Mutation Research 455 (2000) 129139

A refined protocol for conducting the low pH 6.7 Syrian

hamster embryo (SHE) cell transformation assay
L. Custer a, , D.P. Gibson b , M.J. Aardema b , R.A. LeBoeuf b
a

Covance Laboratories, Vienna, VA. 22182, USA

The Procter and Gamble Company, Cincinnati, OH 45239, USA

Abstract
The Syrian hamster embryo (SHE) cell transformation assay evaluates the potential of chemicals to induce morphological
transformation in karyotypically normal primary cells. Induction of transformation has been shown to correlate well with
the carcinogenicity of many compounds in the rodent bioassay. Historically the assay has not received wide-spread use due
to technical difficulty. An improved protocol for a low pH 6.7 assay was developed by LeBoeuf et al. [R.A. LeBoeuf, G.A.
Kerckaert, M.J. Aardema, D.P. Gibson, R. Brauninger, R.J. Isfort, Mutat. Res., 356 (1996) 85127], that greatly reduced many
of the technical difficulties associated with the SHE assay. The purpose of this paper is to describe the most current execution
of the pH 6.70 protocol including protocol refinements made since the publication of a comprehensive protocol for this assay
in Kerckaert et al. [G.A. Kerckaert, R.J. Isfort, G.J. Carr, M.J. Aardema, Mutat. Res., 356 (1996) 6584]. 2000 Elsevier
Science B.V. All rights reserved.
Keywords: Syrian hamster embryo (SHE) cells; In vitro; Cell transformation; Protocol

1. Introduction
The Syrian hamster embryo (SHE) cell transformation assay has been used to assess the carcinogenic
potential of chemicals and to study the mechanisms
of chemical carcinogenesis since the 1960s (reviewed
in [5]). The original SHE cell transformation assay
was conducted at pH 7.17.3. In 1980s LeBoeuf et al.
developed a protocol where cells are cultured at pH
6.70 [9]. This modification resulted in many improvements which include: an increase in the number of
transformants in both control and treated cultures allowing for the routine application of robust statistical
procedures to determine treatment-related increases
Corresponding author. Present address: Bristol-Myers Squibb,
HPW 17-2.05, PO Box 5400, Princeton, NJ 08543, USA. Tel.:
+1-609-818-5381; fax: +1-609-818-3675.
E-mail address: laura.custer@bms.com (L. Custer).

in transformation, a reduction in the influence of different cell isolates and serum lots thus improving the
efficiency of conducting the assay and increasing intra
and inter-laboratory reliability and a transformed phenotype that was easier to score resulting in increased
consistency of scoring of the assay across labs. In
a comprehensive review of the SHE transformation
literature, Isfort et al. [4] found that both the high
(7.17.3) pH (213 chemicals) and low (6.70) pH (48
chemicals) assays had an 80% concordance with the
rodent bioassay for the total of 261 chemicals/agents.
Our database at pH 6.7 now includes 96 chemicals
(data not included) and the concordance remains
around 80%. This continues to confirm the validity of
using the SHE assay to predict the outcome of rodent
bioassays [10]. Since Kerckaert et al. [6] published a
comprehensive protocol for the conduct of the low pH
assay, experience has grown and several modifications
have been used to make the assay easier to conduct

0027-5107/00/$ see front matter 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 7 - 5 1 0 7 ( 0 0 ) 0 0 0 9 8 - 1

130

L. Custer et al. / Mutation Research 455 (2000) 129139

and the results easier to interpret. These modifications include the number of chemical concentrations
tested, the number of dishes per experiment and the
interpretation of results when the average number of
colonies per dish are outside the ideal range. It is the
aim of this protocol update to convey our experiences
in protocol refinements and assay interpretation since
the last protocol paper [6]. General details on the
conduct of the assay are included so that this publication can exist as a stand alone protocol for conduct
of the low pH 6.7 SHE cell transformation assay. The
literature [6] should be consulted for a more detailed
justification of the overall method.

