A QUICK, SIMPLE AND UNBIASED METHOD TO QUANTIFY C2C12

MYOGENIC DIFFERENTIATION
PEDRO VELIC
A, PhD and CHRIS M. BUNCE, PhD
School of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
Accepted 1 February 2011
ABSTRACT: Introduction: C2C12 myoblasts undergo in vitro
myogenesis to form protein-rich multinucleated myotubes.
Determining the fraction of total nuclei incorporated into myotubes is a commonly used method to quantify the extent of differentiation, but it is labor-intensive and susceptible to operator
bias. Methods: We have developed a simple method to quantify
myotube formation using micrographs of JennerGiemsastained C2C12 cultures. Because myotubes are darkly stained
by JennerGiemsa dyes, the extent of myotube formation correlates with an increase in pixels attributed to the darkest tones.
Thus, image histograms were obtained from photographs using
ImageJ software, and the sum of the darkest tones was used
as a measure of myotube density. Results: Measurements of
myotube density mirrored those of fusion index during C2C12
differentiation and after treatment with prostaglandin D2, an inhibitor of C2C12 myogenesis. Conclusions: We propose this
inexpensive, quick, and unbiased method to quantify C2C12 differentiation as a complement of the fusion index analysis.
Muscle Nerve 44: 366370, 2011

The C2C12 myoblast cell line is a widely used

model to study myogenesis in vitro.1,2 Changing
pre-conuent C2C12 cultures from high-serum
(10% fetal bovine serum) to low-serum conditions
(25% horse serum) induces cell cycle exit, commitment to myogenic differentiation, and fusion
between myoblasts to form multinucleated myotubes, thus recapitulating adult skeletal myogenesis.1 As differentiation and cell fusion proceed,
myotubes elongate and become protein-rich structures that express myosin, a-actin, troponin, and
other components of the muscle-contractile
machinery.3
There are several methods to quantify in vitro
myogenic differentiation that can be applied to
C2C12. Determining the total levels of myotubespecic proteins, such as myogenin and myosin or
the activity of creatine kinase, gives an overall and
unbiased measure of differentiation but does not
inform about the shape and size of myotubes.4
The most widely used method consists of staining
the cultures with antibodies against myotube-specic proteins, staining the nuclei, and scoring photographs for the number of nuclei inside and outside myotubes. The fusion index, the proportion
of total nuclei inside the myotubes, is calculated.5
Abbreviations: FI, fusion index; PBS, phosphate-buffered saline; PGD2,
prostaglandin D2, PPARc, peroxisome proliferator-activated receptor
gamma; RGB, redgreenblue
Key words: fusion index, in vitro, myoblast, myotube, rapid quantification
Correspondence to: P. Velica; e-mail: velica@gmail.com
C 2011 Wiley Periodicals, Inc.
V

Published online 15 August 2011 in Wiley Online Library (wileyonlinelibrary.

com). DOI 10.1002/mus.22056

366

Quantification of Myogenesis

Other stains, such as the JennerGiemsa stain, can

also be used as long as nuclei and myotube structures are distinguishable.6 However, in the fusion
index method, large numbers of photographs
must be scored manually. This makes it labor-intensive and susceptible to operator bias, as it is often difcult to determine whether a nucleus is
inside or outside a myotube.
We have developed a quick, simple, and
unbiased method for quantifying C2C12 differentiation. In this assay, cultures are stained with JennerGiemsa dyes, resulting in lightly stained myoblasts and darkly stained myotubes, and they are
photographed using an inverted microscope. The
micrographs are analyzed using the free software
ImageJ to obtain image histograms that show the
distribution of pixels across the range of gray
tones. Because the myotubes are darkly stained,
the number of pixels allocated to the darkest tones
is used as a measurement of differentiation. This
measurement reects myotube density and can be
derived from the same photographs used to calculate the fusion index. During C2C12 differentiation, myotube density increased sharply 3 days after conuence, mirroring the increase in fusion
index, and both methods yielded identical results
when they were used to measure the reduction of
myotube formation caused by prostaglandin D2
(PGD2). Thus, this quick, simple method can be
used as an unbiased analysis for photographs of
stained C2C12 cultures to infer the amount of
myotube formation.

