per cent becomes:
(640.9 : 2202.1) X 100% = 29.1%.
5.3.2. Chemical score
Table V 31 gives the amount of each essential amino acid, present in the yeast
mixture, in two ways. Namely: the amount expressed as g/16 g N and the relative
amount expressed as % m/m of total essential amino acids. The same is done for the
amino acids present in the provisional amino acid scoring pattern proposed by FAO
(1973). The relation between both values is an indication for the quality of the
yeast protein and shows that the Umiting essential amino acids are in the first place
the sulphurcontaining ones and, secondly, leucine.
I l l
Table V 3-1 Relation between relative amounts of essential amino acids in the
yeast mixture and in the FAO provisional amino acid scoring pattern
(1973)
Jtein
: E)xlO0
1
11.8
16.6
17.3
6.8
18.4
12.9
3.0
13.2
100.0
FAO standard
g/16gN
4.0
7.0
5.5
3.5
6.0
4.0
1.0
5.0
36.0
(A
: E)xlOO
II
11.1
19.4
15.3
9.7
16.7
11.1
2.8
13.9
100.0
Ratio
I/II
1.06
0.86
1.13
0.70
1.10
1.16
1.07
0.95
The Chemical score based on these figures is 70%. For that reason the determination
of the biological value (BV) of the yeast was also performed with addition of 0.3%
methionine to the yeast. This addition results in an increase of the relative amount
of the sulphurcontaining amino acids from 3.08 g/16 g N to (3.08 + 30 : 38)
g/16 g N = 3.87 g/16 g N. The Chemical score corresponding to this figure is:
(3.87 : 46.1) X (100 : 9.7) x 100% = 86.6%.
Addition of methionine raises the Chemical score of the sulphur-containing amino
acids to the same level as the second limiting amino acid leucine.
The biological value, calculated according to the formula BV = 102 - 0.634 x X, in
which X is the percentage of amino acid deficit of the limiting amino acid(s), is
increased from 83.0 to 93.1 after addition of methionine.
5.3.3. EAA-lndex
As is mentioned in Chapter U 4.2.3, Oser (1951) has proposed an amino acid index
which may be used for characterising the biological value of a protein foodstuff. In
Table V 3-2 this method is applied to calculate the EAA-Index of the yeast
mixture obtained in our experiments. The amino acid content is compared with the
standard used by Oser, taking into account the same amino acids as considered by
him.
g/16gN
Isoleucine
Leucine
Lysine
Total sulphur
amino acids
Total aromatic
amino acids
Threonine
Tryptophan
Valine
5.36
7.51
7.82
3.08
8.34
5.85
1.34
5.99
45.3
112
Table V 32 Calculation of the EAA-lndex of the yeast mixture
Lysine
Tryptophan
Isoleucine
VaHne
Arginine
Meth. + Cyst.
Threonine
Leucine
Yeast mixture
g/16gN
7.82
1.34
5.36
5.99
5.32
3.08
5.85
7.51
Phenynafianine 4.63
Histidine
lOxlogEAA-
log EAA-
EAA-
2.15
-Index
Index
-Index
Oser 1951
g/16gN
7.0
1.5
7.7
7.2
6.6
6.4
4.3
9.2
6.3
2.4
Ratio
%
111.7
89.3
69.6
83.2
80.6
48.1
136.5
81.6
73.5
89.6
Log ratio
%
2.00000
1.95085
1.84261
1.92012
1.90634
1.68215
2.00000
1.91169
1.86629
1.99388
19.07393
1.90739
80.8%
The biological value (BV), calculated according to the equation EAA-Index =
0.877 x BV + 7.83 gives, when we use the value 80.8 for the EAAIndex, a
biological value of 83.2. It is remarkable that both figures for the calculated
biological value are in such a close agreement although the methods applied are
totally different; not only in the procedure of calculating but also as regards the
composition of the protein used as a standard.
When 0.3% methionine is added to the yeast, the EAA-Index becomes 82.7%
corresponding to a biological value of 85.4.
5.3.4. Protein efficiency ratio
Yeast production on a laboratory scale was, unfortunately, not large enough for all
the nutritional experiments mentioned in Chapter II.
It was, accordingly, decided not to determine the PER value since this method (see
Chapter II) is supposed only to be a screening test, whereas the NPU and TD tests
give more quantitative information.
5.3.5. Net protein utilization (NPU), true digestibility (TD),
and biological value (BV)
As mentioned before, the sulphur-containing amino acids were Umiting in the
yeast mixture, and for that reason the tests were performed with four different
113
diets, namely i) with yeast, ii) with yeast and 0.3% methionine ui) with casein and
iv) without any proteincontaining component (control).
The complete diets and their nitrogen contents are given in Table V 33.
Table V 3-3 Diets in g/100 g for the estimation of NPU and TD values
Yeast Yeast + Casein Control
I II III IV
Yeast
Methionine
Casein
Wheat starch
Sucrose
Cellulose')
Minerals^)
Choline chloride (50%)
Vitamin ADEK-
preparation^)
Vitamin B mixture'')
Soybean oil
N % m/m
21.7
.
- . -
31.3
30.0
6.0
4.0
0.4
0.4
0.2
6.0
1.33
Yeast +
methionine
II
21.7
0.3
.
31.3
30.0
6.0
4.0
0.4
0.4
0.2
6.0
1.38
Casein
III
_. _
.
9.6
43.4
30.0
6.0
4.0
0.4
0.4
0.2
6.0
1.41
53.0
30.0
6.0
4.0
0.4
0.4
0.2
6.0
0.05
Explanation:
1) SolkaFloe, Chemimpo, Amsterdam; 2) According to Jones and Foster (1942);
3) Per 100 g diet: 900 lU vitamin A; 300 lU vitamin D; 6 mg vitamin E and 0.4 mg
vitamin K; 4) Per 100 g diet: 0.5 mg thiamine HCl; 0.6 mg riboflavin; 1.0 mg
pyridoxine HCl; 2.5 mg niacin; 1.5 mg Ca-pantothenate; 0.03 mg biotin; 0.1 mg
folic acid; 0.001 mg vitamin B12 and sucrose.
The animals in this experiment were newly weaned albino rats of the CI VOcolony
(of Wistar origin). They were kept on a stock diet for seven days preceding the test
period. The gain in body-weight during these days was 25 to 30 g per week.
At the beginning of the test, the rats were 27 to 29 days old. The results of the
experiments with three groups of four rats each are summarized in Table V 3-4.
114
Table V 3-4 Results of NPU test with three groups of four rats each
Group A Group B Group C
intake N carcass N intake N carcass N intake N carcass N
I
II
III
IV
1 = ^
2.62
3.44
3.43
0.09
^east, II =y
8.54
10.40
9.77
7.79
east + me
2.49
3.70
3.89
0.09
thionine; III =
9.23
11.03
10.85
7.97
= casein; IV =
2.79
4.50
3.55
0.08
= control.
9.41
11.57
10.22
8.19
The nitrogen contents of the carcasses were derived from the loss of water and the
nitrogen intake is calculated from the amount of intake and its nitrogen content.
From the figures of Table V 34, the NPU value for each group can be derived
with regard to the diet with yeast (1), that with yeast and methionine (II) and the
diet with casein (III), giving the following results:
NPU Group A Group B Group C
I
II
III
34
78
60
54
85
76
46
77
59
These figures prove that without methionine the yeast mixture is inferior to the
casein, but with methionine it is superior.
Table V 3-5 Summarized results for NPU and TD values, based
on a corrected N, related to food intake
1
11
III
IV
Intake
gfood
594
843
771
533
gN
7.90
11.64
10.87
0.26
g
82
115
55
41
Faeces
%N
2.50
2.48
2.14
1.55
gN
2.05
2.85
1.18
0.64
Correc-
ted
gN
0.71
1.01
0.93
0.64
Carcass
gH20 gN
605 27.18
733 33.00
685 30.84
532 23.95
NPU
44.2
80.0
65.8

TD
83.0
84.1
97.7

I = yeast; II = yeast + methionine; III = casein; IV = control
115
Table V 3-5 shows the figures for intake, faeces, and carcass N, those given in Table
V 34 are recalculated for 12 rats per diet. The corrected N, used for calculation of
the TD, IS derived from the faecal N of group IV using a correction factor equal to
the ratio of the food intake, assuming that the endogenous N would be
proportional to the food intake. The figures Hsted in this table prove, however, that
there is a notable difference in the relation g faeces/g food intake between the
control group (IV), the two groups on a yeast-diet (I and II) and the group with
casein (III). For the various diets, this relation is
I II III IV
0.138 0.136 0 071 0 077
On the assumption that the endogenous N is more influenced by the faeces
excreted than by the food intake, the correction factor must be related to the
amounts of faeces. The following results are thus obtained'
Faeces in g
Calculated N
Corrected N
TD values
I
82
2.05
1.28
93 3
II
115
2.85
1.79
90.9
III
55
1.18
0.86
96.8
IV
41
0.64
0 64
In table V 3-6, the values mentioned above are used for calculating the biological
value (BV), which is obtained from the values of the NPU and the TD by means of
the formula BV = (lOOxNPU) TD (see Chapter II 4.4 1).
Table V 36 Two different calculations of TD and BV values,
based on the same experimental results.
I II III
a b a b a b
NPU
TD
BV
44.2
83.0 93 3
53.2 47.4
80.0
84.1 92.9
97.0 88 0
65.8
97.7 96 8
67 3 68 0
Explanation
a endogenous N correction related to food intake
b endogenous N correction related to amounts of faeces
From the figures listed in Table V 3-6, it is evident that, as far as the yeast
mixture is concerned, there is a notable difference between the BV values derived
from the Chemical score (83.0), the EAA-Index (83.2), and the values based on the
biological test (47-53)
116
For the yeast mixture with methionine, the BV value derived from the Chemical
score is 93.1, which is in remarkably good agreement with the values obtained in the
biological test (88 and 97); the BV value derived from the EAA-lndex is 85.4 and
this agrees with the lower value based on the results of the biological test.
