J Mol Hist (2007) 38:177–182

DOI 10.1007/s10735-007-9085-6

ORIGINAL PAPER

Fiber type and myosin heavy chain compositions of adult pretarsal

orbicularis oculi muscle
Nai-Chen Cheng Æ Shu-Lang Liao Æ I-Jong Wang Æ
I-Chan Lin Æ Yueh-Bih Tang

Received: 3 January 2007 / Accepted: 16 February 2007 / Published online: 17 March 2007

Springer Science+Business Media B.V. 2007

Summary Pretarsal orbicularis oculi muscle (POOM) is MyHC IIA content showed a negative correlation with age.
an important structure of eyelid movement in human. The Comparing with previous studies of limb or masticatory
aim of this study was to investigate fiber histomorphology muscles, adult POOM exhibits a unique fiber and MyHC
and myosin heavy chain (MyHC) isoform composition of composition, as well as a different aging pattern.
adult POOM, and to clarify their age-related changes.
Eyelid specimens from 58 subjects (age range, 21 to 91
years) were collected during upper blepharoplasty proce-
dures. Serial cross sections of POOM were ATPase-stained
and examined under miscroscope. Quantitative measures of Introduction
muscle fiber size and fiber type distribution were obtained
in 35 subjects with adequate fiber cross sections. Relative The orbicularis oculi muscle is a complex structure, and
MyHC isoform contents of POOM were retrieved by gel plays an important role in the periorbital anatomy. It is
electrophoresis in all 58 subjects. Examination of the his- divided into three regions based on their locations in the
tochemical staining revealed an abundance of type II fiber eyelid and on the face (Fig. 1). The first region is the
( >85%) in human POOM, with more type IIX than IIA orbital portion, extending from the orbital rim up to the
fibers. Decreased mean area of all fibers and type IIA fibers eyebrow and down to the cheek, which generates forceful
were noted in the old group when compared to the young. closure of the eyelids. The preseptal region extends from
As for MyHC analysis, the relative content of MyHC iso- the upper end of tarsal plate to the orbital rim. The pretarsal
forms exhibited an order of IIX > IIA > I, and the relative region (POOM) extends from the lid margin to the upper
end of the tarsal plate in the eyelid. The pretarsal and
preseptal regions are much smaller than the orbital portion,
N.-C. Cheng
Y.-B. Tang (&) but are functionally important for our daily blinking
Department of Surgery, National Taiwan University Hospital movement (Lander et al. 1996). Human POOM is com-
and College of Medicine, 7 Chung-Shan South Road, Taipei 100, posed with relatively short, overlapping fibers, which are
Taiwan
heterogeneous in length (Freilinger et al. 1990; Good-
e-mail: [email protected]
murphy and Ovalle 1999; Happak et al. 1988; Lander et al.
S.-L. Liao
I-JongWang 1996). The muscle have the smallest diameter of any
Department of Ophthalmology, National Taiwan University skeletal muscle, including other mimetic muscles (Good-
Hospital and College of Medicine, Taipei, Taiwan
murphy and Ovalle 1999, Happak et al. 1988).
I-Chan Lin Muscle fibers are composed of functional units called
Department of Ophthalmology, Keelung Municipal Hospital, sarcomeres. Thick filament (myosin) and thin filament
Keelung, Taiwan (actin) exist within each sarcomere, and the interaction of
these two myofibrillar proteins cause the muscle to con-
N.-C. Cheng
Department of Surgery, National Taiwan University Hospital, tract. The myosin molecule is composed of six polypep-
Yun-Lin Branch, Doliou, Yunlin, Taiwan tides: two heavy chains and four light chains. Myosin

123
178 J Mol Hist (2007) 38:177–182

Not only responsible for the blink movement, POOM is

also an essential structure when performing cosmetic and
reconstructive procedures of the eyelids. Nevertheless, the
detailed histomorphology and protein composition of adult
POOM were not described in the literature. In addition,
several studies showed that different skeletal muscles in
human behave differently upon aging (Kirkeby and Gar-
barsch 2000; Monemi et al. 1998), so the aging process of
POOM, which cannot be inferred from other muscles, also
merits further study. Therefore, we aimed to investigate the
histomorphometry and MyHC composition of adult POOM
and elucidate their age-related changes.

