Vibrational Spectroscopy 63 (2012) 492–498

Contents lists available at SciVerse ScienceDirect

Vibrational Spectroscopy
journal homepage: www.elsevier.com/locate/vibspec

Micro ATR FTIR imaging of hanging drop protein crystallisation

Stefanie E. Glassford a , Lata Govada b , Naomi E. Chayen b , Bernadette Byrne c , Sergei G. Kazarian a,∗
a
Department of Chemical Engineering, Imperial College London, SW7 2AZ, UK
b
Department of Surgery and Cancer, Imperial College London, SW7 2AZ, UK
c
Department of Molecular Bioscience, Imperial College London, SW7 2AZ, UK

a r t i c l e i n f o a b s t r a c t

Article history: Protein crystallisation is of great importance within structural proteomic projects where considerable
Received 13 June 2012 time and effort is spent trying to obtain high resolution structural information from the X-ray diffraction
Received in revised form 28 July 2012 of protein crystals. The crystallisation process is largely empirical due to a lack of understanding of
Accepted 28 July 2012
crystallisation kinetics and it is often very challenging to obtain crystals of sufficient size and quality for
Available online 6 August 2012
X-ray analysis and data collection. This paper presents the use of micro attenuated total reflection (ATR) –
Fourier Transform Infrared (FTIR) spectroscopic imaging as an analytical method for the measurement of
Keywords:
the growth of protein crystals using the hanging drop approach. As well as allowing for the measurement
Total internal reflection
FT-IR spectroscopy
of in situ growth of protein crystals, the high spatial resolution of this method allows the detection of
Protein crystals micro-crystals. The use of this technique could be very advantageous for protein structural studies since
Spatial resolution it allows both detection of the growth of small protein crystals and provides chemical information as the
ATR imaging crystals form.
Spectroscopic imaging © 2012 Elsevier B.V. All rights reserved.

1. Introduction Other novel methods, including the use of porous surfaces and
seeding technologies [4] have been developed along with signif-
The crystallisation of proteins in order to determine their 3D icant progress in nanocrystallography [5] as well as specialised
structure has become an increasingly challenging area of research, techniques such as lipidic cubic phase [6] and bicelle crystallisa-
especially with the continuing development of structural pro- tion [7] suitable for membrane proteins. Despite these exciting
teomics projects. A lot of time and effort is spent generating the developments, hanging drop or sitting drop are still the most com-
target protein and in submitting the protein to both preliminary monly employed methods. Hanging and sitting drop crystallisation
and optimisation crystallisation trials with the aim of gaining a 3D are vapour diffusion methods, where a small drop of protein and
structure from X-ray crystallography, the best available method precipitant solution is equilibrated against a reservoir containing a
for obtaining high resolution structural information [1]. However, higher concentration of precipitant. As the drop equilibrates against
despite many efforts focussing on rationalisation, the crystallisa- the reservoir, the concentration of protein within the drop increases
tion of proteins is still largely an empirical process and involves a to the level required for the nucleation and growth of crystals [8].
significant amount of trial and error [2]. Even when protein crys- As mentioned above, the optimum crystallisation conditions
tals are formed, it is still very difficult to obtain ones that are useful for each protein are identified by extensive screening. The devel-
for X-ray diffraction [2]. Often protein crystals can be too small or opment of new analytical processes to aid in the understanding
slightly imperfect. In addition to this, there is currently no easy way of the crystallisation kinetics by measuring their growth in situ
to distinguish between a protein and salt crystal within the crys- would greatly facilitate the development of rational crystallisation
tallisation drop meaning that time and resources are often wasted approaches. This paper introduces the use of Micro ATR FTIR spec-
optimising conditions, only to discover during final stages that the troscopic imaging as a method that can be used to study the in situ
crystal is in fact not proteinaceous [3]. growth of protein crystals in a hanging drop configuration. The
There are currently several crystallisation methods used in chemical specificity of infrared spectroscopy means that crystals
structural biology laboratories. The traditional approaches employ- can easily be identified as protein and the high spatial resolution
ing vapour diffusion methods such as hanging or sitting drop are the can allow for the detection of micro crystals. The technique could
most widely used, however other methods including microbatch also provide some insight into structural kinetics of crystallisa-
have been successfully used to crystallise a number of proteins. tion through the use of secondary structure analysis of the protein
molecules.
Infrared spectroscopy has already been extensively used for the
∗ Corresponding author. study of protein behaviours such as surface adsorption [9] and
E-mail address: [email protected] (S.G. Kazarian). in solution [10]. Characteristic absorption bands for proteins are