2. Materials
2.1. Animals
Syrian golden hamsters at day 13 gestation can be
obtained from Charles River Labs (Kingston or Wilmington colonies). If hamsters are obtained from a
different source, verify that the day plugs are observed
is counted as day 1.
2.2. Wash solution for cell isolation
Wash solution is prepared by adding 200 U/ml
Penicillin and 200 g/ml of Streptomycin to calcium magnesium free Hanks Balanced Salt Solution
(CMF-HBSS). This solution is stable for 30 days when
stored at 4 C. Penicillin/Streptomycin (Pen/Strep)
and CMF-HBSS can be obtained as sterile solutions
from Life Technologies (Rockville, MD).
2.3. Dissociation solution
Dissociation solution is prepared on the day of cell
isolation by adding 1.25% (v/v) trypsin, 2.5% (v/v)
pancreatin and penicillin/streptomycin at 200 U/ml
and 200 g/ml, respectively to CMF-HBSS. Before
use the solution pH is adjusted to 77.2 using a sterile
7.5% sodium bicarbonate solution. Trypsin, Enzar-T
40, can be obtained from Intergen (Purchase, NY),
thawed and stored as aliquots at 20 C up to 6
months. Pancreatin 4 USP, can be obtained from
Life Technologies thawed and stored as aliquots at
20 C up to 6 months.

2.4. Cell isolation medium

Complete culture medium is prepared by adding
130 ml of fetal bovine serum (FBS), 13 ml of
l-glutamine and 6.5 ml of Pen/Strep to 500 ml of
Dulbeccos Modified Eagles Medium-LeBoeufs
modification (DMEM-L). FBS is obtained from
Hyclone Labs (Logan, UT). A 200 mM stock of
l-glutamine is obtained from Life Technologies
(Gaithersburg, MD). DMEM-L has been previously
described in [7]. It contains low glucose and a reduced concentration of sodium bicarbonate and can
be obtained from Quality Biologicals (Gaithersburg,
MD). Because optimum pH is critical to the success
of the assay, the pH of each new lot of complete media
should be verified before use. To a sterile scintillation
vial, 10 ml of complete medium is added and incubated for at least 4 h at 37 C in humidified air with
10% CO2 . After incubation the vial is tightly capped
with a septum-style cover to prevent degassing and
allowed to reach room temperature. The pH should be
read by inserting the pH electrode through the septum
to eliminate the possibility for CO2 escape and subsequent pH change. The pH should be 6.656.75. Occasionally the culture medium pH needs to be adjusted.
If needed, the pH may be adjusted by addition of sterile 7.5% sodium bicarbonate or 1N HCl, followed by
incubation and re-analysis to confirm the correct pH.
2.5. Complete culture medium
Complete culture medium is prepared by adding
128 ml of non-heat activated FBS and 12.8 ml of
l-glutamine to 500 ml DMEM-L. This medium should
be used within 2 weeks. For optimized cell growth
and transformation complete medium pH should be
6.656.75. If needed pH should be adjusted as described above.
2.6. Cryopreservation medium
On the day of use, dimethyl sulfoxide (DMSO)
is added to complete culture medium to yield 15%
(v/v) DMSO. DMSO can be obtained from Sigma
(St. Louis, MO).
2.7. Other solutions and reagents
0.05% trypsin/0.53 mM EDTA in HBSS (detachment solution), 0.4% trypan blue stain and 7.5%

L. Custer et al. / Mutation Research 455 (2000) 129139

sodium carbonate can be obtained from Life Technology (Gaithersburg, MD). Absolute methanol and
benzo[a]pyrene (B[a]P) can be obtained from Sigma
(St. Louis, MO).

3. Procedure
3.1. Embryo harvest
Usually 34 pregnant hamsters (day 13 gestation)
are used to obtain 3040 embryos for cell isolation.
The dams are sacrificed by CO2 asphyxiation, followed by exsanguination. The ventral surface of the
hamster is swabbed with 70% ethanol. Using sterile
instruments and technique the abdominal skin is incised and retracted. The peritoneal cavity is opened
and both uterine horns are transferred to 100 mm sterile petri dishes containing ice-cold wash solution. The
uterine horns are transported on ice to a class II biohood to maintain sterility.

131

the 50 ml tubes, leaving the tissue behind. Fresh dissociation solution is added and the flask returned to
the stir plate for 10 min. As before, dissociation solution and cells are collected into 50 ml tubes containing FBS. This dissociation step is repeated a total
of 34 times. Cells are collected by centrifugation at
1000 rpm for 10 min (Beckman TJ-6) and pooled in
2050 ml of complete culture medium containing antibiotics. The cells are stored on ice until seeded into
culture flasks. The concentration of viable cells is determined by hemacytometer count using trypan blue
exclusion and cells are seeded at a density of 3 107
viable cells per T-225 flasks (2 107 in a T-150) with
complete media containing antibiotics. The flasks are
incubated at 37 C in humidified air with 10% CO2 for
4 h. Then the flasks refed with 50 ml complete culture
medium without antibiotics and returned to the incubator. If after 24 h of growth the flasks are not 6080%
confluent then they are refed and returned to the incubator. If the cultures have not reached 6080% confluence after 48 h growth then discard the cells and
repeat the cell isolation.