Materials and methods

Biopsy specimens of POOM from 58 adults’ eyelids were

Fig. 1 The three anatomical portions of human orbicularis oculi studied. There were 17 males and 41 females, ranging in
muscle age from 21 to 91 (mean 58.7 ± 16.6). The muscle samples
were obtained from the POOM excised during surgery for
cosmetic blepharoplasty or dermatochalasis correction.
heavy chain (MyHC) isoforms are the key determinants of None of these patients had any known neuromuscular
the maximum shortening velocity in muscle cells, and disease or obvious POOM abnormality at the time of sur-
represent the functional properties of skeletal muscle fibers. gery. All muscle specimens were collected from the central
MyHC in the head region also contains an adenosine tri- portion of the POOM, and bilateral specimens were ob-
phosphate (ATP) binding site and serves as the enzyme tained from 30 subjects. The investigations adhered to the
adenosine triphosphatase for ATP hydrolysis, which pro- tenets of the Declaration of Helsinki and were approved by
vides the energy necessary for muscle contraction. The the institutional human experimentation committee. In-
enzyme, myofibrillar ATPase, is widely applied in histo- formed consent was obtained from each subject. Local
chemical staining for classification of skeletal muscle fiber. anesthetic (2% lidocaine hydrochloride with 1: 1000 epi-
Fibers are separated based on staining intensities because nephrine) was injected in the area of muscle biopsy before
of different pH sensitivity of ATPase, and it was found that surgery in all patients. The muscle specimens were ob-
different ATPase-based fibers contain different MyHC tained within 15 min of the anesthetic injection. After
isoforms. The three MyHC isoforms originally identified isolation of the muscle from the eyelid specimen, each
were MyHC I, IIA, and IIB, and they were thought to muscle sample was grossly inspected, oriented, and stret-
correspond to fiber I, IIA, and IIB based on the traditional ched to resting length to avoid contraction artifact. Muscle
ATPase staining. However, a fourth MyHC isoform, specimens were mounted, rapidly frozen in propane chilled
MyHC IIX, was identified in small mammals, and it was with liquid nitrogen and stored at –80C until further
found later that what had been called MyHC IIB in humans analysis.
was actually MyHC IIX. Humans do not express the fastest The frozen biopsy specimens were thawed to –20C and
MyHC isoform, MyHC IIB. Thereafter, there has been a serial transverse sections were cut in a cryostat (Bright
controversy about the nomenclature of ATPase-based fiber OTF Cryostat, Cape town, South Africa). Myofibrillar
typing. In this article, we follow the new classification, and ATPase-based histochemistry was performed using prein-
refer fibers with MyHC IIX isoform as type IIX to avoid cubation pH values of 4.3, 4.6 and 10.3 to determine the
confusion. Hence, the three MyHC isoforms identified in muscle fiber type composition (Brooke and Kaiser 1970).
human limb muscles are MyHC I, IIA, and IIX, and they Sections were examined under a Nikon Microphot-FXA
correspond to fiber I, IIA, and IIX based on ATPase microscope (Garden city, New York), and suitable areas
staining (Scott et al. 2001). Type I fibers are functionally with transversely cut muscle fibers were photographed. The
slow-twitch fibers, capable of sustained continuous activ- printouts of the ATPase-stained sections were simulta-
ity. Type IIA fibers are fast-twitch fibers, but are also neously inspected for identification of the fiber types. The
capable of sustained activity. On the other hand, type IIX three fiber types (I, IIA, and IIX) were distinguished on
fibers, which are fast-twitch, phasic and anaerobic, fatigue their basis of staining intensities. The proportion of each
easily (Scott et al. 2001). fiber type was determined by counting the number of type