0924-2031/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.vibspec.2012.07.011
 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498 493

sensitive to protein conformation and there are many examples size of the imaging area has been measured to be approximately
in the literature for the assignment of protein secondary structure 64 ␮m × 64 ␮m [16]. Despite this small imaging area, Micro ATR
based on the analysis of Amide I band [11,12,13]. Chemical imaging FTIR spectroscopic imaging has great potential for the studies of
using infrared spectroscopy offers the opportunity to study pro- polymeric materials [17,18,19,20], cross-sections of paintings [21]
tein crystallisation in situ. Macro ATR FTIR spectroscopic imaging and biomedical materials where it has already been employed to
applied to the study of protein crystallisation in a high throughput analyse cross sections of hair [22], breast cancer tissue [16] aorta
manner has previously been shown by Chan et al. [3]. This work and arteries [23] as well as to chemically image live cancer cells
demonstrated the applicability of ATR FTIR spectroscopic imaging [15].
to distinguish between multiple crystallisation conditions simul- The Micro ATR FTIR set up is particularly amenable to the study
taneously using a macro approach, where the spatial resolution is of hanging drop protein crystallisation. The Ge ATR crystal is remov-
limited to 40 ␮m using a zinc selenide ATR accessory or 15–20 ␮m able, as reported previously in reference [15]. Therefore, drops of
with a diamond accessory [3]. Macro ATR FTIR imaging has also protein solution and precipitant solution can be deposited directly
been used for the study of the effect of different surface properties onto the Ge crystal and a reservoir can then be placed underneath
on protein adsorption and crystallisation by preparing a gradient the drop in exactly the same manner as that which is applied
monolayer on the surface of a silicon ATR accessory [14]. However, in structural biology labs. Confocal Raman microscopy has also
the spatial resolution in that study was limited to approximately been applied to the in situ measurement of hanging drop protein
84 ␮m [14]. It would, therefore, be useful to obtain chemical images crystallisation to monitor the change in concentration of protein
of protein crystals with a higher spatial resolution to allow the within the crystallisation drop [24,25]. Micro ATR FTIR imaging is
detection of micro crystals as well as attempt to measure protein an imaging technique and thus requires shorter measurement and
crystal growth in situ. sample preparation times to obtain the same number of spectra
ATR FTIR spectroscopic imaging is an extremely versatile analyt- from the sample in comparison with Confocal Raman microscopy.
ical technique that allows for the measurement of chemical specific Also, despite the small imaging area of Micro ATR FTIR, it is possible
images of a sample, without the need for additional dyes. An FTIR to measure the growth of multiple crystals using this approach as
spectrometer equipped with an accessory that contains a single will be demonstrated below.
bounce internal reflection element (IRE) coupled with an FPA detec- This novel application of Micro ATR FTIR opens up exciting
tor enables spatially resolved chemical information to be obtained opportunities for the dynamic study of protein crystallisation by
from the sample by measuring ∼4000–16,000 spectra simultane- measuring the growth of crystals in situ. The high spatial reso-
ously from different locations within the sample. The nature of the lution of the technique allows for the measurement of the early
ATR approach is such that infrared light only interacts with the stages of crystal growth through the detection of micro crystals and
sample within a few micrometers from the surface of the inter- imaging their growth. This could facilitate the identification of con-
nal reflection element. This limited probing depth means that it is ditions optimal for crystal growth, quickly and easily without need
particularly applicable to the study of aqueous samples but it does to damage the crystal or move the drop. It has potential to be further
require good contact between a particular sample and the surface developed for the study of protein structure in proteins where it is
of the ATR element [15]. very hard to obtain X-ray standard crystals, through FTIR spectro-
Micro ATR FTIR combines a Cassegrain type objective and an scopic analysis of protein secondary structure. This method could
infrared microscope operating in reflection mode, shown in Fig. 1. also be applied to other areas of research where crystallisation is a
The angle of incidence of this type of objective is ca. 30◦ meaning key factor, for example pharmaceuticals.
a Germanium ATR element (n = 4) is used to ensure total inter- This paper shows the micro ATR FTIR spectroscopic imaging of
nal reflection of the IR beam occurs. The high refractive index of three different proteins. Lysozyme and thaumatin are both eas-
the Germanium element combined with the IR microscope objec- ily crystallised and therefore good model proteins with which to
tive means that spatial resolution of approximately 4 ␮m can be develop a new technique. The third protein studied was Lobster ␣-
achieved [16]. This has enabled the study of many samples previ- crustacyanin which has previously been studied to understand the
ously excluded from FTIR imaging studies due to inadequate spatial natural blue colouration of raw lobster shells and is more challeng-
resolution. The field of view with this approach is limited and the ing to crystallise [26].