METHODS

(ATCC) cultures were

Cell Culture. C2C12
expanded in Dulbecco modied Eagle medium
(Lonza) supplemented with 10% (v/v) fetal bovine
serum (Gibco), 2 mM glutamine (Gibco), 100 U/ml
penicillin, and 100 lg/ml streptomycin (Gibco)
(growth medium) at 5% CO2 and 37
C. For C2C12
differentiation, 5  104 cells were seeded in six-well
plates (Costar) and cultured in growth media until
reaching 7080% conuence (day 0). Media were
then replaced with Dulbecco modied Eagle medium supplemented with 5% (v/v) horse serum
(Gibco), 2 mM glutamine, 100 U/ml penicillin, and
100 lg/ml streptomycin (differentiation medium).
Cells were kept in differentiation medium until the
end of the assay, typically between 4 and 7 days.
MUSCLE & NERVE

September 2011

FIGURE 1. JennerGiemsa stain reveals dark myotube structures during C2C12 differentiation. C2C12 cultures were differentiated
using low serum conditions after reaching 7080% confluence (day 0). Images are representative photographs of cultures fixed and
JennerGiemsa stained at days 1, 0, 1, 3, 5, and 7. Darkly stained multinucleated myotubes appear at day 3 among the lightly
stained myoblasts. Myotube size and number increase at days 5 and 7. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Myotube formation was monitored daily using an

inverted microscope. All treatments and/or timepoints were performed in biological triplicates.
JennerGiemsa Staining and Fusion Index Scoring. Cells
in wells were washed with phosphate-buffered
saline (PBS), xed with 100% methanol for 5 min,
and left to air dry for 10 min. If not stained immediately after xing, wells were covered with their
lids and stored at room temperature. Jenner staining solution (BDH, Poole UK) was diluted 1:3 in 1
mM sodium phosphate buffer (pH 5.6), and 1 ml
was incubated in the wells for 5 min followed by
washing with distilled water. Wells were then incubated with 1 ml Giemsa stain (BDH) diluted 1:20
in 1 mM sodium phosphate buffer (pH 5.6) for 10
min at room temperature before washing again
with water. Importantly, all wells within a single
experiment were stained simultaneously using the
same batch of Jenner and Giemsa dyes. Each well
was photographed in four randomly selected
regions using a digital camera (Olympus Camedia
C-5050) adapted to an inverted microscope. Protein-rich myotubes can be identied by a darker
purple color, whereas nuclei stain pink. For each
eld, the number of nuclei incorporated in myotubes and the total number of nuclei were scored.
Fusion index (FI) was calculated as the percentage
of total nuclei incorporated in myotubes.
Histogram Analysis. Images had a resolution of
2560  1920 pixels and were in a redgreenblue
(RGB) color model. Data are presented in gray
tones, as ImageJ converts RGB images to gray
scale. Gray tone images can also be used, but color
is useful when performing the aforementioned FI
scoring. Digital photographs of JennerGiemsastained cultures were analyzed using the free
image-processing software ImageJ, version 1.40g
(http://rsbweb.nih.gov/ij). Image histogram analysis can be obtained by selecting Analyze/Histogram or pressing ctrlH. The x-axis of the
image histogram represents the range of 255 gray
tones (0 black, 255 white), and the y-axis is
the number of pixels attributed to each tone. For
Quantification of Myogenesis

each image, the histogram data were copied into a

spreadsheet, and the average pixel count for each
tone was calculated for the replicate images and
used to plot an average histogram for each condition or time-point. Histograms derived from myotube-rich cultures have a larger proportion of pixels attributed to the darkest tones.
For quantication of myotube density the mean
number of pixels in each image attributed to tones
075 was calculated between replicates for each
condition or time-point. The calculations for several images can be rapidly performed using a
spreadsheet template. The threshold of darkest
tones (75) is chosen arbitrarily but depends on the
histogram shape. The shape may vary between
experiments and should include the interval of
tones where the difference between conditions is
greatest. Importantly, the same threshold must be
used for all images in a single experiment.
RESULTS
Histogram Analysis of Differentiating C2C12 Cultures. C2C12 myoblasts were cultured in six-well

plates and induced to differentiate over the period

of 7 days after change to low serum conditions
(day 0). At days 1, 0, 1, 3, 5, and 7, cultures were
methanol-xed and, at the end of the time-course,
stained simultaneously using the same batches of
Jenner and Giemsa dyes. Dark purplestained multinuclear myotubes were observed from day 3
onward alongside mononuclear unfused myoblasts
(Fig. 1). The size and length of myotubes
increased gradually until day 7. Each replicate well
was photographed in four randomly selected
regions, and the images were analyzed using the
ImageJ software. For each photograph an image
histogram was obtained. Image histograms represent the number of pixels attributed to each of the
255 gray tones, in which 0 is black and 255 is
white. For each time-point, the average pixel count
for each of the 255 tones was calculated to generate an average image histogram (Fig. 2). At day
1, when cell density was low, image pixels were
mostly distributed to the lighter tones. However, as
MUSCLE & NERVE