3.2. Dissociation of embryonic tissues

3.3. Cyropreservation of SHE cells
The embryos are removed from the uterine horns
and transferred to a new petri dish containing ice-cold
wash solution. During harvest the tissue is kept on
ice at all times. Aseptically, each embryo is decapitated, eviscerated and delimbed. The remaining tissue
is placed in a 100 ml beaker containing 5 ml of wash
solution. After pooling the tissue from two hamsters
in each beaker, it is minced into small (3 mm) pieces
using scissors. The minced tissue is placed in a sterile
trypsinization flask (Bellco, Vineland, NJ) containing
a magnetic stir bar and transferred to a stir plate. To
remove as many blood cells as possible, the tissue is
rinsed with 100 ml of wash solution at a slow stirring
speed for 5 min. The tissue is allowed to settle and the
wash solution is discarded. This rinse step is repeated
using room temperature dissociation solution.
To begin collecting cells, add 50100 ml of dissociation solution to the flask and slowly stir for 10 min.
If the tissue is stirred too fast the cell membranes
lyse and a thick cell lysate forms at the bottom of the
flask. Meanwhile, add 2 ml of cold FBS to a series
(1518) of 50 ml centrifuge tubes on ice. The flask is
removed from the stir plate and the dissociation solution is transferred by pipet (approximately 50 ml) to

When the cells are 6080% confluent they are cryopreserved. Culture medium is aspirated, the cells
rinsed twice with 10 ml CMF-HBSS and 4 ml (T-225
flasks) detachment solution added. To stop the trypsin
activity, cold complete medium without antibiotics is
added to each flask and the cells are transferred to
sterile 50 ml centrifuge tubes. The cells are pelleted
by centrifugation as described above and pooled in
5080 ml complete culture medium without antibiotics. The concentration of viable cells is determined
as described above and adjusted to 5 106 viable
cells/ml. An equal volume of cryopreservation media is added to produce 2.5 106 cells/ml in culture
medium containing 7.5% DMSO. The cell suspension
is dispensed, 1 ml/vial into prechilled cryostorage
vials and the cells are placed into a 70 C freezer
for 24 h before transfer to a liquid nitrogen freezer.
3.4. Confirmation of transforming ability
Prior to use each new cell isolate is evaluated for
its ability to undergo morphological transformation
in response to treatment with the known carcinogen

132

L. Custer et al. / Mutation Research 455 (2000) 129139

benzo[a]pyrene. Typically B[a]P concentrations of

2.5, 5 or 10 g/ml produce a statistically significant
increase over the solvent control with transformation
frequencies averaging 1.3 0.4% and solvent controls frequencies averaging 0.3 0.2%. Cell isolates
are used as long as statistically significant differences exist between the positive and solvent controls.
Stocks of cryopreserved SHE cells are usually used
for 69 months post-isolation (we have not evaluated
cell isolates beyond 9 months after cryopreservation).
If there is a cell isolate with MT frequency that is on
the lower end of the range, different lots of serum can
be used to find a lot that increases the MT frequency.
3.5. Feeder cell preparation
Typically, the use of antibiotics is limited to dissociation of embryonic tissue but some laboratories
routinely supplement complete culture medium with
penicillin 50100 U/ml, streptomycin 50100 g/ml.
This should be done with caution because penicillin
may suppress transformation in SHE cells as has been
shown in C310T1/2 cells [2].
Cryopreserved SHE cells (12 vials) are quickly
thawed at 37 C and transferred to a T-225 flask in
5070 ml of complete culture medium without antibiotics. The cells are incubated for 16 days at 37 C with
10% CO2 in humidified air until the cells are 5090%
confluent. The cells are detached by rinsing twice with
CMF-HBSS, followed by 4 ml of detachment solution. When the cells have detached, trypsin activity is
stopped with addition of 1 ml of cold FBS and the cells
are transferred to a T-25 flask on ice. The cells are exposed to an X-ray dose sufficient to inhibit replication
but not a lethal dose. Following X-irradiation, cell
concentration is adjusted to 2 104 viable cells/ml in
complete culture medium and 2 ml of this suspension
is placed into each 60 mm petri dish. Confirmation
that the cells are still viable but no longer replicating
is made by preparing five plates with feeder cells only
(for all cytotoxicity and definitive assays). A background of attached feeder cells should be visible but
no colonies. If colonies are seen in the feeder cell only
dishes, the study must be repeated using feeder cells
given a higher amount of irradiation. A lack of healthy
feeder cells on the dishes indicates that the cells were
given a lethal dose of X-rays. This will result in a low
plating efficiency of the target cells, thus the study