2. Experimental
IR radiaon
2.1. Materials

Three proteins, lysozyme, thaumatin and ␣-crustacyanin were

studied, each with different crystallisation conditions at 20 ◦ C.
Hen egg-white lysozyme (L 7651), thaumatin from Thaumato-
coccus danielii (T 7638) and all the chemicals used in the reservoir
solutions were obtained from Sigma–Aldrich. ␣-crustacyanin was
provided by Peter Zagalsky of Royal Holloway University of
Protein soluon Ge Crystal London.

2.2. Crystallisation conditions

A solution of 40 mg/ml lysozyme in 0.1 M sodium acetate buffer,

pH 4.5 was prepared. The corresponding reservoir solution con-
Reservoir tained 6% w/v NaCl in 0.1 M sodium acetate buffer, pH 4.5.
Thaumatin solution at 30 mg/ml in 0.8 M sodium potassium tar-
Fig. 1. Schematic of Micro ATR imaging optics with protein crystallisation experi- trate, pH 6.8 was crystallised using a reservoir solution containing
ment. 0.1 M BIS-TRIS Propane (BTP) in the same buffer. ␣-crustacyanin
494 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498

Fig. 2. Micro ATR images of lysozyme crystallisation and spectra extracted from region marked by box. A = 0 min (spectra only), B = 40 min (spectra only), C = 80 min,
D = 110 min, E = 140 min, F = 20 h. The ATR images show distribution of absorbance of the Amide II band at 1540 cm−1 . Spectra A and B show absorption of the crystallisation
solution prior to formation of protein crystals, therefore the main band at ca. 1640 cm−1 is from the bending mode of water.

solution at 5 mg/ml in 28% w/v Polyethylene glycol monomethyl 3. Results and discussion
ether 2000 was equilibrated against a reservoir solution of 0.1 M
BIS-TRIS pH 6.5. 3.1. Lysozyme
All crystallisation drops were deposited on the Ge ATR element
and equilibrated against 200 ␮l of their reservoir solutions. 2 ␮l Lysozyme was used as a model protein for the development of
drops each of lysozyme and thaumatin were set up in a 1:1 pro- the micro ATR-FTIR method since it is readily crystallisable. ATR
tein to reservoir ratio while alpha crustacyanin was set up in a 2:1 FTIR imaging measurements were taken as soon as the trial was
ratio. set up and at subsequent 40 min intervals. Over time the growth of
a lysozyme crystal could be seen in the bottom left corner of the
2.3. Micro ATR FTIR spectroscopic imaging measurements image, shown in Fig. 2. Spectra extracted from the region of the
protein crystal (indicated by box in Fig 2) show that the growth
All ATR-FTIR imaging measurements were carried out on a con- of the protein crystal is marked by the appearance of the charac-
tinuous scan spectrometer (Varian FTS-7000) coupled to an IR teristic Amide I and II bands. The ATR FTIR images were generated
microscope equipped with a liquid nitrogen cooled focal plane by plotting the distribution of absorbance of the Amide II band at
array detector (FPA). A removable Germanium ATR crystal (Refrac- 1540 cm−1 within the imaging area. From the ATR FTIR images, the
tive index, n = 4) was attached to a Cassegrain objective mounted maximum size of the crystal measured can be approximated to
on the IR microscope. The 64 × 64 pixels FPA detector has a mea- 29 ␮m, however it is clear than this is only a part of a crystal as
sured imaging area of 64 ␮m × 64 ␮m in the micro ATR mode. Each the crystal had partly grown outside the field of view.
measurement consists of 250 scans co-added at 8 cm−1 spectral res- Visible images of the measuring surface of the Germanium crys-
olution and has been ratioed against a background spectrum which tal were taken after 20 h, shown in Fig. 3. Since optical images
was measured prior to protein deposition for each new sample. The require the removal of the objective from the IR microscope, only
time taken for each measurement was 7 min. one image was taken during the whole trial in order to minimise
The crystallisation drop was deposited directly onto the mea- disturbance to the crystallisation drop. Moving the Germanium
suring surface of the Germanium ATR element which was then crystal can also affect the ratio with background spectra and hence
mounted onto the IR microscope objective. A small reservoir of should be avoided as much as possible. Water vapour can be an
solution was placed beneath the crystallisation drop and the system issue with IR spectroscopy and removing the objective exposes the
sealed using silicon grease, in a similar manner to commercially spectrometer to the local environment through the gap in equip-
available hanging drop crystallisation plates (Corning Life Sci- ment where the objective would be. This increases the amount of
ences). This set up is shown in Fig. 1. water vapour within the spectrometer which interferes with the
infrared light path, affecting the quality of spectral measurements.
2.4. Optical images Therefore, to minimise the effect of this and allow the system
to return to equilibrium, a time interval of 30 min was observed
Optical images for lysozyme crystals were obtained using the before taking a measurement each time the ATR element was
40× objective on the IR microscope. In the case of thaumatin and removed.
Lobster ␣-crustacyanin, optical images were obtained using a Leica From the visible image, the presence of multiple lysozyme crys-
optical microscope. tals can be seen, with an approximate size of 50 ␮m, indicating
 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498 495