September 2011

367

FIGURE 2. Average image histograms during C2C12 differentiation. Stained cultures from days 1, 0, 1, 3, 5, and 7 postconfluence were photographed (four fields per replicate, three
replicates per time-point), and each image was analyzed using
ImageJ to obtain an image histogram. The x-axis shows the
range of gray tones ranging from 0 (black) to 255 (white),
whereas the y-axis shows the number of pixels attributed to
each tone. Lines represent the average pixel count for each
tone, and the error bars represent the standard deviation of the
three replicates. As C2C12 cultures differentiated to form dark
myotubes, the more the pixels became allocated to the darkest
tones. The difference in pixel distribution is evidenced in the
zoomed histogram showing tones 075. The light pink color of
the myoblast cytosol falls between tones 110 and 150. [Color
figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

cell density increased from day 0 onward, pixel distribution gradually shifted to the darker tones.
The amount of pixels in the darkest tones was
greatly increased at days 3, 5, and 7, correlating
with the appearance of darkly stained myotubes.
The amplied histogram showing tones 075 evidences the shift in pixel distribution to the darker
tones during C2C12 differentiation.
Increase in Pixel Number in Dark Tones Correlates
with Myotube Formation and Fusion Index. For the
FI, photographs were scored for total and myotube-incorporated nuclei (Fig. 3A). Total nuclei

number rapidly increased up to day 1 followed by

a slow decrease until day 7. The number of nuclei
inside myotubes sharply increased between days 1
and 3 and remained high thereafter, suggesting
the decrease in total nuclei number was due to
death of unfused cells. As a result, the FI showed
an accentuated increase between days 1 and 3, followed by slower increases at days 5 and 7 (Fig. 3B).
For the quantication of myotube density, the sum
of pixels attributed to tones 075, in the darker
end of the spectrum, was calculated for each photograph. The choice of tone 75 as a threshold is arbitrary but should be a tone at which the difference of pixel count between curves is greater. In
this case a threshold between tones 50 and 75
would be acceptable (Fig. 2). However, the same
threshold must be used in all images of the same
experiment. Myotube density remained low until
differentiation day 1, increased sharply between
days 1 and 3, and was followed by slower increases
until day 7 (Fig. 3C). The changes in myotube
density correlated with the appearance of darkly
stained myotubes and mimicked the evolution of
the FI during C2C12 differentiation (Fig. 3B).
Thus, the sum of pixels attributed to the darkest
tones appears to be a comparable measure of
C2C12 differentiation.
Exposure of C2C12 cells to drugs and signaling
molecules as well as other types of experimental
manipulations are common approaches to studying
myogenesis. The images shown in Figure 4A were
used to calculate the FIs plotted in Figure 4B, as
previously reported by our group.6 In the earlier
study, C2C12 cultures were differentiated for 4
days in the presence of 10 lM PGD2, an inhibitor
of myogenesis; 10 lM ciglitazone, a synthetic peroxisome proliferator-activated receptor gamma
(PPARc) agonist; 1 lM GW9662, a PPARc antagonist; and the combination of 1 lM GW9662 and 10

FIGURE 3. Changes in nuclei number, fusion index, and myotube density during C2C12 differentiation. Photographs of stained cultures for days 1, 0, 1, 3, 5, and 7 post-confluence were analyzed. Gray open circles or diamonds represent individual replicates for
each time-point, whereas black filled circles and squares connected by lines represent the average of the replicates. (A) Circles: total
nuclei number; diamonds: number of nuclei inside myotubes. (B) Fusion index calculated as the fraction of total nuclei present inside
myotubes. (C) Myotube density calculated as the sum of pixels attributed to tones 075 (see image histograms in Fig. 2).
368