must be repeated using feeder cells given a lower level

of irradiation. In most laboratories, an acceptable dose
of X-rays will be 30005000 rad. At Covance Laboratories this was achieved by X-irradiation (130 kV)
times of 1530 min, using a model 43855A Faxitron
cabinet X-ray system (Wheeling, IL).
3.6. Target cell preparation
One day prior to use a vial of cryopreserved SHE
cells is rapidly thawed at 37 C and transferred to
a T-25 flask containing 4 ml of complete culture
medium. After 12 h incubation at 37 C with 10%
CO2 in humidified air, unattached cells are aspirated,
the flask is refed with 5 ml complete medium and
returned to the incubator overnight. The cells are
detached as described above and held on ice until
seeding. The number of viable cells is determined
by hemacytometer using trypan blue exclusion and
the cell concentration adjusted such that 2 ml of cells
seeded produce 35 10 colonies per dish. Plating
efficiency varies with cell isolate and serum lot but is
usually 2050% for untreated cells. To each 60 mm
petri dish containing feeder cells seeded 24 h previously, 2 ml of target cells are added and the dishes
returned to the incubator for 24 h.
3.7. Dose substance preparation
An appropriate solvent for each test chemical is selected and the pH and osmolality of the dosing solution
is measured prior to performing the dose range-finding
assay. The following solvents are tested in order of
preference: DMSO, culture medium, ethanol and acetone. Routinely, serial dilutions of the test chemical are
made in the solvent of choice at 500 the final culture
dish concentration. Each of the chemical solutions are
diluted 1:250 into culture medium to produce 2 dosing stocks. To prevent assay interference, final solvent
concentrations should be no greater than 0.2%. Final
1 concentrations are achieved by adding 4 ml of the
2 dosing stock to each 60 mm dish containing feeder
and target cells in 4 ml of culture medium. The final
solvent concentration in all dishes is 0.2%. Because
maintenance of pH 6.656.75 is critically important
for optimum growth and transformation, the pH of the
highest soluble dosing solution is determined using a
standard pH meter both at the time of preparation and

L. Custer et al. / Mutation Research 455 (2000) 129139

after at least 4 h incubation. Prior to incubation the

dosing solutions should be pH 7.0, if not the pH is
adjusted using 1N HCl or NaOH. Osmolality is determined using a standard freezing point depression
osmometer. Dosing solution osmolality should not exceed the solvent control by more than 70100 mOSM.
3.8. Cytotoxicity assay
Before conducting the transformation assay, a dose
range-finding experiment is performed to establish an
appropriate concentration range for the test chemical. This involves exposing SHE target cells in clonal
growth for either 24 h or 7 days to a range of concentrations of test chemical, usually at least 10 concentrations. Typically, 10 dishes are treated per dose
concentration. Following chemical exposure, colonies
are fixed using methanol and stained with Giemsa. The
average number of colonies per dish, plating efficiency
and relative plating efficiency are calculated for each
dose group. All of the colonies on the dish (including partial colonies on the side of the dish, diffuse or
small colonies, etc.) are counted to obtain an accurate
assessment of plating efficiency. Plating efficiency is
equal to the number of colonies per dish divided by
the number of target cells seeded per dish times 100%.
Relative plating efficiency (%RPE) is calculated by
dividing the plating efficiency of each dose concentration by the plating efficiency for the solvent control.
The cytotoxicity of each treatment group can be
measured by the reduction in relative plating efficiency
and visual observation of colony size and density.
For chemicals that reduce plating efficiency, the highest dose concentration tested should cause at least a
50% reduction in %RPE. If the size or density of the
colonies is reduced so that an evaluation of the phenotype of the colonies is not possible, then this is
considered cytotoxic and the highest dose concentration tested should decrease colony size or density by
approximately 50%. In accordance with OECD and
ICH guidelines for genotoxicity tests, the highest dose
tested should not exceed 5 mg/ml or 10 mM whichever
is lower. For precipitating non-toxic chemicals the top
dose should be the lowest dose that results in precipitation. For toxic, precipitating chemicals the top dose
will be the dose that results in at least 50% reduction in RPE unless the precipitate precludes scoring.
If precipitation precludes scoring then the top dose