ATR image). Opportunities exist in which this could be determined

in order to overcome this limitation. If the exact location of the ATR
imaging area was known then it may be possible to ensure crys-
tal growth in this area through the use of surface modifications or
seeding technologies, for example.

3.2. Thaumatin

Thaumatin is another protein that is well characterised and crys-

tallises readily, although over a longer time than lysozyme. ATR
FTIR imaging measurements were taken at time = 0, 24 and 28 h
(Fig. 4) where the ATR images were generated by plotting the dis-
tribution of the Amide II band at 1525 cm−1 within the imaged area.
Spectra extracted from the highlighted region are also shown and
a visible image thaumatin crystals on the germanium surface. After
24 h several small crystals have grown within the ATR imaging area,
varying between 6 and 15 ␮m in size. Then at 28 h, the size of these
crystals has reduced to a maximum of 6 ␮m. This would suggest
Fig. 3. Visible image of lysozyme crystals on surface of Ge ATR element. Dashed box that as the crystals increase in size, the effect of gravity means that
shows approximate size of imaging area. the area of crystal in contact with the Ge surface decreases. Again
an optical image was taken at the latest stage of the trial to min-
that half of one of these crystals was captured using micro ATR imise disturbance to the sample. The image shows many protein
FTIR imaging. This highlights the main limitation of this method; crystals across the Ge surface; however from the image it is hard
the limited imaging area. Therefore, for proteins that grow rapidly to distinguish individual crystals.
to form big crystals, it may not always be possible to capture their Measurements of thaumatin were repeated and ATR FTIR
growth using micro ATR FTIR imaging. There is no guarantee that a images are shown in Fig. 5 along with spectra extracted from the
nucleation site will occur within the imaging area and hence there highlighted region. The spectra appear to be noisier than the pre-
is still a degree of trial and error within this approach. The exact vious measurements due to the presence of a larger amount of
location of the imaging area on the germanium crystal is also not water vapour in the system. As before, after 24 h several protein
known precisely, the black dashed box in the visible image shows crystals can be observed in the ATR FTIR image with a maximum
an estimated position (from the shape and size of the crystal on size of approximately 15 ␮m, then after 48 h the number and size

Fig. 4. (i) Micro ATR images of thaumatin crystals. (ii) Extracted spectra from region marked by box on the image: A = 0 h (spectra only), B = 24 h (crystal size 6–15 ␮m),
C = 28 h (crystal size < 6 ␮m). The ATR images show distribution of absorbance of the Amide II band at 1525 cm−1 . Spectrum A shows absorption prior to formation of protein
crystals. (iii) Visible image of thaumatin crystals on surface of Ge ATR element.
496 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498

Fig. 5. (i) Micro ATR images of thaumatin crystals. (ii) Extracted spectra from area marked by box on image: A = 24 h (crystal size ∼ 15 ␮m), B = 48 h (crystal size ∼ 10 ␮m)
The ATR images show distribution of absorbance of the Amide II band at ca. 1525 cm−1 . (iii) Visible image of thaumatin crystals on surface of Ge ATR element.