5.4. Bacteriological examination
The mixture of the various batches produced in the course of the study was not
examined for bacteriological contamination. However, it seemed interesting to
investigate whether yeast, after production under sterile conditions, could easily be
prevented from bacterial contamination during procedures of centrifuging, washing
and drying in a laboratory. Therefore, a sample from an arbitrary batch production
was taken for bacteriological testing.
In Table V 41 the results obtained are compared with the FAO standards,
published in PAG GuideHne No. 5 (1969-1972).
Table V 41 Results of sanitary control of produced yeast, compared with
standards from PAG Guideline No. 5 (1969).
Results
1 x 10^
< 1
<10
<10^
<10
<10
< 1
absent in
absent in
/g
/g
/g
/g
/g
/g
/g
lOg
25 g
Pass
criteria
2x10 * /g
<10 /g
<10^ /g
<10 ' /g
<10 /g
absent in 0.1 g
absent in 1 g
absent in 10 g
absent in 25 g
As wiU be seen from Table V 4 1, the results of bacteriological testing of the yeast
fulfill the requirements.
5.5. Toxicological indication
5.5.1. Introduction
Chapter 11 described the extensive programme of long-term studies, necessary
Total aerobic count
Aerobic spore count
Clostridia group
Lancefield group D
streptococci
Mould spores
Enterobacteriaceae
Staphylococcus aureus
Escherichia coli
Salmonella
117
before any given product can be permitted to be used for animal or human
consumption.
The yeast available was a mixture of yeast batches produced on a laboratory scale
under vanous experimental conditions from a number of different raw materials
It can be expected that yeast, produced on a pilot plant scale under standard
conditions may have different qualities depending on the raw material used.
For this reason, our safety evalution of the yeast produced was restricted to a
simple screening test which, of course, does not give fuU information but only
indicates the toxicity of the product. When such a simple test has been done with a
number of batches and the results obtained are favourable, more extensive studies
should be conducted with yeast produced under standard experimental conditions
on a pilot plant scale Such studies are described in Chapter II.
5.5.2. Screening test
The yeast mixture was incorporated into a diet at a level of 60%, the control diet
contained the corresponding amount of protein from soybean meal Table V 51
gives the composition of the diets.
Table V 5-1 Diets in g/100 g for the sub-acute feeding trial
Yeast diet Control diet
Yeast
dl-Methionine
Soybean meal
Wheat starch
Cellulose')
Minerals^)
Chohne chlonde (50%)
Vitamin AEDK--preparation )
Vitamin B mixture*)
CaCOs
KH2PO4
Soybean oil
60.0
0.2
.
33 0
20
0 5
1 0
0.6
0 2
25
_. _
.
0 2
50.0
24 2
2.0
0 5
1.0
0 6
0 2
2.3
1 0
180
Explanation
1) SolkaFloe, Chemimpo, Amsterdam, 2) According to Jones and Foster (1942),
3) Per 100 g diet 1350 lU vitamin A, 450 lU vitamin D, 9 mg vitamin E and 0.6
mg vitamin K, 4) Per 100 g diet 0 5 mg thiamine HCl, 0.6 mg riboflavin, 1.0 mg
pyridoxine HCl, 2.5 mg niacin, 1.5 mg Ca-pantothenate, 0 03 mg biotin, 0.1 mg
fohc acid, 0 001 mg vitamin Bj 2 and sucrose
118
Each of the diets was fed to a group of 8 male and 8 female rats of the
CIVOcolony (of Wistar origin). Diets and tap water were provided ad libitum.
Table V 52 shows the average body weights and Table V 5-3 the food intake and
food efficiency.
Table V 52 Average body weights during screening test
Mean body weights in g of rats at end of week no:
Control diet
Yeast diet
Control diet
Yeast diet
0
48
48
48
48
1
71
63*
64
61
2
ma l e s
99
84**
f e ma l e s
83
78
3
117
101*
97
90
4
143
121**
111
107
* = P<0.05; ** = P<0.01 (Student t-test)
Table V 53 Food intake and food efficiency during screening test
Control diet
Yeast diet
Control diet
Yeast diet
Food intake in
in week
1
5.7
4.4
5.4
4.3
no
2
9.0
8.2
8.0
7.7
g/rat/day
3 4
ma l e s
9.7 10.3
9.7 10.5
f e ma l e s
8.6 8.6
8.3 8.9
Food efficiency
during
gain
95
73
63
59
week 1 to 4
food
243
230
214
204
gain
food
0.39
0.32
0.29
0.29
From these Tables it wUl be seen that, for both sexes, body weights of animals on
the yeast diet were lower than of those on the control diet, but that the food
efficiency was less favourable only with males (Student ttest).
The food intake on the yeast diet was lower than that on the control diet in the
119
first week only. As may be seen in Table V 5-4, there is no significant difference in
the heamoglobin concentration between the two groups. The organ weights of the
groups on the yeast diet showed slightly increased ratios for liver and kidneys.
Table V 5-4 Terminal body weight, haemoglobin concentrations (Hb) and
organ-tobody weight ratios in screening test
Body weight Hb Organ-to-body weight x 100%
in g/rat in g/100 ml heart liver kidneys spleen
ma l e s
Control group 146 13.5 0.39 4.37 0.90 0.28
Yeast group 123** 13.8 0.40 5.04***0.98 0.29
f e ma l e s
Control group 112 15.3 0.40 4.13 0.90 0.26
Yeast group 105 14.9 0.42 4.96** 0.98* 0.27
* = P<0.05; ** = P<0.01; *** = P<0.001 (Student t-test)
At autopsy the males of the yeast-fed group showed pale livers. In the female
control group rats a greenish discolouration of the kidneys was observed.
MicroscopicaUy, no changes were observed in the kidneys that could be attributed
to treatment with yeast. Compared with the control animals, the males and females
of the yeast-treated group generally exhibited a liver cell alteration characterized
by a smooth, less vacuolar cytoplasm. This finding was not associated with any
morphological sign of ceUular injury and, therefore, not considered to represent a
toxic effect.
To summarize the results of the present study, it appears that, in spite of the very
high level of 60% yeast in the diet, no toxic effects occurred. However, the males
showed depressed growth and a somewhat decreased food efficiency. The relative
weights of liver and kidneys were increased, and the microscopy of the liver showed
slight cytoplasmic changes, which were considered to be of a non-degenerative
character without pathological significance.
120
CHAPTER VI
DISCUSSION AND CONCLUSIONS
1 Discussion of the results
The previous chapters have shown that it is possible to utilize fats, fatty avids and
mixtures thereof for the cultivation of yeast biomass The properties of the yeast
produced were examined extensively, the results lead to the conclusion that, with
the strain selected, a product can be obtained that may be used as a protein source
m animal diets
The media developed have proved to be very useful For Mg, Zn and Mn some
quantitative data were obtained, they show that the uptake of these elements is
between 86% and 88% The composition of the vitamin solution applied is not
based on our experiments It might well be that a reduction of the amounts of
vitamins will give satisfactory results Experiments in this direction are under way
The temperature chosen has its drawbacks, as has been discussed in Chapter III It
IS remarkable, however, that for the large scale productions of SCP described so far,
a fermentation temperature of 30C is employed Nevertheless it seems worthwhde
to devote more attention to this aspect, especially because various microbial strains
are described than can work at higher temperatures
The yield of the yeast production, le the amount of yeast obtained from 100
grams of fat, fatty acid or fat product, did vary with the raw material used The
highest yield, namely 75 g, was obtained with fatty acids of soya origin, used in a
continuous process In batch processes, this product gave a yield of 64 to 68 g
The emulsions of fats gave a lower yield This is most probably due to the fact that
the air supply was limiting The fermentors used did not allow a higher air flow
because of excessive foaming of the fermentation liquid For grease yield values of
69 g and 58 g were obtained The yields obtained with tallow and technical fat were
59 g and 57 g Special attention has stiU to be given to the foaming capacity of fat
emulsions, in constructing fermentors to be used for fat fermentation
2 Economic considerations
Numerous investigators have studied the apphcation of plant and microbial protein
in the diets of veal calves (Van Hellemond, 1967, Colvin, 1968, Chassin, 1969,
Goril and Nicholson, 1969, Roy, 1970, Huber et al , 1970, Van Weerden 1970,
Moril et al, 1971, Nitsan, 1971, Kirchgessner and Roth, 1973,Gropp, 1973, etc )
On the basis of results, reported by the authors concerned it may be expected that
at least 10% of the skimmed milk powder used in the diets for veal calves can be
replaced by SCP products
121
Moreover, it is evident from the results mentioned that, when skimmed milk
powder is partly replaced, the diets have to be changed as regards other components
as well because the composition of skimmed milk powder differs from that of SCP
products with respect to protein, carbohydrates and fats.
Table VI 2-1 Composition and prices of skimmed milk powder, whey powder and
refined animal fat, on September 3rd 1974
Product Price in Components in kg per 100 kg product
Nfls/lOOkg Protein Carbohydrate Fat Rest
50.4 0.5 13.1
73.4 0.8 14.4
100.0
1) Official EEC price for the period March 1974 to March 1975
This is iUustrated in Table VI 21 which gives the composition and prices on
September 3rd for skimmed milk powder, whey powder and refined animal fat. The
composition is derived from the "Feedstuff Yearbook 1973" (Noordervliet) and
the information about prices has been obtained from personal communication with
various manufacturers on September 3rd, 1974.
Table VI 22 shows the composition and the costs of two diets for calves. The
calculation of the price of the yeast is based on the assumption that the total costs
of both diets should be the same. The composition of the yeast is that mentioned
in Chapter V of this thesis. The values given in this table iUustrate once more,
that the diets have changed not only with respect to content of skimmed
mUk powder and yeast. One should therefore not compare these products with
respect to their protein content only. In fact, as far as the overaU composition is
concerned, 10% of skimmed milk powder and 2.6% of fat can be replaced by 8% of
yeast and 4.6% of whey powder.
This is a very important fact when the costs of the products are considered.