133

would be the first dose that results in precipitation and

is scorable. The cumulative database for the assay indicates that most chemicals are detected without exceeding 2540% reduction in RPE (i.e. RPEs ranging
from 60 to 75%). Based on this experience, we now
routinely choose dose concentrations expected to result in actual RPEs of 40, 60, 75 and 85% with decreasing chemical concentration in the transformation
assay. If the chemical is non-toxic causing less than
10% reduction in %RPE, four concentrations will be
tested up to the maximum described above, solubility
permitting.
Due to normal variability in toxicity from assay to
assay, for steep toxicity curves, additional dose concentrations in the range between 100 and 50% toxicity
should be considered to ensure that sufficient toxicity
is achieved in the final assay.
3.9. Seeding equalized number of target cells
It has been demonstrated that morphological transformation (MT) frequency increases with decreasing
number of colonies per dish. A suggested explanation for this observation is that fewer cells produce
less fibronectin which has been shown to decrease
transformation frequency in SHE cells [6,12]. To
avoid false positive responses due to large variations
in colony numbers an acceptable range of 2545 average colonies per dish has been chosen for routine
assay performance. Therefore, to maintain an average
of 2545 colonies per dish the number of target cells
seeded must be increased for chemical doses which
produce 30% or greater reduction in RPE. To accomplish this, for toxic chemical concentrations, adjusted
numbers of target SHE cells are seeded for the SHE
cell transformation assay. Cytotoxicity data obtained
from the dose range-finding assay are used to determine which dose concentrations require target cell adjustment and to what extent. The %RPE is calculated
for each dose concentration in the range finding assay. Target cell adjustment is performed for each dose
concentration with %RPE less than 70% and is calculated as follows. The adjusted number of target cells
seeded per dish is equal to 100 the numbered target
cells seeded in the vehicle control divided by %RPE.
Rarely do the doses used in the preliminary toxicity
test result in exactly 50 or 70% reductions in RPE so
an extrapolation is made to arrive at doses that are

134

L. Custer et al. / Mutation Research 455 (2000) 129139

expected to result in appropriate toxicity. Usually dose

levels are chosen that are expected to result in one
that causes a 5060% reduction in RPE and one that
results in approximately a 30% reduction in RPE.
3.10. Transformation assay
Good assay sensitivity requires that at least 1000 or
more colonies per dose group are analyzed. With this
population size a dose response of 3.54 over background is statistically significant using a one-tailed
Fishers Exact Test at p < 0.05 [3]. To obtain at
least 1000 colonies and maintain 2545 colonies per
dish, a minimum of 40 dishes per dose group must
be dosed. Historically for ease of handling, data were
pooled from two transformation assays with 20 dishes
per dose group. Experience over time has shown that
conducting a single large assay with at least 40 dishes
is faster and more efficient, particularly for chemicals
with steep toxicity curves where toxicity might vary
substantially between experiments.
The transformation assay is conducted in an analogous manner to the dose range finding assay. Exceptions to this include increased number of plates
per treatment group, use of adjusted target cell numbers and inclusion of a positive control chemical.
X-irradiated feeder cells are seeded into forty 60 mm
dishes per treatment group and incubated overnight as
described in the Section 3.5 above. Sufficient target
SHE cells to produce 35 10 average colonies per
dish are seeded into dishes containing feeder cells and
incubated overnight as explained in the Section 3.9
above. The next day, test chemicals are formulated
and dosed as discussed above. Benzo[a]pyrene at 2.5,
5.0 or 10 g/ml is the standard positive control for
either exposure regimen. After treatment the dishes
are incubated undisturbed for 78 days. Alternatively,
after 24 h the dosing solution is replaced with 8 ml of
complete culture medium and the dishes returned to
the incubator for 67 days clonal expansion.
The assay is terminated when colonies in the vehicle
control are approximately 4 mm in diameter, i.e. 67
days following chemical treatment. Culture medium
is removed and the dishes rinsed with HBSS. The
colonies are fixed using methanol and stained with
Giemsa diluted 1:10 into phosphate buffer pH 7.27.4.
Dishes with only feeder cells should contain viable
cells but no colonies.