Fig. 6. (i) Micro ATR images of Lobster alpha crustacyanin crystallisation. (ii) Extracted spectra from boxed area. A = 0 h (spectra only), B = 48 h (Crystal size ∼ 6 ␮m), C = 96 h
(Crystal size ∼ 15 ␮m) The ATR images show distribution of absorbance of the Amide II band at ca. 1540 cm−1 . Spectrum A shows absorption prior to formation of protein
crystals. (iii) Visible image of Lobster alpha crustacyanin crystals on surface of Ge ATR element.
 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498 497

Fig. 7. (i) Micro ATR images of Lobster alpha crustacyanin crystallisation. (ii) Visible image of Lobster alpha crustacyanin crystals on surface of Ge ATR element. (iii) Extracted
spectra from boxed area. A = 1 h, B = 21 h, C = 48 h. The ATR images show distribution of absorbance of the Amide II band at ca. 1540 cm−1 . Spectrum A shows absorption prior
to formation of protein crystals.

of thaumatin crystals measured has reduced; the biggest one was of the protein crystals on the Ge surface. These correlate with the
approximately 6 ␮m. This confirms that as time increases and the previous trial; by 48 h a large amount of protein had aggregated on
crystals become sufficiently large, they are too heavy and less the surface with several micro crystals present.
surface area is in contact with the ATR element. A visible image The measurement of the crystallisation of ␣-crustacyanin by this
(Fig. 5) of the germanium surface at 48 h shows thaumatin crystals approach demonstrates the applicability to proteins that are more
of approximately 20 ␮m. These measurements of thaumatin crys- challenging to crystallise and highlights the ability to detect the
tallisation show that micro ATR can provide information during the formation of micro crystals. Previously in macro ATR imaging stud-
initial growth of the crystals meaning that it is a technique suitable ies, the spatial resolution was such that protein crystals had to be
for the early stages of crystallisation. Further analysis of the pro- much more significant in size before they could be detected. The
tein spectra obtained during this early, critical stage in the process results presented here demonstrate the micro ATR-FTIR technique
could yield information about crystallisation kinetics. Additionally, allows the study of proteins that do not form crystals large enough
this analysis has the advantage of differentiation between salt and or of high enough quality for X-ray diffraction studies as well as
protein crystals than is possible using other techniques. detection of the crystals in the early stages of growth.

3.3. Lobster alpha crustacyanin 4. Conclusions

To determine if Micro ATR imaging could be applied to more This paper demonstrates the novel application of micro ATR
challenging proteins, the hanging drop crystallisation of lobster FTIR spectroscopic imaging to the study of hanging drop protein
␣-crustacyanin was studied. Crystallisation drops of lobster ␣- crystallisation. The deposition of a protein solution directly onto
crustacyanin normally result in the formation of a large number a Germanium ATR objective means that the growth of protein
of small blue crystals which in some cases may be too small for use crystals can be measured in situ without the need for additional
in X-ray studies. measures to identify them as actual protein crystals, during the
␣-Crustacyanin is known to take approximately one day to initial stages of crystal growth. The measurement of multiple pro-
crystallise under standard laboratory conditions. ATR FTIR imag- tein crystals simultaneously has also been shown to be possible.
ing measurements were taken at 0, 48 h and 96 h, with ATR The high spatial resolution of Micro ATR FTIR images has allowed
images, generated from plotting the distribution of Amide II band the detection of protein crystals approximately 6 ␮m in size. The
at 1540 cm−1 , shown in Fig. 6 along with spectra extracted from high spatial resolution measurement of crystal growth in situ could
the highlighted area. After 48 h, a large amount of the protein has offer exciting opportunities for the study of target proteins, espe-
aggregated onto the surface of the Ge element although several cially those which do not produce crystals of the size required
micro-crystals can also be detected, shown by the small regions for X-ray crystallography. It could also have potential application
of higher concentration of protein. A further measurement at 96 h within other areas of research where the understanding of crys-
shows several large regions of high concentration of proteins. From tallisation is important such as drug design and development in
the visible image (Fig. 6) it can be seen that the crystals did not grow pharmaceuticals.
bigger than 10 ␮m, therefore these regions are most likely several
crystals that have agglomerated together (largest region from ATR Acknowledgements
images ∼15 ␮m). It should be noted that the optical image obtained
shows a larger area of the Ge element surface than the imaging area S.E.G. is funded by a BBSRC Targeted Priority Studentship in the
using ATR FTIR imaging therefore a direct comparison of crystals is area of Bioprocessing. S.G.K. acknowledges the research funding
not possible. Measurements were repeated for a new crystallisa- from the European Research Council under the European Com-
tion drop, at time 1, 21 h and 48 h. ATR images and spectra from the munity’s Seventh Framework Programme (FP7/2007-2013)/ERC
highlighted region are shown in Fig. 7 along with an optical image advanced grant agreement no. [227950]. N.E. C. acknowledges
498 S.E. Glassford et al. / Vibrational Spectroscopy 63 (2012) 492–498