To achieve a yield of 70% an amount of 1.35 kg of fat is necessary for production
of 1 kg of yeast. For the price of the fats and fat products in this thesis, an average
can be taken of Nfls 1.20 per kg. Since the price of the yeast, given in Table VI 2-2
is Nfls 3.38 per kg, the production costs should be below Nfls 1.67 per kg in order
to make the process economicaUy feasible. As a matter of fact it is not at aU
impossible that this can ultimately be achieved.
Refined
animal fat
265 36.0
Skimmed
mUk powder')
Whey powder 72-74 11.4
146-150
Table VI 2-2 Composition and costs of diets for calves with and without yeast
Skimmed
milk powder
Whey powder
Refined
animal fat
Rest
Product
kg/100 kg
50.0
20.0
22.6
7.4
100.0
Costs
Nfls
132.50
14.60
33.45
p.m
180.55
p.m
Components in kg
Protein
18.0
2.3
- . -
- . -
20.3
per 10
Carbohydrate
25.2
14.7
- . -
-.
39.9
0 kg produ
Fat
0.25
0.15
22.6
.
23.0
ct.
Rest
6.55
2.85
- . -
7.4
16.8
Yeast
Skimmed
milk powder
Whey powder
Refined
animal fat
Rest
8.0
40.0
24.6
20.0
7.4
106.00
17.95
29.60
+ p.m.
180.55
+ p.m.
14.4
2.8
___
-.
20.2
20.2
18.1
.
.-
39.9
27.00 3.0 1.6 2.6 0.8
0.2 5.2
0.2 3.5
20.0
7.4
100.0 180.55 20.2 39.9 23.0 16.9
3. Final conclusion
The utilization of fats and fat products as a substrate for the production of SCP by
means of yeast fermentation may well be considered realizable. Production on a
pUot plant scale could now be worked out, although this wUl call for high
investment costs.
The calculations made in section VI 2 show that, at this very moment such an
investment might look worthwhile. It is, however, impossible to predict how prices
wiU change over the years which will be needed for development of yeast production
on an industrial scale. To illustrate this statement. Table VI 3-1 compares the prices
of the various products concerned in the beginning and at the end of the present
study.
123
Table VI 31 Average prices of various fats and fat
products in Nfls per 100 kg product
Grease
Tallow
Soya fatty
acids
Technical
fat
Septe
1971
. ^ 0
63
43
25
mber September
1974
>
116
113
120
70
Furthermore it is evident that the difference between the estimated costs of the
yeast production and the estimated market price is too small to invite people to
take the risk of the necessary investments (see Table VI 2-2). In this respect one
should remember that the price for the skimmed mUk powder, given in Table V
2 1 is not actuaUy officially accepted in the agricultural policy of the Common
Market.
The EEC regulation 716/74 fixes the aid for skimmed milk powder for use as
animal feed at Nfls 1.15 per kg. Taking this reduction into account, the price of
yeast calculated from the diets given in Table VI 22 becomes Nfls 1.94 per kg and
this reduces the possible production costs from Nfls 1.67 to Nfls 0.23 per kg. This
means that, after all, the feasibOity of the process is not only influenced by the free
market prices, but that it also largely depends on Common Market regulations
which ultimately are based on political decisions.
It has been announced that the new 100,000 t/y plant of the Bntish Petreoleum
Ltd. (B.P.) for the production of SCP in nparaffins, which is to be built in Sardinia,
will be a joint enterprise of B.P. and the Italian Government. This will be a guarantee
that, if necessary, political decisions can be made to protect the price of the SCP
produced. It is quite understandable that this very guarantee for the price is a
necessary condition to raise the funds for this new project of B.P.; it is furthermore
desirable in view of the contribution that the product implies to world
protein production.
The view that the production of microbial protein is of the utmost importance for
the world food situation is another essential pre-requisite for the realization of the
production of single ceU protein on fats and fat products.
124
LIST OF REFERENCES
Abbot, J.C. (1966) in Worid Protein Resources, p. 1; Am.Chem.Soc;
Washington D.C. 1966
Adler, E. et al. (1939); Biochem. J. 33, p. 1028
Allen, J.E. (1959); Spectrochim. Acta 10, p. 800
Alford, J.A. et al. (1964); J. Lipid Research 5, p. 390
Alford, J.A. et al. (1967); J. Appl. Bact. 30, p. 488
Altschul, A.M. (1974); New Protein Foods Vol I A; Acad. Press, London.
AOAC Methods of Analysis; Washington D.C. 1965; 10th ed.
Arx, J.A. von (1972); Ant. v. Leeuwenhoek 38, p. 289
Atkins, L. et al. (1943); Ind. Eng. Chem. Anal. 15, p. 141
Auro, J. et al. (1957); Ind. Eng. Chem. 49, p. 1237
Autret, M. (1969); Voeding 30, p. 156
Barton - Wright E.C. (1944); Biochem. J. 38, p. 314
Barton - Wright E.C. (1945); Analyst 70, p. 283
Belcher, C.B. and Brooks K.A. (1963); Anal. Chim. Acta. 29, p. 202
Bender, A.E. and Miller D.S. (1953); Biochem.J. 7, p. 53
Bender, A.E. (1958); Proc. Nutr. Soc. 17, p. XXXIX
Bender, A.E. (1965); Proc. Nuti. Soc. 24, p. 190
Berntrop, J.C. (1902); Zeitschr. Angew. Chemie p. 121
Berry J.A. (1933); J. Bact. 25, p. 433
Blakeslee, L.C. et al. (1973) in World Food Production, Demand and Trade;
Iowa State University Press
Bonnelycke, B.E. et al. (1969); Anal. Biochem. 27, p. 262
Bos, P. and De Bruyn J.C. (1973); Ant. v. Leeuwenhoek 39, p. 99
Bowen, H.J. (1966); Trace Elements in Biochemistry, p. 123; Academic Press
London
Bradshut, R.B. (1965); The Kjeldahl Method for Organic Nitrogen; Academic Press
New York
Britisch Petroleum Co. Ltd; B.P. Patent P.V. No. 958.871 d.d. 30 -12-1963.
B.P. Patent No. U.S. 3560341 - S d.d. 15-08-1968;
B.P. Patent No. B.E. 788902-Q d.d. 15-03-1973.
Bruins, W.J. and Schroder, O.J.Th.M. (1972); Mikrochim. Acta, p. 674
Buchanan, J.R. and Muraoka, T.T. (1964); A.A. Newsletter 24 , Sept. 1964, p. 53
Champagnat, A. (1964); BT. Patent No. 1.385.722 C 12d, d.d. 0 7-12-1964
Chassin, J. (1964); Thesis, Hebrew University, Rehovoth, Israel
Colvin, B.M. (1968); J. Dairy Sci. p. 868
Corman, J. (1957); Appl. Microbiology 5, p. 313
Damaschke, K. (1958); m F.T. Todt, Elektrochemische Sauerstoffmessungen, p. 42,
W. de Gruyter, Beriin
Derse, P.H. (1958), J.Ass.Off.Agric. Chem 41, p 192
Dostalek, M et al. (1972),Proc. IV I.F S Perm. Techn Today, p 497
Edel, W. and Kampelmacher, E H (1969), BuU. W H 0 41, p. 297
Edlbacher, S. and Bauer, Dt (1938), Zeitschr. Physiol. Chem. 254, p 275
Eggum, B 0 (1965), Zeitschnft fur Tierphys. 22, p 32
Eykman,C. (1901),Centr. Bakt.Parasitenk 29, p 841
Elvehjem,C.A.( 1931), J. Biol. Chem. 90, p 111
Engel, C G.J M (1971); Landbouwk. Tijdschrift 83, p. 200
Erdtsieck, B. (1967), Thesis University of Technology, Eindhoven
FAO (1957), Nuti.Stud No. 16, Rome.
FAO (1965), Nuti Mtg.Rep.ser No 37, Rome
FAO (1970), Nutr.Stud.No. 24, Rome
FAO (1973), Nutr.Mtg Rep.ser.No. 52, Rome
Flynn, L M. et al. (1951), Anal. Chemie. 23, p. 180
Frens, A.M (1961), Tijdschnft voor Diergeneeskunde 86, p. 1636
Food Balance Sheets (1966), FAO, Rome
Gaden, E.L. (1962), Biotechn. Bioeng. 4, p. 99
Geyer, R.P. et al. (1948), J.Lab.Clin.Med. 33, p. 153
Giohtti, G. and Cantoni, C J (1966), J.Appl.Bact.29, p 395
GoriU, A.D.L. and Nicholson, J.W.G. (1969); Can.J.Animal Sci. 49, p. 315
Gradova, N.B. (1973), Proc.l 3rd Int. Spec Symp on Yeast, Helsinki, p. 154
Greaves, J.E. et al. (1928), J Bact. 16, p.409
Groot, A.P. de, et al. (1970a), Fd.Cosmet.Toxicol.8, p.267
Groot, A P. de, et al. (1970b), Fd.Cosmet.Toxicol.8, p.499
Groot, A.P. de, et al. (1971), Fd.Cosmet Toxicol.9, p.787
Groot, AJP. de (1973); Proc.Int.Symp.on SCP, Rome, m Press
Gropp, J (1973), Bayerische Landwirtschafthches Jahrbuch 50, p.82
Heijdeman, M.T. in Tlie Biological Efficiency of Protein Production, p. 303;
Cambridge U.P.