123
J Mol Hist (2007) 38:177–182 179

I, IIA and IIX fibers in a computerized printout of the refers to those older than 70. Wilcoxon ranksum test was
section with pH 4.6 preincubation. We selected muscle used to analyze the differences of the histomorphometric
fibers with a long-to-short axis ratio of not more than 1.5, parameters between groups, and Wilcoxon matched-pairs
and sections with inappropriate staining or less than 30 signed-ranks test was employed for intra-group compari-
transversely cut fibers were discarded from further analy- sons. The level of significance was set at P < 0.05. On the
sis. Histomorphometric parameters were obtained in 35 other hand, since more muscle samples were available for
subjects (8 males and 27 females) with a mean age of MyHC analysis, the relationship between age and relative
58.5 ± 15.2 (range: 21 to 82). The average number of fiber MyHC contents of POOM was analyzed by linear regres-
analyzed was 56 ± 25 (mean ± standard deviation). sion. Some reports have suggested gender differences in
According to a previous report, there is little difference in fiber type compositions in limb muscles (Essen-Gustavsson
mean fiber area once 30 or more fibers are measured and Borges 1986; Staron et al. 2000), so gender was re-
(McCall et al. 1998). garded as a confounding factor in the linear regression
Appropriate regions within the pH 4.6-preincubated model. The level of significance was also set at P < 0.05.
section were chosen and imaged at 20g magnification. The
cross section area of each muscle fiber was determined by
tracing the periphery of the fiber using an image analysis Results
program (Image-Pro Plus, Media Cybernetics, Silver
Spring, Maryland). Only fibers with clear borders and no The POOM muscle fibers were loosely arranged in parallel
tendency toward longitudinal cut were measured. Mean elongated fascicles of variable thickness, which were
fiber area was calculated as (% type I fiber · type I widely separated by connective tissue. The fibers were
area + % type IIA fiber · type IIA area + % type IIX significantly smaller than those in limb muscles. The
fiber · type IIX area) · 1/100. ATPase staining in the biopsy specimens showed good
MyHC analysis was performed on all biopsy samples contrast among different fiber types. Serial sections dem-
using sodium dodecyl sulfate-polyacrylamide gel electro- onstrated the change of ATPase staining in the same fibers
phoresis (SDS–PAGE). The protocol of electrophoresis with preincubation at pH 4.3, 4.6 and 10.3. Type I fibers
was based on the procedures of Talmadge and Roy (1993). stained darkly at pH 4.3 but lightly at pH 10.3. Type II fiber
Briefly, five to six serial cross sections (20 lm thick) from showed an inverse pattern, and fiber IIA and IIX could be
each biopsy were placed in 0.5 ml of a lysing buffer con- readily differentiated at pH 4.6 (Fig. 2). In all examined
taining 1.0% 2-mercaptoethanol, 4.0% SDS, 16.0% 1.0 M samples, type II fibers were predominant and type I fibers
Tris, 20% glycerol, 0.2% bromophenol blue (pH 6.8) were smaller in size and number. The checkerboard
and were heated for 2 min at 100C. The extracts were staining pattern due to mixed distribution of equivalent
cooled and centrifuged at 12,000g for 1 min. Small number of type I and II fibers in limb muscles could not be
amounts of the supernatants were loaded on 8% separating observed in human POOM.
gels, run overnight (20–21 h) at 150 V. The gels were In the 35 subjects enrolled for histomorphometric
dried and silver stained. MyHC isoforms were identified analysis, the overall percentage by fiber number was 12.5%
according to their molecular masses compared with those for type I, 37.1% for type IIA, and 50.4% for type IIX. The
of marker proteins and their migration patterns. Relative ratio of IIA or IIX fiber was significantly higher than that of
MyHC isoform content was subsequently determined by type I fiber, and the ratio of IIX fiber was higher than that
densitometry, assisted by the same software used for of IIA fiber (P < 0.001, Wilcoxon matched-pairs signed-
histomorphometric analysis (Image-Pro Plus). ranks test). The mean cross section area of each fiber was
For each subject, histomorphometric parameters were 531 ± 89 lm2 for type I, 745 ± 83 lm2 for type IIA, and
derived by analyzing the histochemically stained muscle 712 ± 63 lm2 for type IIX. The average area of all fibers
cross sections. Because no significant difference was was 702 ± 50 lm2. IIA or IIX fiber was significantly larger
observed between the right and left eyelid when samples than type I fiber (P < 0.0001, Wilcoxon matched-pairs
from both sides were available, the data were pooled for signed-ranks test).
further analysis. It has been found that total muscle mass We classified 11 patients who were less than 50 years into
peaks at about the age of 24 years. A modest 10% decrease the young group, and 11 patients who were more than
in whole muscle size occurs between 24 years and 50 years 70 years into the old group. The sex distributions in these two
of age, while an additional diminution of 30% occurs be- groups were not statistically different (Fisher’s exact test,
tween 50 years and 80 years of age (Lexell et al. 1988). P = 0.659). The fiber type distributions between the two
Therefore, two age groups (young and old) were selected groups showed no statistical difference. The old group had a
among the subjects for further comparison. The young significantly smaller type IIA fiber area (710 ± 47 lm2)
group refers to those younger than 50, and the old group comparing to the young group (800 ± 96 lm2 ) (11.3%