3.11. Evaluation of morphological transformation

Relative plating efficiency is determined after the
colonies have been stained. The total number of
colonies is recorded for each culture dish and %RPE
is calculated as described in the Section 3.8 above.
Colonies from the highest treatment group to produce
at least 50% RPE and three lower concentrations are
selected for evaluation of morphological transformation. Sometimes due to toxicity, there are substantial
numbers of small diffuse colonies that could not be
scored for morphological transformation. In addition, there may be colonies on the side of the dish
or where part of the colony is missing and cannot
be fully evaluated for morphological transformation.
When this occurs the total number of these scorable
and unscorable colonies should be recorded and used
to calculate the total colony number and to calculate
%RPE. Only the number of scorable colonies are
used in the calculation of transformation frequency.
This prevents under estimating the transformation
frequency due to the presence of a large number of
colonies of insufficient size and/or density to score.
Using a stereomicroscope, individual colonies are
evaluated for transformed morphology. Because the
primary cells represent diverse lineages, typically
several different normal colony morphologies exist
in every dish. Normal colonies exhibit orderly, often
flowing, growth patterns with minimal cell stacking
(see photographs in [6,11]). Epithelial-like normal
colonies usually have more rounded cells with a
smooth perimeter, while more fibroblast-like colonies
tend to have some criss-crossing cells at the perimeter. As observed with normal colonies, more than one
transformed colony morphology exists. Morphologically transformed (MT) colonies exhibit extensive
random-oriented three-dimensional growth with cell
stacking and criss-crossing both at colony center
and perimeter (see photographs in [6,11]). Individual cells within the colony are more basophilic than
normal cells and have decreased cytoplasm to nucleus ratios [11]. Epithelial-like transformed colonies
primarily exhibit extensive cell stacking, while more
fibroblast-like transformed colonies tend to have
very basophilic cells with an extensive criss-crossing
pattern of growth. For control of bias, initially the
vehicle control and two concentrations of the positive
control are scored blind. After assuring that at least

L. Custer et al. / Mutation Research 455 (2000) 129139

one concentration of the positive control produces

a statistically significant response, the dose groups
(including the negative and positive controls added
back in) are coded and scored blind. The dishes are
decoded and MT frequency is calculated for each
dose group. MT frequency is equal to the number of
MT colonies divided by the total number of colonies
scored for transformation times 100.
3.12. Statistical analysis
The data are evaluated for significant treatment related effect and positive dose-response trend [3]. The
validity of pooling data from multiple experiments
for analysis of treatment related effect has previously
been verified [9]. Also a binomial distribution across
independent trials has previously been verified using
goodness-of-fit analysis of vehicle and positive control data [8]. A one-sided Fishers Exact Test [1] is
used to examine treatment related effect. This method
uses a pairwise comparison of MT frequency of each
dose group and the vehicle control. A p < 0.05 level
of significance demonstrates a treatment related effect
on MT frequency. The chemical is considered positive if at least two dose concentrations cause a statistically significant increase in MT frequency relative to
the vehicle control in either the 24 h or 7 day exposure. If only one dose is statistically significant then an
unstratified binomial exact permeation trend test for
a significant positive dose-response trend (Armitage
[1]) is performed [3]. Chemicals are considered positive with a single positive dose and a positive trend
test. If treatment with a chemical results in one significant dose level but is negative in the trend test, it may
useful to do another study with doses that bracket the
positive dose to clarify the original response. Treatment with chemicals that do not result in statistically
significant increases in MT frequency in both the 24 h
or 7 day exposures are considered negative.
3.13. Assay acceptance criteria
For a valid test the following criteria must be met.
Ideally every group should have an average of 2545
colonies per dish. However, with experience it is
concluded that fewer than 25 colonies per dish is
acceptable if the dose group is negative and at least
1000 colonies have been evaluated (since too few

135

colonies on the plate will tend to overestimate a transformation response). Likewise, an average greater
than 45 colonies/dish is acceptable if the dose group
is statistically significant, regardless of the number
of colonies evaluated (since too many colonies per
plate will tend to underestimate the transformation
response). At least 1000 colonies per treatment group
should be evaluated for sufficient statistical power
but fewer than 1000 colonies is acceptable if the
dose group has a statistically significant increase in
transformation response. The MT frequency for the
solvent control should be between 00.6% and at least
one concentration of the positive control should be
statistically significant. Four treatment groups should
be evaluated for each chemical and at least 50% cytotoxicity achieved. Plating efficiency of untreated or
vehicle control groups is usually 2045%.
3.14. Examples of SHE transformation data
Numerous examples of typical SHE data have
been published previously (LeBoeuf et al. [9] among
others). Included in this report are three compounds
tested recently as coded chemicals. Two compounds
are rodent carcinogens and they were positive in the
SHE assay: cyclosporin A and cyclophosphamide.
Sulfisoxazole which is a rodent non-carcinogen is also
included and it was negative in the SHE assay. Since
both cyclosporin A and cyclophosphamide were positive with the 24 h exposure, a 7 day exposure was
not conducted. Sulfisoxazole was negative in the 24 h
exposure and therefore, it was tested in the 7 day
continuous exposure which was also negative.
3.15. Cyclosporin A
Cyclosporin A was tested at six concentrations up
to 125 g/ml (Table 1, Fig. 1). None of the concentrations tested caused any reduction in RPE. The top concentration (125 g/ml) was selected based on colony
density. All of the concentrations above 125 g/ml resulted in colonies that were unscorable because there
were too few cells/colony. All of the concentrations
tested caused significant transformation p < 0.05 except for the low concentration (50 g/ml). With five
of six concentrations tested resulting in significant
increases in transformation, cyclosporin A was concluded to be positive in the SHE assay.