Hellemond, K K van (1967). ILOB Report No. 183, Wageningen
Herbert, D et al. (1956), Journ Gen Microbiol. 14, p. 601
Holden, A.V. and Marsden, K. (1969), J.Chromatogr. 44, p 481
Huber, J.T. et al. (1970), J Dairy Sci 53, p.675
Humphrey, A.E. (1968), Proc Western Hemisphere Nutrition Congres 11, p 92
lUPAC, (1966), Standard Methods of the Oils and Fats Section of the lUPAC,
5th Ed and Supplements, Butterworth, London
Johnson, M J (1972), Proc. IV 1 F S. Ferm.Techn.Today, p. 473
Jones, J H and Foster, C. (1942), J Nutr 24, p 245
126
Kamer, J.H. van der and Jansen, N.F. (1949); Anal. Chim.Acta 3, p.397
Kamer, J.H. van der and Ginkel, L. van (1952); Cereal Cehm 29, p. 239
Kharatyan, S.G. et al. (1974); Proc. 4th Int. Symp.on Yeast, Vienna, p.75
Kinson, K. and Belcher, C.B. (1964); Anal.Chim.Acta 30, p.64
Kirchgessner, M. and Roth, F. (1973); Zuchtungskunde 45, p. 208
Kofrani, E. (1967); Die Nahrung 11, p.863
Kurdina, R.M. et al. (1972); Chem.Abstracts 77, sect. 16, No.150557
Leuthardt, F. (1941); Ergeb. Physiol. 44, p.588
Lunven, P. et al. (1973); Br.J.of Nutr. 30, p.l89
MacHargue and Calfee, R.K. (1931); Pl.Physiol.Lancaster 6, p.559
MacKenzie, et al. (1958); Gen.Microbiol. 2, p.l97
Markovetz, H.J. and KaUie R.E. (1964); J. Bact. 87, p.968
Mateles, R.J. (1971); Biotechn.Bioeng.l3, p. 581
MerriU, A.C. and Watt, B.K. (1955); Agr.Handbook, U.S.Dept.of Agric.
Meyrath, J. and Matt, J. (1973); Proc. 3rd Int.Spec.Symp. on Yeast, Helsinki,p.l55
Miller, D.S. and Bender, A.E. (1955); Br.J.Nutr.8, p.382
Miller, E.L. et al. (1965); Br.J.Nutr.l9, p.565
Miller, E.L. et al. (1964); Biotechn.Bioeng.6, p.299
Mitchell, H.H. (1923); J.Biol.Chem.58, p. 873
MitcheU, H.H. and Block, R.J. (1946); J.Biol.Chem.l63, p.599
Moser, A. (1973); 3rd Symp.f.Techn.Mikrobiol.Beriin
MoriU,J.L. (1971);J.Dairy Sci.54,p.l060
Morris, E.O. (1958); in The Chemistry and Biology of Yeast p.25
Academic Press London
Mossel, D.A.A. and Ratto, M.A. (1970); Appl.MicrobioI.20, p.273
Mossel, D.A.A. and Waart, J. de (1968); Ann. Inst.Pasteur LiUe 19, p.13
Munk, V. et al. (1969); Biotechn.Bioeng. 11, p.383
Munk, V. et al. (1969); Folio Microbiol.l4, p.334
NEN Publications of NNI, Polakweg 2, Rijswijk
Nitsan, Z. (1971); J.Dairy Sci.54, p.l294
Noordervliet, (1973); Voedingsmiddelenjaarboek, Ed. Noordervliet, Zeist.
Odintsova, N. (1973); Proc.3rd Int.Spec.Symp.on Yeast, Helsinki, p.34
Olson, B.H. and Johnson, M.J. (1949); J.Bact.57, p.235
Oser, B.L. (1951); J.Bm.Dietet.Ass.27, p.396
Oser, B.L. (1970); in Evaluation of Novel Protein Products, p.267
Pergamon Press. Oxford
PAG Guideline No. 5 (1972); PAG Guideline for Edible Heat-Processed Soy Grits
and Flour
127
PAG Guideline No. 6 (1972); PAG Guideline for Preclinical Testing of Novel
Sources of Protein
PAG Guideline No. 11 (1972); PAG Guideline for the Sanitary Production and Use
of Dry Protein Foods
Roy, J.H.B. (1970); J.Sci.Fd.Agric. p.346
Romer, E.G. et al. (1972); Microchim. Acta p.674
Sachse, R. (1878); Chem. Zeitung 2, p.732
Sarett, H. and Cheldelin, P. (1944); J.Biol.Chem.l55, p.l53
Schattock, P.M.F. (1955); Ann.Inst.Pasteur Lille 7, p.95
Schmidt, G. and Tannhauser, S.J. (1945); J.Biol.Chem.l61, p.83
Schoori, N. (1929); Chem. Weekblad 26, p. 130
Shacklady, CA. (1969); Proc.Nutr.Soc.28, p.91
Shacklady, CA. and Gatumel, E. (1972); in H. Gounelle de Pontanel, Symp.Aix en
Provence 1972; Centre de Recherche Foch, Paris
Sheggs, H.R. and Wright, L.D. (1944); J.Biol.Chem.l56, p.21
SneU, E.E. and Strong, J.M. (1939); Ind.Eng.Chem.Anal.l 1, p.346
Southgate, D.A.T. and Dumin, J.V.G.A. (1970); Br.J.Nutr.24, p.517
Suomalainen, H. and Ouri, E. (1971); The Yeast Vol.2, Academic Press, London
Thomas, K. (1904); Arch.Anat.Physiol.219, Physiol.Abt. p.7
Thomas, K.C and Jones, A.M. (1971); J.Appl.Bacteriol.34, p.717
Vewin, (1974) Statistisch overzicht van de waterleiding in Nederland 1968-1969,
in Press
Vischer, E. and Chargaff, H. (1948); J.Biol.Chem.l76, p.716
Waart, J. de et al. (1968); J.Appl.Bact. 31, p.276
Waart, J. de and Smit, F. (1967); Lab.Pract.l6, p.l098
Wagner, et al. (1969); Biotechn.Bioeng. 11, p.393
Wal, P. van der (1968); World Conf.Amin.Prod.il, Maryland
Walt J.P. van der and Kerken, A.E. van (1961); Anth. van Leeuwenhoek 28, p.81
Wang, DJ.C. (1968); Chem.Eng.p.53
Weerden, E.J. van (1970); ILOB Report No. 254, Wageningen
Weerden, E.J. van (1970); ILOB Report no. 254, Wageningen
WeibuU, M. (1892); Zeitschrift Angew.Chemie, p.450
WeibuU, M. (1894); Zeitschrift Angew.Chemie, p.l99
Wenzeler, R.J. (1947); Cellular Comp.Physiol.l7, p.263
Wetzler, T.F. et al. (1962); Am.J.Publ.Health 52, p.460
Wickerham, L.J. (1946); J.Bact.52, p.293
Wickerham, CJ. and Burton, K.A. (1928); J.Bact.l6, p.363
Wiken, T.O. et al.(I951); Anth. van Leeuwenhoek 17, p.279
128
Wiken, T.O. et al. (1954); Arch. Mikrobiol.20, p.201
Wiken, T.O. (1954); Kalium Symposium p. 123, Int.KaU Inst.Bern
Wiken, T.O. (1972); The Symba Process, Proc.IV I.F.S. Ferm.Techn.Today, p.569
Wilcke, H.L. (1966); in Worid Protein Resources p. 16, Am.Chem. Soc,
Washington D.C.
Wilson, P.N. (1971); Symp. Meat in the Future, ICST London
Worgan, J.T. (1973); in The Biological Efficiency of Protein Production, p. 339;
Cambridge U.P.
SUMMARY
This thesis describes the results of investigations on the production of yeast on fats
and fat products as sole sources of carbon and energy The motivation of this
research is elucidated in Chapter 1, where inter aha the matter of protein calorie
malnutrition is discussed extensively
TheoreticaUy, the amount of protein produced is sufficient to provide every human
being with the minimum quantity of protein necessary to prevent starvation
However this quantity is not now available for approximately two thirds of the
world population
Moreover, there is an ever increasing demand for tasty animal proteins, particularly
meat, due to the growing prosperity in several parts of the world This will almost
certainly result in a world surplus of fats which, because of deterioration, become
inedible
The position of the Netherlands, in terms of world trade in fats, is rather unique
and it may be expected that, especially in the Netherlands, the availibility of
inedible fats and fat products wiU increase considerably
Hence, production of yeast on the very raw materials just mentioned could serve a
twofold purpose The environment wdl not be overloaded with a growing surplus of
inedible fats, and a microbial protein may be produced that could replace skimmed
milk-powder in fodder, thus contributing indirectly to the production of protein
suitable for human consumption
In addition to the world shortage of protein and the availibility of fats. Chapter I
deals with some general aspects of the production of 'single cell protem' (SCP) and
with some specific problems connected with the utilization of fats as substrates for
yeast cultivation
Chapter II describes in detail the analytical methods applied throughout the work
reported in this thesis The majority of these analyses were performed m the various
departments of the Central Institute for Nutrition and Food Research
(CIVO-TNO), Zeist, Netherlands Some of them, particularly a number of studies
concerning the safety evaluation of yeast, are to be performed as soon as sufficient
amounts of yeast grown on fats on a pdot plant scale are avadable On account of
their general importance, however, they are already described m this Chapter
Moreover, they diustrate the claims on the quality of the yeast, which in our view
have to be met before the product is brought on the market as animal fodder
Chapter 111 is divided into two sections The first sets out the various steps in the
process of production of yeast biomass i e cultivation in slants, shake flasks, a 1
liter fermentor and a 20 liter fermentor The second specifies the experimental
conditions A number of experiments had to be performed before we found that
glucose yeastextract agar was the most suitable substrate for the slant cultures
In the shake flask experiments, addition of a definite amount of urea was found
essential in order to maintain the pH value of the culture liquid between 4 and 5
130
The 1 liter fermentor was constructed by using common laboratory glassware of the
Sovirel type. After special arrangements had been made for the inlet tubes, and for
the bearing system, this fermentor proved to be quite efficient and, in comparison
with factory made fermentors, rather cheap. In addition to glassware, various
instruments were used for recording and controling the pH and the temperature, and
analyzing the effluent gas.
The 12 to 15 Hter fermentations were done with a standard Biolafitte fermentor,
combined with a Terzano fermentation vessel and some special instrumentation.
For the harvesting of the yeast, various centrifuges and a freeze drying apparatus
were used.
The second section of Chapter 111 moreover describes a number of experiments
leading to the selection of a strain of Endomycopsis lipolytica Wickerham et al. for
further investigations.