Quantification of Myogenesis

MUSCLE & NERVE

September 2011

FIGURE 4. Comparison of myotube density and fusion index after C2C12 treatment. C2C12 cells were differentiated for 4 days in the
presence of 10 lM prostaglandin D2 (PGD2), 10 lM ciglitazone, 1 lM GW9662, or a combination of 1 lM GW9662 and 10 lM PGD2.
Data respective to this experiment have been presented previously by our group.6 (A) Representative photographs of JennerGiemsastained cultures. (B) Fusion index. (C) Myotube density calculated as the sum of pixels attributed to tones 075. Data represent the
average of three biological replicates 6 standard deviation. *P < 0.05 relative to vehicle control. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]

lM PGD2. As reported previously, PGD2, alone or

in combination with GW9662, caused a 72% and
71% decrease in FI relative to vehicle treatment,
respectively, whereas ciglitazone and GW9662
alone had no effect. Herein we have reanalyzed
the photographs using the myotube density assay
and the results are shown in Figure 4C (average
histograms are shown in Fig. S1, refer to Supplementary Material). In agreement with the FI, myotube density was reduced by 67% and 72% in cells
treated with PGD2 alone or in combination with
GW9662, respectively, and was not affected by
GW9662 alone. However, in contrast to the FI,
ciglitazone treatment caused a 33% decrease in
myotube density relative to vehicle control. This
discrepancy is due to ciglitazone causing a proportional 30% decrease in both total and myotubeincorporated nuclei (see Fig. S2, Supplementary
Material), resulting in an unaltered FI relative to
vehicle control. Nevertheless, the absolute area
covered by myotubes was reduced, as detected by
the myotube density assay.

DISCUSSION

Quantifying myotube formation is an essential aspect of using the C2C12 myoblast system. Herein
Quantification of Myogenesis

we have presented a method to quantify myotube

density using photographs of stained cultures.
First, the JennerGiemsa staining process is inexpensive, simple, and can be performed in less than
1 hour even with large sample numbers. These
dyes contain methylene blue, which stains nuclei a
light pink color, and eosin, which stains muscle
bers a dark purple color, thus allowing for easy
distinction of the protein-rich myotubes.
The stained cultures are then digitally photographed and analyzed using ImageJ to obtain the
average image histograms for each condition. The
average histogram can be quickly calculated by
copying the histogram data from the software into
a spreadsheet and, when plotted, reveals the differences between treatments in pixel distribution
across the range of gray tones. Because myotubes
appear as dark structures, the more myotube coverage (either by increased number, increased size, or
increased accumulation of protein), the more pixels appear allocated to the darkest tones. Myotube
density is calculated as the sum of pixels allocated
to the darkest tones.
Both FI and myotube density can be calculated
from the same photographs and, as shown here,
both measurements yielded comparable results
during a time-course of C2C12 differentiation and
MUSCLE & NERVE

September 2011

369

in an experiment with treatments that altered the

level of myotube formation. However, ciglitazone
treatment caused no change in FI, whereas a signicant reduction in myotube density was
detected. This highlights that both methods, despite monitoring C2C12 myotube formation, do
not measure the same phenomena of the process.
For instance, hyper- or hypotrophism of myotubes
or higher accumulation of protein, without the
incorporation of nuclei, can lead to discrepancies
between FI and myotube density measurements.
Therefore, the myotube density method should be
used as a complement to the FI and does not dispense a critical analysis of the shape and size of
myotubes.
Myotube density as determined by image histograms has the advantage of being a quicker and
unbiased method. Therefore, we propose that this

370

Quantification of Myogenesis

method should be used as a complement to FI

analysis, because it offers the opportunity to detect
differences that are otherwise ignored and can
reduce operator bias.
REFERENCES
1. Yaffe D, Saxel O. A myogenic cell line with altered serum requirements for differentiation. Differentiation 1977;7:159166.
2. Yaffe D, Saxel O. Serial passaging and differentiation of myogenic
cells isolated from dystrophic mouse muscle. Nature 1977;270:
725727.
3. Cabane C, Englaro W, Yeao K, Ragno M, Derijard B. Regulation of
C2C12 myogenic terminal differentiation by MKK3/p38alpha pathway. Am J Physiol Cell Physiol 2003;284:C658C666.
4. Khurana A, Dey CS. Subtype specic roles of mitogen activated protein kinases in L6E9 skeletal muscle cell differentiation. Mol Cell Biochem 2002;238:2739.
5. Bello NF, Lamsoul I, Heuze ML, Metais A, Moreaux G, Calderwood
DA, et al. The E3 ubiquitin ligase specicity subunit ASB2beta is a
novel regulator of muscle differentiation that targets lamin B to proteasomal degradation. Cell Death Differ 2009;16:921932.
6. Velica P, Khanim FL, Bunce CM. Prostaglandin D(2) inhibits C2C12
myogenesis. Mol Cell Endocrinol 2010;319:7178.

MUSCLE & NERVE

September 2011