funding from the UK Engineering and Physical Sciences Research [14] S. Glassford, K.L.A. Chan, B. Byrne, S.G. Kazarian, Langmuir 28 (2012)
Council EP/G027005/1. We thank Peter Zagalsky of Royal Holloway 3174–3179.
[15] M.K. Kuimova, K.L.A. Chan, S.G. Kazarian, Appl. Spectrosc. 63 (2009) 164–171.
University of London for supplying Alpha crustacyanin and Dr. K. L. [16] S.G. Kazarian, K.L.A. Chan, Appl. Spectrosc. 64 (2010) 135A–152A.
Andrew Chan for his help. [17] A. Gupper, P. Wilhelm, M. Schmied, S.G. Kazarian, K.L.A. Chan, J. Reussner, Appl.
Spectrosc. 56 (2002) 1515–1523.
[18] A. Gupper, P. Wilhelm, G. Kothleitner, K.J. Eichhorn, G. Pompe, Macromol. Symp.
References 205 (2004) 171–180.
[19] D.J. Nagle, G.A. George, L. Rintoul, P.M. Fredericks, Vib. Spectrosc. 53 (2010)
[1] A. McPherson, Methods 34 (2004) 254–265. 24–27.
[2] N.E. Chayen, Curr. Opin. Struct. Biol. 14 (2004) 577–583. [20] G.C. Eder, L. Spoljaric-Lukacic, B.S. Chernev, Anal. Bioanal. Chem. 403 (2012)
[3] K.L.A. Chan, L. Govada, R.M. Bill, N.E. Chayen, S.G. Kazarian, Anal. Chem. 81 683–695.
(2009) 3769–3775. [21] M. Spring, C. Ricci, D.A. Peggie, S.G. Kazarian, Anal. Bioanal. Chem. 392 (2008)
[4] V. Bolanos Garcia, N. Chayen, Prog. Biophys. Mol. Biol. 101 (2009) 3–12. 37–45.
[5] E.R. Bodenstaff, F.J. Hoedemaeker, M.E. Kuil, H.P.M. de Vrind, J.P. Abrahams, [22] K.L.A. Chan, S.G. Kazarian, A.A. Mavraki, D.R. Williams, Appl. Spectrosc. 59
Acta Crystallogr. D: Biol. Crystallogr. 58 (2002) 1901–1906. (2005) 149–155.
[6] V. Cherezov, Curr. Opin. Struct. Biol. 21 (2011) 559–566. [23] F. Palombo, S.G. Cremers, P.D. Weinberg, S.G. Kazarian, J. Roy. Soc. Interface 6
[7] S. Faham, J.U. Bowie, J. Mol. Biol. 316 (2002) 1–6. (2009) 669–680.
[8] N.E. Chayen, Acta Crystallogr. D: Biol. Crystallogr. 54 (1998) 8–15. [24] K. Noda, H. Sato, S. Watanabe, S. Yokoyama, H. Tashiro, Appl. Spectrosc. 61
[9] A. Sethuraman, G. Belfort, Biophys. J. 88 (2005) 1322–1333. (2007) 11–18.
[10] K.A. Oberg, A.L. Fink, Anal. Biochem. 256 (1998) 92–106. [25] J. Vongsvivut, S. Watanabe, K. Noda, H. Sato, H. Tashiro, Scienceasia 34 (2008)
[11] W.K. Surewicz, H.H. Mantsch, D. Chapman, Biochemistry 32 (1993) 400–408.
389–394. [26] N.E. Chayen, M. Cianci, J.G. Grossmann, J. Habash, J.R. Helliwell, G.A. Nneji, J.
[12] P.I. Haris, F. Severcan, J. Mol. Catal. B Enzym. 7 (1999) 207–221. Raftery, P.J. Rizkallah, P.F. Zagalsky, Acta Crystallogr. D: Biol. Crystallogr. 59
[13] A. Barth, C. Zscherp, Q. Rev. Biophys. 35 (2002) 369–430. (2003) 2072–2082.