The concentration of the fats varied from 1.8 to 2% m/m. The preparation of the
emulsions occurred in two steps. Firstly, preparation of an emulsion containing 18
or 20% m/m fat, for which a dairy emulsifier type Renny was used, and sterUization
of the emulsion. The amount of emulsifier applied was inversely proportional to the
diameter of the oil droplets of the emulsion, as could be proved by simple
calculations. Secondly, the emulsion obtained was diluted with water, and salt and
vitamin solutions were added. This was performed just before the complete
medium was inoculated. The composition of the salt solutions was based on the
assumption that the various elements in microbial ceUs are, within certain limits,
present in a constant interrelationship. In the case of Mg, Mn and Zn, experimental
corroboration of this hypothesis was found.
For growth of the yeast a temperature of 30 C was chosen. The pH of the culture
Hquid was kept constant at 4.3 by means of ammonia solution addition.
Chapter IV presents the apphcation of the standard procedure, developed in the 1
Hter fermentor. Emulsions of grease, tallow and technical fat were tested in batch
fermentations and, furthermore, batchtobatch and continuous fermentations
were performed with fatty acids obtained from soya oU refineries.
Normally the yield, i.e. the amount of dry yeast obtained from 100 g of fat, was 65
to 70 g but, when oxygen was limiting the growth, the yield dropped to between 57
and 59 g.
After mixing various batches of the yeast produced it was subjected to
organoleptic, chemical, bacteriological, nutritional and toxicological investigations.
The results of these examinations are described in Chapter V.
It was found that, due to this organoleptic properties, the yeast was acceptable for
appUcation in animal fodder.
The fat content was 32% m/m, and the fatty acid profile showed 40% m/m
mono-unsaturated fatty acids, 30% m/m polyunsaturated fatty acids and only
0.5% m/m odd numbered fatty acids.
The protein content was 40% m/m, and the amino acid composition, as compared
vrith the FAO provisional standard, was good. The nucleic acid content was only
131
45% m/m The 'Chemical score' was 70%, because of a low content of
sulphurcontaining amino acids, but could be increased to 86 W( by addition of
0 3% m/m methionine to the yeast Tlie biological values derived from these values
for the Chemical score were 83 and 93 respectively Calculated on the basis of the
EAA-lndex, the biological values were 83 and 85
Experiments with rats showed a biological value of 53 for the yeast and 97 for the
mixture of yeast and 0 3% m/m methionine
The bacteriological properties of an arbitrary yeast sample met the requirements
laid down in the PAG guideline No 5
From a screening test performed for 28 days with rats on a diet which contained
60% m/m of the yeast mixture as the sole source of protem, the conclusion could
be drawn that this yeast had no toxic effects
In Chapter VI, the final one, the statement is made that it is possible to produce
yeast biomass on fats and fat products as sole sources of carbon and energy with
the selected strain and according the procedure developed The properties of the
yeast, grown on a laboratory scale, are undoubtedly promising for application in
animal diets
However more investigations are needed The yeast should be produced on a pdot
plant scale in order to get cell material sufficient for toxicity research and
experiments with live stock
In view of the results obtained so far it seems quite probable that such examination
of the yeast wdl show that it fulfils the requirements of PAG guidelines for
application of novel sources of protein in fodder
Finally in this Chapter the economic feasibUity of the process is discussed The
conclusion is that the application of the process of cultivating yeast on fats and fat
products to make SCP that can replace skimmed milk powder in diets for calves is
determined by the prices of raw materials and those of skimmed milk powder Both
types of prices largely depend on pohtical decisions
132
SAMENVATTING
In dit proefschrift worden onderzoekingen beschreven die zijn gedaan om na te
gaan of vet en vetprodukten kunnen worden gebruikt als enige koolstof en
energiebron voor gistproduktie.
De motivatie voor dit onderwerp wordt toegelicht in hoofdstuk 1, waarin tevens het
wereldvoedselprobleem en speciaal het eiwit- calorie tekort uitvoerig wordt
besproken. Het blijkt, dat de hoeveelheid eiwitvoeding die wordt geproduceerd, in
theorie genoeg is om ieder mens te voorzien van de minimale hoeveelheid die nodig
is om niet van honger te sterven, doch dat in de praktijk ongeveer tweederde van de
wereldbevolking minder krijgt dan deze minimale hoeveelhied. Bovendien is er een
toenemende vraag naar dierlijk eiwit, voornamelijk vices, tengevolge van een
stijgende welvaart. Deze vleesproduktie kost niet alleen een extra grote hoeveelheid
plantaardig eiwit ten gevolge van de lage conversiegraad, maar zal ook een grotere
vetproduktie met zich meebrengen, die kan leiden tot een overschot aan dierlijke
vetten die, als gevolg van zeer snel bederf, niet meer eetbaar zijn, z.g. oneetbaar vet.
De positie van Nederiand is, wat betreft de handel in dit soort vetten, nogal
uitzonderiijk en daarom kan worden verwacht, dat vooral in Nederland de
hoeveelheid oneetbaar dierlijk vet aanzienlijk zal toenemen. In dit licht bezien heeft
de produktie van gist, waarbij vet en vetprodukten als grondstof worden gebruikt,
een tweeledig doel: het mUieu wordt niet belast met vetoverschotten en bovendien
wordt een eiwitrijk produkt verkregen dat indirect kan bijdragen aan de
eiwitbehoefte van de wereldbevolking, doordat de raagere melkpoeder, die thans in
veevoeder wordt gebruikt, gedeeltelijk kan worden vervangen.
Vervolgens wordt in hoofdstuk 1 ingegaan op enkele algemene aspecten van de
produktie van eencellige eiwitrijke microorganismen, de zg. SCP, waartoe o.a.
gisten, bacterien en algen behoren. Hierbij wordt duidelijk, dat een aantal
technologische moeilijkheden nog moet worden overwonnen voor de produktie van
SCP op economische schaal - waarvoor een jaarproduktie van 100.000 ton per
bedrijf wordt aangenomen mogelijk zal zijn.
Tenslotte wordt in dit hoofdstuk ingegaan op de extra problemen die de
fermentatie van vet en vetprodukten met behulp van gisten met zich meebrengen.
In hoofdstuk II worden de analytische methoden beschreven, die bij het onderzoek
zijn toegepast. Bovendien zijn enkele methoden opgenomen, met name op het
gebied van de toxicologische evaluatie, die kunnen worden gebruikt wanneer t.z.t.
grotere hoeveelheden gist, geproduceerd op semi-technische schaal, beschikbaar
zullen zijn. Zij zijn reeds thans opgenomen wegens hun importantie voor
SCPevaluatie en als Ulustratie van de onderzoekingen, die moeten worden verricht,
alvorens SCP als grondstof voor veevoeder in de handel mag worden gebracht.
Hoofdstuk III bestaat uit twee delen; in het eerste zijn de verschUlende fasen
beschreven, waarin het produktieproces van de vetfermentatie kan worden onder-
verdeeld n.l. de groei in de cultuurbuis, de schudkolf, de 1 1 fermentor en de 20 1
133
fermentor In het tweede worden de onderzoekingen besproken, die hebben geleid
tot de keuze van de procescondities
Als medium voor de cultuurbuis bleek glucosegistextractagar het meest geschikt
te zijn Om bij de schudkolfproeven de pH tussen 4 en 5 te houden was het nodig
een bepaalde hoeveelheid ureum aan het cultuurmedium toe te voegen Het bleek
mogelijk met laboratoriumglaswerk van het Sovireltype een fermentor voor een
capaciteit van 1 hter op te bouwen Voor fermentaties van 12-15 liter werd een
Biolafitte fermentor gebruikt Deze fermentors werden gecombineerd met appara-
tuur voor het meten en regelen van de temperatuur, de pH en het zuurstofgehalte in
de lucht die voor de fermentatie was gebruikt Het oogsten, wassen en drogen van
de geproduceerde gist geschiedde met een aantal centrifuges en een vnesdroogappa-
radt van 5 1 capaciteit
Het tweede deel van hoofdstuk III begint met de beschnjvmg van de proeven, die
geleid hebben tot de keuze van de giststam, die in het onderzoek is gebruikt n I een
Endomycopsis Hpolytica Wickerham et al stam Daarna wordt uitgelegd dat het
emulgeren van het vet in twee trappen geschiedt, eerst wordt een emulsie met 18 of
20% m/m vet gemaakt met behulp van een zuivelemulgator, type Rennie, en daarna
wordt deze emulsie gesterdiseerd Aan de hand van rekenvoorbeelden wordt
duidelijk gemaakt, dat de hoeveelheid toe te passen emulgator omgekeerd evenredig
IS met de diameter van de vetdruppels in de emulsie Vlak voor het gebruik wordt
de vetemulsie met water verdund tot een eindconcentratie van 1,8 of 1% m/m
waarbij tevens de benodigde mmeralen en vitamines worden toegevoegd
De samenstelling van het zoutmengsel is gebaseerd op de vooronderstelling dat de
verschUlende elementen in een min of meer constante verhouding in plantencellen
voorkomen Voor de elementen Mg, Mn en Zn wordt deze hypothese ondersteund
door analytische gegevens, verkregen tijdens een coiitinufermentatie
Gefermenteerd wordt bij een pH van 4 3 die constant wordt gehouden door het
toevoegen van ammonia, en bij een temperatuur van 30C
Hoofdstuk IV beschrijft de toepassing van de standaardprocedure, ontwikkeld in de
fermentor van 1 liter In de fermentor van 20 liter werden emulsies van oneetbaar
varkensvet, oneetbaar rundvet en techmsch vet gefermenteerd
Bovendien werden met vetzuren, verkregen bij de raffinage van sojaolie, een batch
proef, een continue fermentatie en twee series van zg batchto batch fermentaties
uitgevoerd Uit de resultaten blijkt dat uit 100 g vet 65 tot 70 gram droge gist kan
worden verkregen, behalve als zuurstofbeperking optreedt In dat geval wordt de
opbrengst 57 to 59 gram per 100 gram vet
De bij de verschiUende fermentaties verkregen hoeveelheden gist werden vermengd
en gebruikt voor organoleptisch, chemisch en bacteriologisch onderzoek, terwijl
bovendien proeven met ratten werden gedaan om de voedingswaarde te bepalen en
eventuele schadelijkheid vast te stellen De resultaten van deze onderzoekingen
staan m hoofdstuk V vermeld
De organoleptische kwaliteit van de gist lijkt zonder meer geschikt voor toepassing
in veevoeder
134
Het vetgehalte was 32% m/m; het vetzuurpatroon gaf aan dat 40% m/m van de
vetzuren een dubbele binding had terwijl 30% m/m uit meervoudig onverzadigde
vetzuren bestond. Het gehalte aan vetzuren met een oneven aantal koolstofatomen
was slechts 0.5% m/m.