123
180 J Mol Hist (2007) 38:177–182

Table 1 Histomorphometric measurements of human POOM in the

young group (less than 50 years) and the old group (more than
70 years)
Young (n = 11) Old (n = 11) P value

Type I fiber (%) 11.3 ± 6.4 11.8 ± 5.8 0.82

Type IIA fiber (%) 34.6 ± 6.2 39.9 ± 10.3 0.18
Type IIX fiber (%) 54.1 ± 10.6 48.3 ± 7.6 0.082
Type I fiber area (lm2) 528 ± 111 514 ± 98 0.57
Type IIA fiber area (lm2) 800 ± 96 710 ± 47 0.0078*
2
Type IIX fiber area (lm ) 724 ± 34 677 ± 76 0.12
Mean fiber area (lm2) 726 ± 47 674 ± 51 0.0197*
The type IIA fiber area and mean fiber area of all fibers were sig-
nificantly larger in the young group. Data denote mean ± standard
deviation
*
P < 0.05

The relative content of the three MyHC isoforms

(MyHC I, MyHC IIA, and MyHC IIX) were obtained from
all 58 subjects (Fig. 3). The overall relative content of
MyHC I was 21.8 ± 4.7%, 37.4 ± 5.5% for MyHC IIA,
and 40.8 ± 5.7% for MyHC IIX. The relative contents of
MyHC IIA or IIX were significantly higher than that of
MyHC I (P < 0.0001, Wilcoxon matched-pairs signed-
ranks test). With gender controlled as a confounding factor,
and linear regression plots showed a negative correlation of
relative MyHC IIA content with age (r = –0.30, P = 0.01).
As for MyHC I and MyHC IIX, no significant correlation
was reported (Fig. 4).

Discussion

In humans, skeletal muscle fibers can be classified based on

their histochemical characteristics. Among the available
staining methods, ATPase histochemistry is most com-
monly used for fiber classification. Because the histo-
Fig. 2 Serial sections of histochemical staining of adult POOM for chemically defined muscle fibers differ metabolically and
ATPase with preincubation in (A) pH 4.3; (B) pH 4.6; and (C) pH functionally, the difference of fiber type composition from
10.3 solutions. Note the scantiness of stained type I fiber in plane A, muscle to muscle reflects the functional demands placed on
and the abundance of the darkly-stained type II fiber in plane B and C.
I, type I muscle fiber stained darkly after acid preincubation and the individual muscle. Based on a contraction rate of 12–13
lightly after alkaline preincubation. IIA, type IIA fiber is lightly times per minute, the human POOM contracts over 750
stained after pH 4.6 preincubation, which can be distinguished from times in every waking hour (Goodmurphy and Ovalle
the darkly-stained type IIX fiber (IIX). Bar = 50 lm 1999). In adults, type I fiber accounts for only 11 to 16% of
all fibers in POOM. A predominance of type II fiber makes
decrement, P = 0.0078, Wilcoxon ranksum test), whereas POOM well suited to high speeds of contraction, though
type I and IIX area were not statistically different. The mean poorly suited to sustained contractions (Freilinger et al.
area of all fibers was 674 ± 51 lm2 in the old group, which 1990; Goodmurphy and Ovalle 1999; Happak et al. 1988).
decreased 7.2% comparing to that of 726 ± 47 lm2 in the However, the characteristics of fiber IIA and IIX in human
young group (P = 0.0197, Wilcoxon ranksum test). The POOM were not described in these studies. Nelson and
hierarchy of the percentage fiber type area in the young or old Blaivas (1991) examined muscle samples from 20 children,
group was both IIX > IIA > I from the largest to the smallest ranging from 9 months to 16 years. The mean cross section
(Table 1). area of each fiber was found to be 158.4 lm2 for type I,