136

L. Custer et al. / Mutation Research 455 (2000) 129139

Table 1
Concentration (g/ml)

Total colonies

Average (colonies/dish)

Cyclosporin A 24 h exposure
DMSO
1171
33
50
1355
35
62.5
1348
37
87.5
1333
35
100
1375
36
112.5
1291
34
125
1334
35
B[a]P 2.5 g/ml
1190
32
Cyclophosphamide monohydrate 24 h exposure
Media control
881
30
DMSO
1348
34
1000
1071
31
1250
1126
31
1500
985
26
2000
1088
32
2500
885
29
B[a]P 1.25 g/ml
1143
32
Sulfisoxazole 24 h exposure
DMSO
1534
39
25
1428
37
50
1332
33
100
1518
42
250
1574
43
500
1655
41
1000
1582
40
B[a]P 10 g/ml
904
36
Sulfisoxazole 7-day exposure
DMSO
957
32
80
1038
35
160
1020
34
]240
942
31
320
862
29
400
972
32
480
1055
35
B[a]P 10 g/ml
904
30

Transformed coloniesa

RPE (%)b

MTF (%)c

p-value

2
7
15
10
11
9
14
15

100 (36)d
108
114
108
111
106
106
100

0.171
0.517
1.113
0.750
0.800
0.697
1.049
1.262

0.1307
0.0028e
0.0321e
0.0226e
0.0460e
0.0044e
0.0013e

2
1
2
2
9
5
16
17

100 (25)d
100 (31)d
84
84
71
55
48
94

0.23
0.07
0.19
0.18
0.91
0.46
1.81
1.49

0.6113
0.5910
0.0481e
0.3211
0.0006e
0.0022e

3
8
5
2
3
7
4
13

100 (52)d
93
85
65
66
59
54
92

0.200
0.560
0.375
0.132
0.191
0.423
0.253
1.430

0.0374e

3
2
3
2
5
7
3
13

100 (46)d
108
107
98
90
68
43
94

0.313
0.193
0.294
0.212
0.580
0.722
0.284
0.930

0.0076e

Morphologically transformed colonies.

Relative plating efficiency, (test group plating efficiency/solvent control plating efficiency) 100.
c Morphological transformation frequency MTF, (number of MT colonies/total number of scorable colonies) 100.
d Plating efficiency PE, (number of colonies per dish/number of cells seeded per dish) 100.
e MTF values are significantly greater than control MFT values at p < 0.05 as determined by the Fishers Exact Test.
b

3.16. Cyclophosphamide monohydrate

Cyclophosphamide was tested in the 24 h exposure
up to a concentration of 2500 g/ml (Table 1, Fig. 2).
Treatment with this compound resulted in cytotoxicity that was dose responsive. The top dose reduced
the %RPE to 48%. Two of the six concentrations tested
caused a significant increase in transformation (1500
and 2500 g/ml). Based on these results cyclophos-

phamide was concluded to be positive in the SHE

assay.
3.17. Sulfisoxazole
In the 24 h exposure sulfisoxazole was tested at
concentrations up to 1000 g/ml (54% RPE) with no
significant increase in MTF. With a 7 day exposure
sulfisoxazole was tested to 480 g/ml (43% RPE)

L. Custer et al. / Mutation Research 455 (2000) 129139

137

Fig. 1. Morphological transformation frequency (MTF) of cyclosporin A with a 24 h exposure as a function of its toxicity (relative plating
efficiency: RPE). MTF is indicated by the filled circles, RPE by the open circles. Indicated dose groups (asterisks) are significantly greater
(p < 0.05) than the solvent control group.

Fig. 2. Morphological transformation frequency (MTF) of cyclophosphamide monohydrate with a 24 h exposure as a function of its toxicity
(relative plating efficiency: RPE). MTF is indicated by the filled circles, RPE by the open circles. Indicated dose groups (asterisks) are
significantly greater (p < 0.05) than the solvent control group.