Het eiwitgehalte was 40% m/m en het aminozuurpatroon was goed in vergelijking
met het door de FAO voorgestelde standaardpatroon; het gehalte aan nucleinezuren
was slechts 45% m/m. De 'Chemical score' was 70% tengevolge van het lage gehalte
aan zwavelbevattende aminozuren, maar kon worden verhoogd tot 86.6% door
toevoeging van 0,3% methionine. De van deze waarden afgeleide biologische
waarden waren resp. 83 en 93. Gebaseerd op de EAA-lndex waren de biologische
waarden 83 resp. 85. Proeven met ratten gaven een biologische waarde van 53 voor
de gist als zodanig, doch deze werd 97 door toevoeging van 0,3% methionine.
Debacteriologische kwaliteit van het gistmengsel voldeed aan de eisen, die volgens de
richtlijnen van de FAO aan dergelijke produkten moeten worden gesteld.
Bij een onderzoek naar de toxiciteit bleek dat ratten, die gedurende 28 dagen
werden gevoerd met een dieet dat voor 60% m/m uit gist bestond. geen toxische
verschijnselen vertoonden.
In hoofdstuk VI wordt de conclusie getrokken dat het mogelijk is gisten te
produceren, die vetten of vetzuren als enige koolstof- en energiebron gebruiken,
wanneer met de juiste stam wordt gewerkt volgens de in dit proefschrift beschreven
standaardmethode. De eigenschappen van de op laboratoriumschaal geproduceerde
gist zijn veelbelovend voor de toepassing in veevoeder.
Het onderzoek is echter nog niet afgesloten. Op semi-technische schaal moeten nog
grotere hoeveeUieden worden gemaakt waarmee uitgebreider toxicologisch onder-
zoek en proeven met vee kunnen worden gedaan. Gezien de reeds verkregen
resultaten is het goed mogelijk dat zal blijken, dat de geproduceerde gist volledig zal
voldoen aan de door de FAO gestelde eisen.
Tenslotte wordt in dit laatste hoofdstuk de economische toepasbaarheid van het
proces nagegaan. Hierbij blijkt dat de mogelijke toepassing voornamelijk wordt
bepaald door de prijzen van de grondstoffen en de prijs van magere melkpoeder, die
op hun beurt sterk afhankelijk zijn van politieke beslissingen.
135
PeaHDMB
B flHccepTaukiM onHCbi aat oTCH p e a y / i b x a r b i Hc c n s f l o -
BaHHki B n p o H3 BOf l c x BB flpommski Ha mHpax M mn p o -
n p o f l y K x a x , n a n ef l k i Hc x BGHHor o HcxoHHHHa y r n s -
p o f l a H BHBp r wH. r i oBOf l H sxoki p a So x B p a 3 . n a r a -
ex c H B r . naBB I , rfie, nsmf l y n p o HHM, n o f l p o f i n o
o f i c y mf l a s x c H Bo n p o c Hs f l o c x a x o MHo r o nnx ahMH
dS/l HOBbi nM Ha/ l OpHRMH.
TeopSXH HBCHM , HO/IMHBCXBO np O H 3 B GflB H HO TO 6B; i Ha
flocxaxoHHO Mxodbi n p s f l o x a p a x MX b r o n o f l a HHe -
x . H . " y p o BBHb 6 e 3 o n a c Ho c x M" . Qf l naKO, a x o H O -
/IMHBCXBO 3 HaCXORLUBB BpBMH HBf l OCXynHO npHf i xi M"
3HXB/ i bHo fl,nn flayx x p a x BH Hacs/ i sHMFi S B MH OT O
Luapa H3 - 3 a flayx n p m H H . B o - n e p a b i x , /iraflw B 6 O -
r a x b i x c x p a n a x noxpGf i / i Ht ax n a n Ho r o 6o/ i bi ue dB/ r na
HBM HBOdXOflHMblH MHHkl MyM; BG- BXOpbl X, p a c n p B f l B -
JIBHMB HMBtOmBrOCn HD/ l HHBCXBa, BC/l H BOOfilUB Cy -
i UBCXByax c u c x s Ma p a c n p Bf l B/ i BHHH, HecoBepi uBHHo .
Bo/i bLi i B x o r o , BOB pac x GX c n p o c na BKycHbi B W H -
BOXHblB 6 e / l HH, B OCOfi sHHOCXH Ha MHCO, BC/ l Bf l CX-
BktB p a c x y me r o 6 . n a r o c o c x o n HHH B HSHOxopbi x n a c -
XHX 3BMH0r 0 LUapa. 3 X 0 , nOSXH BSpOHXHO, n p HBB-
flBX H MWpOBOMy M3/l Hl UKy WMpa.
ri onOWBHHB Hkl f l Bp/ i aHf l OB , C XOHHH SpSHKFI MKpOBOkl
XQprOB. nH HBCbBfl OdHbl MM mH po np Qf l y HX aMM , flOBO/Ib-
HQ yHHHa. nbHo H , c n s f l y a x o mwf l a x b , HXO OC OS B H H O
B HMf l Bp/ i aHf l aX 3HaHkl Te/ l bH0 Bbi paCXBX Ha.nKHHblM
3 a n a c HBc t ef l odHbi x mwpoB M wHp o n p o f l y HXOB.
136
Hs a x o r o c j i B f l y s x , HXO npoH3BOj : i cxBa flpowwBH Ha
Bbi i xi BynoMRHyxoM c y p b e 6bi / i o dbi B A B O M H S no. nB3Ho.
r i yXBM BXOpHHHOH Of i paSoXHH H3/1HLJJH0B XBXHMHBC-
HHx mnpoB H wHponpof l y HXOB noms x dbi xb n p o n a B B -
flSH MH HpodHa/ l bHbl H 6e/ 10H, HOXOpblH MOT 6bl 3 a MB-
HHXb MO.noHHbiH nopoi xi OH B HopMB fl/iH CHo x a , x a -
HkiM o d p a a o M H O C B B H H O y n a c x B y f l a n p o H3 BOf l c x BB
6B. nHa, r o f l H o r o fl,nf\ nHxaHHH HBn o Be n a .
HpoMB MHp o Bo r o HBf l o c x a x Ha 6B.nHa H Ha; i HHHH W H -
pa H WHponpof l y HXOB, r . naBa I n a c a Bx c f i BLUB H B -
Hoxopbi x acnBHXOB n p o HSBOf l c x a a of l HOHJ i ax oHHor o
6B.nHa ( s i n g l e c e l l p r o t e i n ' S C P ' ) H H B H O X O -
pblX CnB14kl(})HHBCHHX n p o f i / l BM, CBHSaHHblX C HC"
no ; i b30 BaHHBM c b i p b H, ynoMH H y x o r o n a n c y S c x p a x
finr\ x a n o r o n p o H3 BGf l c x Ba .
r . naBa I I flBxa/ibHO onHCbi aas x aHa. nHXHHBCHHe
MBXOflbi y nox ps d/ i BHHbi B a p a f i o x a , onHcaHHOM B
3X0H flMccepxai4HH .