123
J Mol Hist (2007) 38:177–182 181

Fig. 3 MyHC isoforms,

separated by SDS–PAGE in
human POOM samples

270.8 lm2 for type IIA, and 499.4 lm2 for type IIX. In type IIX), enlargement of muscle fiber of all types form
comparison of our results of adult (mean cross section area childhood to adulthood were prominent, especially type I
531 lm2 for type I, 745 lm2 for type IIA, and 712 lm2 for and IIA fibers. Considering the scantiness of type I fibers
and the age-related reduction of IIA fiber area demon-
strated in this study, IIA fiber seems to play a vital role in
the development and aging in POOM.
Although early research had indicated that there was a
selective loss of type II fibers in human limb muscles
during aging (Larsson et al. 1978), many subsequent
studies proved it not to be the case (Kirkeby and Garbarsch
2000; Lexell et al. 1988). It was found that a loss of fiber
number occurs with aging, but the decrement is equally
apparent among different fiber types (Lexell et al. 1988).
Unlike loss of fiber number, the cellular atrophy associated
with senescence is specific to each fiber types. Type II
fibers experienced a 26% reduction in cross section area,
while the size of type I fibers did not differ between young
and old individuals (Lexell et al. 1988). Even among
subtypes of type II fibers, aging appears to elicit prefer-
ential atrophy. In a follow-up study of vastus lateralis
muscle biopsy, the atrophy of type IIA fibers (14%) was
less than that of IIX fibers (25%) (Aniansson et al. 1986).
On the other hand, human jaw muscles undergo muscle and
region-specific changes in fiber composition during aging
(Monemi et al. 1999). Alterations in the masseter and the
lateral pterygoid muscles in the elderly are opposite to
those reported for limb and trunk muscles, whereas chan-
ges in the anterior and posterior bellies of the digastric
muscle resemble those of limb and trunk muscles (Monemi
et al. 1998, 1999). The present study demonstrated that
overall fiber type composition did not change with aging in
human POOM. Similar to limb or digastric muscle, the size
of type I fibers also remained relatively constant with
aging. However, a significant reduction of type IIA fiber
area in human POOM (11.3%) was noted, while the
reduction of cross section area of fiber IIX (6.5%) was not
statistically significant. Unlike limb muscles, the extent of
selective type II fiber atrophy in human POOM appeared
less prominent, and the phenomenon is more significant in
Fig. 4 Correlations of relative contents of different MyHC isoforms
type IIA fibers.
and age (n = 58). A statistically significant decrease of MyHC IIA By examining biopsy samples from vastus lateralis
content with aging was noted muscle of 7 young and 7 elderly subjects, decrease of