138

L. Custer et al. / Mutation Research 455 (2000) 129139

Fig. 3. Morphological transformation frequency (MTF) of sulfisoxazole with a 24 h exposure as a function of its toxicity (relative plating
efficiency:RPE). MTF is indicated by the filled circles, RPE by the open circles.

Fig. 4. Morphological transformation frequency (MTF) of sulfisoxazole with a 7-day exposure as a function of its toxicity (relative plating
efficiency: RPE). MTF is indicated by the filled circles, RPE by the open circles.

L. Custer et al. / Mutation Research 455 (2000) 129139

and again did not result in any significant increases in

transformation (Table 1, Figs. 3 and 4). Overall, sulfisoxazole was concluded to be negative in the SHE
assay.
4. Conclusion
The protocol described in this article incorporates
refinements since the publication of a comprehensive
pH 6.70 protocol by Kerckaert et al. [6]. The refined
protocol is designed to minimize the time and complexity of the procedure required to complete a valid
SHE assay. To date over 500 agents or combinations
of agents (chemicals, radiation etc.) have been tested
in the SHE cell transformation assay. With an expanding data set results from the assay continue to show
a high concordance, sensitivity and specificity when
compared with the 2-year rodent bioassay. The assay
also continues to show good reproducibility with an
increasing number of compounds producing similar
results when tested in multiple laboratories. It is hoped
that the refinements of the assay procedures described
in this paper make the assay easier to conduct and the
results easier to interpret for other laboratories using
this assay as a component of an overall assessment of
a chemicals carcinogenic potential.
References
[1] P. Armitage, Statistical Methods in Medical Research,
Blackwell Scientific Publications, Oxford, 1971, pp. 135138.
[2] J.S. Bertram, Reduction in the formation of carcinogeninduced transformed foci by penicillin G sodium in the
C3H/10T1/2 CL8 cell lines, Cancer Lett. Sep. 7 (5) (1979)
98289.

139

[3] CYTEL, Software Corporation, StatXact, Cambridge, MA.

[4] R.J. Isfort, G.A. Kerckaert, R.A. LeBoeuf, Comparison of the
standard and reduced pH Syrian hamster embryo (SHE) cell
in vitro cell transformation assay to predict the carcinogenic
potential of chemicals, Mutat. Res. 356 (1996) 1163.
[5] R.J. Isfort, R.A. LeBoeuf, Application of in vitro cell
transformation assays to predict the carcinogenic potential of
chemicals, Mutat. Res. 365 (1996) 161173.
[6] G.A. Kerckaert, R.J. Isfort, G.J. Carr, M.J. Aardema, A
comprehensive protocol for conducting the Syrian hamster
embryo cell transformation assay at pH 6.70, Mutat. Res. 356
(1996) 6584.
[7] R.A. LeBoeuf, G.A. Kerckaert, Enhanced morphological
transformation of early passage Syrian hamster embryo cell
cultures in medium with a reduced bicarbonate concentration
and pH, Carcinogensis (Lond) 8 (1987) 689697.
[8] R.A. LeBoeuf, G.A. Kerckaert, J. Poiley, R. Raineri,
An interlaboratory comparison of enhanced morphological
transformation of Syrian hamster embryo cells cultured under
conditions of reduced bicarbonate concentration and pH,
Mutat. Res. 222 (1989) 205218.
[9] R.A. LeBoeuf, G.A. Kerckaert, M.J. Aardema, D.P.
Gibson, R. Brauninger, R.J. Isfort, The pH 6.7 Syrian
hamster embryo cell transformation assay for assessing the
carcinogenic potential of chemicals, Mutat. Res. 356 (1996)
85127.
[10] R.A. LeBoeuf, G.A. Kerckaert, M.J. Aardema, R.J. Isfort,
Use of Syrian hamster embryo and BALB/c 3T3 cell
transformation for assessing the carcinogenicity potential of
chemicals, in: D.B. McGregor, J.M. Rice, S. Venitt (Eds.),
The Use of Short and Medium-term Tests for Carcinogenic
Hazard Evaluation, IARC Scientific Publications No. 146,
1999, pp. 409425.
[11] G.M. Ridder, S.B. Stuard, G.A. Kerckaert, D.B. Cody,
R.A. LeBoeuf, R.J. Isfort, Computerized image analysis
of morphologically transformed and non-transformed Syrian
hamster embryo (SHE) cell colonies: application to objective
SHE cell transformation assay scoring, Carcinogenesis 18
(1997) 19651972.
[12] E. Rivedal, T. Sanner, Promotional effects of different phorbol
esters on morphological transformation of hamster embryo
cells, Cancer Lett. 17 (1982) 18.