Bcnbl UHHCXBO 3XHX aHa/ l H30B SblJIO Bbi ncnHBHO B
pasHbi x ox f l e. nax l 4 BHx p a . n b Ho r o - Ha y MHo - Mc c . n Bf l o -
a a x B n b K o r o Mn c x Hx y x a r i HxaHHH M flHiUH ( C I V O-
T NO) , 3 B H C X , HHf l ep/ i aHf l bi . HsHoxopbi B H 3 H H X , a
HMBHHO pflfi HCC/ I Bf l OBaHHH HaC amii^H X C fl On p Bf l B/ l B-
HHfi d s s o n a c Ho c x H flpowmBH, n p e f l c x o n x abi noj i HHXb
Han x c nbHO d y f l a x HOKon/ i BHo flocxaxoHHos HOX I H-
HBCXBO flpowwBH, BbipaLUBHHbix Ha mHps B Maci xi xa-
dax 3HcnepHM6Hxa. nbH0H y c x a HOBHM. BsMf l y of l Hano
Hx odu^Br o 3HaHBHHn OHH onnc aHbi yms a a x o n
r / i a BB. HpoMB x o r o OHH noHCHHi ox X B x p o d o Ba n HH
137
8l
- BBdO OU ' H lA lRHaMXHa(t)(t)e OHHLfOaOtf BOLTBEBHO dOX
- Hauidac}) x o x e ' BHHa i i p o o o u o Hd u aRHSi / BHhauo RHBLT
- a t f o ML/Rp awaxoHO HoaoHMHUHmt f ou H wBHgA dx WRH
- tT' oaa H HBH ' OJOX aLfOOU " I Ba TA OS BUHX BHHBBOtf
- A dopOOL/ HBXO OJOHdOXBdopBl / OJOHhRgO EH HBL/atfO
L/Rp doxnaui dat } ) HRaodxHU- OHt f o HHhBxnawdat t ) X RB
odx Hu- - 02 Bi /tf M HHhBXHawdat j ) x Ra o d x Hu - L , BLTIJ' d o g
- Hd u BoxasaROHUO ai r at f EBd ui oadau a o j o x awodj - i
5 M i7 A tfwaui HX
- ooHt f MW MoaodA xsLr A n Hd aHHahSHE HXHaoHBxoA Rp
- o x h wax o ' RHHaahow BBXoahHi i OH o j OHHa i / a t f a d u o
aHHaL/ aepot f WHWHtr' oxpoaH OHHaaxoal xi A o S OOL / B E B H O
RpUOH aHHBBHXBdxoa BH BOXRUO atfOX WOHhRpO Hd|J
dA xRLT A H
XRHHOl /HBH BL/tf WOXBdxOpA o W HhlB tfOXtf OU UIHWBO
l/Rp WOXHBdxOHE UIR B OmWOd tf H HOEOHDILrj O dBJ B
o x h HL/ HmA dBHpo Rw uiBh B^ wadu ' a o x Ha wHd a u o HB
tfBd Hx o a a o d u sooi r r nHdi j nawwodt r ' a a x o f f o BEHo d u
a HHj d a He H Bt f o d a i / j A B H MH K O X O H o j o HHa a x o HHt f a
HBH a o x HA t f o d u o d n m H aodnw a n HBa o Eq i r OUOH OHWOUI
- Eo a oLfRg Rg o x h ' RHa n i / o u Ra RXRp RHmi/otr' a Rd o x
- OH ' BMBOL T OA aRHSi / BXHawndauoHE ax BOXCHBHKOXA
HxoBh Ho d o x a og Ro o Bwo n g Moaawwodt T Ba x o t f o a
- EHo d u a o o a h o d u a R U B X E aRHEBd BOXCHBJ BI / EH X H H
EH Ho a d a u g " HXOBh aat f B H Bnai / at f EBd m B B B I / J
BXOHO
BLftf BwdoH ati'MB B CHL/Boj dox B Hahi A uRB x at f A g
x wA t f odu wah ' a t f wa d u RHa d o a x a i i a o t f A q x Rp RHWi/otr'
' t f BLf JEB mBH BH ' aRdOXOH 'HOWWOdtf A aXOBhBH H
HBHHHD C ({j BpMBHXDpari M npO MbILU.nB H HO TO M C nO/ I H B H H R ,
flOBOXlbHO flBLUBBblM.
HpoMB c x BHf i HHHo r o o d o p y f l o Ba HHR , dbiJiM Mc nor i bBO-
aaHbi pa3/ i HHHbi 6 n p n d o p b i fl/iR HOHxpo/ i R PH H X B M -
n s p a x y p b i , a x aHws fl/in aHa. nH3a Bbi xsHai OLUBro r a -
s a . B c/ i ynar a 2 0 - n Hx p o Bb i x (j ] BpnBHxai 4HH a p a c n o -
pRWBHHH dbi /i H c x aHf l apx Hbi H t t ^spMBHXop B i o l a f f i t e
a H0MdHHaL4HM 0 ({)BpMBHXMpoBOHHbiM c oc y f l OM T e r z a -
n o , a xaHWB HBHox opoB cnei 4Maxi bHoe o d o p y f l o a a H -
HM8. /],.nR c d o p a flpOWmSH npHMBHRJIMCb paBHbIB LJ,BH-
xpMctJyrH H cyd/ i Mri aMMOHHbi M c y i u Hx a / i b .
Bx opoH paaf l BJi r. naBbi I I I xaHwa onnc bi aaBX p n f l o -
n b i x o B, HOXopbi B BBf l yx K Bbi dopy o f l Ho r o l u x a n a
E n d o my c o p s i s l i p o l y t i c a ( Wi c k e r h a m e t a l . ) fl/iH
fla.nbHBHUJMX HCC/l Bfl O aaHMM .
KoHL4BHxpaL4HFi mHpoB Ho n s d a / i a c b B npefienax 1 , 8 %
- 2 % m/ v . DpHr OXO B/IBHMB 3My/ l bCHM npOBOf l M/ i aCb
B flByx a x a n a x : a o - n a p e b i x , n p Hr o x o a / i BHHB 3My . nb-
CHH, cof l BpwaLi i sH 20 % m/ v WHpa n p n noMoixiM n o -
. noHHor o 3My/ i bCHct ) MHaxopa x n n a R e n n i e , H c x s p n -
nH3ai4MM 3X0H a n y n c HM. Bo - a x o p b i x p a Bd a a / i s HMs a -
xoM ar i y. nbCHH aof l OH H flodaaxisHMB c c nnHbi x H B M -
xaMMHHbi x p a c x B o p o B . 3 x 0 dbi.no Mcno/ i HBHO HBn o -
cpBf l cxBBHHO n ep Bf l MHOHy.nFii4HBH r o x o B o r o c y d -
c x p a x a .
JX^H nof l HRXMR flpommeH dbui a Bbi dpana X BMn s p a x y p a
30 . BB/ IHHMHa PH Hy/ l bXypOBOH yHHflHOCXH n o c x o R H -
Ho nof l f l BpwHBa. nac b na anaHBHHM 4 , 3 n y x a n n p n d a -
B/iBHHH aMMMaHa. Co c x a B ccnRHbi x p a c x B o p o B dbm
139
On p ef l B/ I BH, HCXOflR HC npBf l nO/ l OWBHHH , HXO pasHbi B
a/l BMBHXbl BHyxpH MH Hpo d H 3/I bHblX H/IBXOH HaXOfl RXCR
B H3BBCXHbl X npSf l B/ i aX B nOCXOHHHOM B3 a HM0 0 XH0 -
LUBHHH. B c / i y n a e Mg , Mn H Z n , dbi / i o HaHflOHO S H C -
napHMEHXa/ l bHOB nOfl X BBpWflBHMB aXOkl r Hn OX8 3 b l .
B r / i a a e I V npHBOf l HXCR npMMBHBHHB c x a n f l a p x H o r o
MBx o f l a , p a B p a d o x a HHo r o na 1- / i HxpoBOM I J J BPMBHXO-
PB .
3My/ i bCHH c a. na, d a p a H b s r o mnpa H x s x HHHBCHo r o
WHpa dbi /i H Hcnbi xaHbi a n o p u HRx ^ BpMBHxauHH H fla-
/ I BB, nOCXORHHO, OX nopU,HH H nopUMH, (t l BpMBHXa-
14HH dbi/iH Hcno. nHBHbi c wHpHbiMH HHC/ i ox aMH, n o / i y -
HBHbiMH c a a s o f l a paci JHHHpoBaHHn c o B a o r o n a c / i a .
OdblHHO flOXOfl, X . B . HO/IHHBCXBO flpOWWBH, n o / i y -
HBHHbix M3 100 r mn p a , paaHR/ i cR 6 5- 7 0 r , o f l Ha -
Ho , Ho r f l a HHC/ i opof l o r p a HHHHBa n nof l HRXne flpom-
mBH, 3X0X floxofl yMBHbLua/ i cR flo 57 - 59 r .
ri OC/IB CMBIXIBHHH pa3/ l MHHbl X nopUHH npO H 3 B Sf l B H HblX
/ i pommBH, OHH dbi /i H n o f l Be p r n y x b i flBrycxauHOHHbm,
XMnHHBCHHM, d 3 HX B p H 0/10 T M H B C HH M H X 0 HCH HO/lO PH-
HBCHHM Hccr i Bf l o aaHHRM. Pa s y / i b x a x b i B X H X n c c / i B f l o -
aaHHH onnc aHbi a r / i a s s V. Bbi /i o y c x a Ha B. n BH0 , HXO
d / i a r o f l a p a C B OH M flBrycxaLi,MOHHbiM c a o HCx a a M,
flpommn o n a a a / i HCb npHr of l Hbi MH fl/iR Hc no/ i b3oaaHHR
a nni uB mn B OX H U X .
Cof l BpmaHHB mnpa c o c x a B/ i n / i o 32 % M / M , a c n a n x p
WHpHbix HMc/ i ox n o n a s a n 4 0 % M/ M MOHOHBHacbi i nsH-
Hbix mnpHbi x HHc / i o x , 30 % M / M no/ i HHeHacbi LueHHbi x
WHpHblX HHC/lOX H XO/l bHO 0 , 5 % M/ M HBHBXHblX mn p -
140
HblX HHC/ l OX. C o f l s p w a H H B d S/ I HS COCXa B/ l R/ l O 4 0 %
M / M , 3 COCXa a 3 M M H 0 H M C / 1 0 X , n o CpaBHBHHI O c
n p B f l B s p n x B / i b H b i M o d p 3 3 i 4 0 M ct3A D, dhi / i x o p o L UHM. X M -
MMHBc n a R d a / 1 / i o a a R o q a n n a p a s H H / i a c b 7 0 % n a - a a
HHSHOr O C0 f l Bp W3 HHR CB p y COf l B pW3U4MX 3M M HO HH C XlO X ,
HO MOWHO dbl / 10 BB POf l HRXb flO 8 6 , 1 % P p H d 3 B/16 H H-
BM H flpOWWaM 0 , 3 % MB X H0 HHH3 . B M 0/ I O T M H B C Ha H
l 4 BHHOCXb , Bbl CHHXaHHaR H3 BXHX 3 Ha HBHHH fl/lH X H "
MMHBCHOH d a / l / l OBOH OI4BHHM p 3 B H R / i a C b COOX BBX CX -
BBHHO 8 3 H 9 3 . Bbl CHMXaHHbI B Ha OCHOBB HHJ^BHCa
E A A , d HO/ l Or HHBCHHB 3 Ha HBHMR p 3 BHR/ I MCb COOX" .,
B6 X CX BBHH0 8 4 M 8 6 .