123
182 J Mol Hist (2007) 38:177–182

MyHC IIA content and increase of MyHC IIX content were and immobilization on structure and function of human skeletal
found comparing the elderly to the young (D’Antona et al. muscle fibres. J Physiol 552:499–511
Essen-Gustavsson B, Borges O (1986) Histochemical and metabolic
2003). Unlike the limb muscles, only a negative correlation characteristics of human skeletal muscle in relation to age. Acta
of the relative MyHC IIA content and age was noted in Physiol Scand 126:107–114
human POOM. Although MyHC IIX content appeared to Freilinger G, Happak W, Burggasser G, Gruber H (1990) Histo-
increase with aging in the linear regression plot, the P chemical mapping and fiber size analysis of mimic muscles.
Plast Reconstr Surg 86:422–428
value was not significant. Previous studies have demon- Goodmurphy CW, Ovalle WK (1999) Morphological study of two
strated an age-related reduction in amplitude and peak human facial muscles: orbicularis oculi and corrugator supercilii.
velocity of spontaneous and voluntary blinks (Peddireddy Clin Anat 12:1–11
et al. 2006; Wouters et al. 2001). Therefore, reduction of Happak W, Burggasser G, Gruber H (1988) Histochemical charac-
teristics of human mimic muscles. J Neurol Sci 83:25–35
the IIA fiber area and the relative MyHC IIA composition Kirkeby S, Garbarsch C (2000) Aging affects different human
in human POOM may play a role in the decrease of the muscles in various ways. An image analysis of the histomor-
muscle contraction speed during aging. phometric characteristics of fiber types in human masseter and
In summary, our study demonstrated that human POOM vastus lateralis muscles from young adults and the very old.
Histol Histopathol 15:61–71
is a type II fiber- predominant muscle with relatively small Lander T, Wirtschafter JD, McLoon LK (1996) Orbicularis oculi
fiber cross section area. The proportion of the ATPase- muscle fibers are relatively short and heterogeneous in length.
based fiber types remained unchanged when young and old Invest Ophthalmol Vis Sci 37:1732–1739
POOM were compared. Old subjects showed decreased Larsson L, Sjodin B, Karlsson J (1978) Histochemical and biochem-
ical changes in human skeletal muscle with age in sedentary
cross section area of all fibers and type IIA fibers compared males, age 22–65 years. Acta Physiol Scand 103:31–39
to young subjects. In accordance with the histomorphom- Lexell J, Taylor CC, Sjostrom M (1988) What is the cause of the
etry result, MyHC I exhibited the smallest proportion of the ageing atrophy? Total number, size and proportion of different
three isoforms. Relative MyHC IIA content decreased with fiber types studied in whole vastus lateralis muscle from 15- to
83-year-old men. J Neurol Sci 84:275–294
aging, while relative contents of MyHC I and IIX did not McCall GE, Byrnes WC, Dickinson AL, Fleck SJ (1998) Sample size
change. The results presented here, in combination with required for the accurate determination of fiber area and
findings in the previous studies (Aniansson et al. 1986; capillarity of human skeletal muscle. Can J Appl Physiol
Kirkeby and Garbarsch 2000; Lexell et al. 1988; Monemi 23:594–599
Monemi M, Eriksson PO, Eriksson A, Thornell LE (1998) Adverse
et al. 1998, 1999), allowed us to conclude that human changes in fibre type composition of the human masseter versus
POOM differs from limb or jaw muscles not only in the biceps brachii muscle during aging. J Neurol Sci 154:35–48
fiber type and MyHC composition, but also in their age- Monemi M, Eriksson PO, Kadi F, Butler-Browne GS, Thornell LE
related changes. The present study revealed some funda- (1999) Opposite changes in myosin heavy chain composition of
human masseter and biceps brachii muscles during aging. J
mental characteristics of human POOM and its changes Muscle Res Cell Motil 20:351–361
during aging, which are of value for future investigations Nelson CC, Blaivas M (1991) Orbicularis oculi muscle in children.
on disorders in blink system or periorbital rejuvenation Histologic and histochemical characteristics. Invest Ophthalmol
techniques. Vis Sci 32:646–654
Peddireddy A, Wang K, Svensson P, Arendt-Nielsen L (2006)
Influence of age and gender on the jaw-stretch and blink reflexes.
Acknowledgement The study was supported by grants from Na- Exp Brain Res 171:530–540
tional Science Council, Executive Yuan, Taiwan (NSC 92-2314-B- Scott W, Stevens J, Binder-Macleod SA (2001) Human skeletal
002-299) and National Taiwan University Hospital. muscle fiber type classifications. Phys Ther 81:1810–1816
Staron RS, Hagerman FC, Hikida RS, Murray TF, Hostler DP, Crill
MT, Ragg KE, Toma K (2000) Fiber type composition of the
References vastus lateralis muscle of young men and women. J Histochem
Cytochem 48:623–629
Aniansson A, Hedberg M, Henning GB, Grimby G (1986) Muscle Talmadge RJ, Roy RR (1993) Electrophoretic separation of rat
morphology, enzymatic activity, and muscle strength in elderly skeletal muscle myosin heavy-chain isoforms. J Appl Physiol
men: a follow-up study. Muscle Nerve 9:585–591 75:2337–2340
Brooke MH, Kaiser KK (1970) Three human myosin ATPase systems Wouters RJ, van den Bosch WA, Mulder PG, Lemij HG (2001) Upper
and their importance in muscle pathology. Neurology 20: eyelid motility in blepharoptosis and in the aging eyelid. Invest
404–405 Ophthalmol Vis Sci 42:620–625
D’Antona G, Pellegrino MA, Adami R, Rossi R, Carlizzi CN,
Canepari M, Saltin B, Bottinelli R (2003) The effect of ageing

123