Onbl Xbl C Hpbl C3MH n OH3 3 a / l H d MO/ l Or MMBCHy i O OI ^ BHHy
53 fl/IR flpOWWBH d B 3 BCHHHX n p H d 3 B/I B H H M , H 9 5
fl/lH CMBCH flpOWWBH C 0 , 3 % MBXHOHHHOM. B S H X B -
p HO/ l Or MHBCHMB CB 0 HCX B 3 y f l O B/I B X B O p H lOX X p c d O B a -
HHHM, M3/ 10WBHHbl M B p y HOBOf l CX BB PA G Ms B . Ms
COpXMpO BO HHOr O HCn b l X3 HHR, n p OBBf l SHHOF O H3
Hp b l C3 X , HSXOf l HBI XJHXCH B XBHBHHB 2 8 flHBH Ha flH"
B X B , HOXOp a n C0 f l Bp W3 / i a 6 0 % M / M flpOWWSH, H 3 H
Bf l HHCX BBHHOr O HCXOHHMHa d B / l H a , MOWHO 6bl / 10 3 3 -
H/ l HDHHXb, HXO flpOWWH HB OH33bl B3t OX XOHCHHBCHOr O
flBMCXBHH.
B n OC/ l Bf l HBM V I r / i a a S flB/13BXCH Bb l BOf l , HXO MOW-
HO n p o H s a o f l H X b flpowwBayio d n o M a c c y n a w n p a x M
W H p o n p o f l y n x a x , n a n Bj : i MHCXBBHHo r o H c x o H H H H a y -
rnopofia M s n a p r H H , n p HMSHRR MBd p a HHb i n i xi xaM H
B C OOXB B XC XB M H 0 pa 3 pa d o X 3 H HbIM pSWHM OM .
C a o H c x a a flpowwsM, Bbi pai x; BHHbi x a / i a d o p 3 x o p H O M
141
M3 CL i l Xa d B, HBCOMHBHHO n B p CnB HX H B Hbl fl/lH n p MMB "
HBHHH B HOpMOBbi X flHBXaX. Df l Ha HO, HBOdXOf l HMbI
fl3/lbHBHLUHB H C C/ I Bf l O B 3 H H R . HyWHO Bb i p a CXHXb
flpowwH a no/ i npoMbi i i J / i BHHOM Ma c L i i x s d B , Hx o d b i n o -
/ l y HHX b HX a fl0CX3X0HH0M HO/ I HHBCXBB fl/] R X OHCH-
HO/ l Or HHBCHHX H C C/1 Bf l O B 3 H H H H Onbl XOB CO CHOXOM.
BBMf l y flO CMX n o p n O/ i y HBHHb l X p B 3 y / l b X 3 X 0 B ,
a BCb M3 a s p OR X H O, HXO XSHOe HCn HX 3 HH6 n OH3 WBX,
HXO flpcwwH c o o x B B X c x B y i o x y H 3 3 a H H R M PA G n o
npHMBHBHHt O d B/ l HDB MHHp o d HOr O n p O H CXO Wf l BH H H
( S C P ) a H o p M a x .
Ha HOHB M, B 3 X 0 H T/ 13 B B Odc y Wf l at OXCR 3 H0 H0 MMHB C -
HMB B03M0: - KH0CXH . TaHMM Od p a BOM C/ l Bf l y BX Bb l BOf l ,
HXO n p HMBHBHHB npOL4BCCa Hy / l b X H B 314 H H flpOWWBM
Ha w H p a x M w H p o n p o f l y H x a x , B 146/ i RX n p o M B B O f l c x -
B3 S C P , M0 r y i 4 MX 3 3 MBHHX b 0 d B 3W M pB H Hbl H MD/ I OH-
HblH n o p Ol i l OH a HOpMB fl,nf\ X B / I RX , 3 3 BHCMX XO/ l bHO
OX pSUJBHHH n O/ l MXMHBCHOr O n o p R f l H 3 .
Translated by M r. R. B e lz, C IVO-TN O, Zeist
Verified by M r. W. P . va n den B e rcke n , Slavic
In s t i t u t e , Un i ve rs i t y, Ut re ch t .
142
CURRICULUM VITAE
De schrijver van dit proefschrift is geboren op 15 oktober 1922 te Rotterdam.
In 1940 behaalde hij aan de Eerste Gemeentelijke Hogere Burgerschool te
Amsterdam (nu Ir. Lely lyceum) het einddiploma HBS-B en in datzelfde jaar
begon hij chemie te studeren aan de Gemeentelijke Universiteit van Amsterdam. Na
een gedwongen onderbreking tijdens de Tweede Wereldoorlog legde hij in juni 1946
het candidaatsexamen af en in november 1948 deed hij het doctoraalexamen.
Van 1948-1954 was hij werkzaam in de Industrie resp. bij het Medinos Prodent
Research Laboratorium te Amersfoort en bij Shell-Pernis.
Van 19541968 werkte hij bij het onderwijs, en wel van 1954-1955 aan de
afdeling Chemische Techniek van de HTS in Dordrecht, van 1955-1965 aan het
Heymans College in Groningen, terwijl hij van 1965-1968 hoofd was van de schei-
kunde afdeling van het Murray CoHege in Sialkot, Pakistan, daarheen uitgezonden
door de Zending van de Gereformeerde Kerken in Nederland.
Na zijn terugkeer in Nederland trad hij in dienst van de Voedingsorganisatie TNO,
waarvoor hij in 1968/1969 uitgezonden werd naar Saoedi Arable om als 'senior
officer' een laboratorium voor voedingsmiddelencontrole op gang te helpen brengen
in een gemeenschappelijk project van de FAO (Food and Agriculture Organization
of the United Nations) en het Centraal Instituut voor Voedingsonderzoek
(CIVO-TNO).
Sinds 1969 is hij werkzaam op het CIVO-TNO te Zeist, waar hij belast is met de
uitvoering van het toezicht op de raffinage en distribute van oneetbare dierlijke
vetten volgens het Bijzonder Vetbesluit.
Daarnaast was hij van 19691974 werkzaam als docent aan de Avondcursus van de
Amersfoortse Laboratoriumschool.
143
STELLINGEN
1
Het is niet te verwachten dat een horizontaal draaiende buis-fermentor geschikt is
voor produktie van SCP op economisch rendabele schaal.
A. Moser et al. (1973) 3rd Symp. on Techn.
Microbiol., Berlin
2
Ten onrechte vergelijkt Konfrani de 'chemical score' van een eiwit met de
stikstofbalanswaarde van dat eiwit.
E. Konfrani (1968) Die Nahrung 11, p. 863
3
Het voordeel van de methode van RatcHffe en Rodehurst, toegepast bij de oxydatie
van alkoholen, waarbij isolatie van het chroomtrioxydedipyridine complex
overbodig is, komt niet tot uiting in de beschrijving die House van deze methode
geeft.
Modern Synthetic Reactions (1972); p. 265
4
Het moet ernstig worden betwijfeld of de nadelen van een zogenaamde 'multi-
element' lamp voor A.A.S. bepalingen opwegen tegen het prijsvoordeel, wanneer
met deze lampen vier of meer elementen kunnen worden bepaald.
5
Het door sommigen geuite verwijt dat de vleesproduktie geschiedt met een
eiwitrendement van minder dan 10% is in tegenspraak met statistische gegevens.
Dit proefschrift, hoofdstuk I
6
Op grond van eenvoudige berekeningen kan worden aangetoond dat de hoeveelheid
emulgator, nodig voor de bereiding van een stabiele olie in water emulsie,
omgekeerd evenredig is met de gewenste diameter van de oliedruppels
Dit proefschrift, hootdstuk II
7
In geen enkel leerboek voor het VWO wordt gewezen op de mogehjkheid, dat
toevoeging van een reactant aan een gasevenwicht de vorming van diezelfde reactant
tot gevolg kan hebben, zonder dat dit in strijd is met de wet van Le Chateher van 't
Hoff
8
Het IS noodzakelijk bij de opgave van de biologische waarde van een voedingsmiddel
te vermelden hoe deze waarde is bepaald
SG Kharatyan et al (1974) Prot 4th Int
Symp on Yeast I Vienna
9
De correctiefactor, die wordt toegepast bij de berekemng van de hoeveelheid
endogene stikstof, nodig voor de bepaling van de verteerbaarheid van een eiwit, zou
beter betrokken kunnen worden op hoeveelheden droge faeces dan op de
hoeveelheden gegeten voedsel
Dit proetschnft, hoofdstuk V
10
Wanneer bij de bewerking van oHehoudende zaden met alleen rekening gehouden
zou worden met het ohegehalte van deze zaden maar ook met het eiwitgehalte
daarvan, zouden deze zaden een grotere bijdrage kunnen leveren aan de produktie
van eiwit
11
De vorming van een boterberg zou kunnen worden voorkomen door de melkvetten
op te slaan als boterolie en de produktie van roomboter af te stemmen op een
pessimistisch geschatte vraag.
12
Wanneer in India de conceptie van de 'Heilige Koe' zodanig zou veranderen dat het
met langer heiligschennis zou zijn runderen te slachten, /ou de veestapel in India
ongetwijfeld een we/enhjke bijdrage kunnen leveren aan de produktie van eiwit.
13
Zoals een khnisch chemicus onmisbaar is geworden voor het herstel van patienten,
zou een voedingschemicus onmisbaar geacht moeten worden voor het voorkomen
van ziekte ten gevolge van een onjuiste voeding
14
Het werk van bepaalde groepen, wier activiteiten zijn gericht op het welzqn van de
derde wereld, zou een meer zakelijke en minder gevoelsmatige inhoud knjgen mdien
deze groepen kennis zouden willen nemen van de 'State of Food and Agriculture',
jaarhjks uitgegeven door de FAO
15
Het IS verheugend te constateren dat de divergentie in de medische terminologie
ook de medici zelf blijkbaar gaat hmderen.
British Medical J (1974), p 235
16
Bij een democratisch besluit, waarbij geldt 'one man one vote' wordt geen rekening
gehouden met het verschil in kwaliteit der stemmers, hetgeen het gevaar in zich
bergt dat een dergelijk besluit niet op deskundigheid is gebaseerd.
17
Het poneren van stellingen die in geen enkele relatie staan tot de inhoud van het
proefschrift is het enige ludieke aspect ervan; het zou daarom te betreuren zijn als
deze goede gewoonte werd afgeschaft.
J. van der Veen Delft, 20 november 1974.