www.elsevier.com/locate/toxinvit

Toxicology in Vitro 19 (2005) 389–397

Comparison of MatrigelTM and gelatin substratafor feeder-free culture of undifferentiated mouse embryonic

stem cells for toxicity testing

A.R. Greenlee *,1, T.A. Kronenwetter-Koepel, S.J. Kaiser, K. Liu 2

Marshfield Clinic Research Foundation, 1000 North Oak Avenue, Marshfield, WI 54449, USA

Received 10 May 2004; accepted 18 November 2004

Abstract

Murine embryonic stem (mES) cells have been used to evaluate cytotoxicity and developmental injury following exposure to

embryotoxic agents. However, maintaining a homogeneous population of undifferentiated mES cells for this purpose has been com-

plicated by the need for continuous co-culture with murine embryonic fibroblast (mEF) cells or limited passaging on plastic surfaces

coated with gelatin. Here, we compare the synthetic basement membrane MatrigelTM with 0.1% gelatin substratum for feeder-free

propagation of undifferentiated mES cells. Biomarkers of pluripotentiality, chromosome number, caspase-3 expression, and cardio-

myocyte differentiation were monitored for mES cells cultured on MatrigelTM or 0.1% gelatin up to passage 7 (P7). Our results suggest

that choice of substratum had no significant effect on population doubling time, cell viability, stage-specific embryonic antigen-1

(SSEA-1) expression, or early passage formation of beating cardiomyocytes (all PP 0.09). In other comparisons, however, Matri-

gelTM supported significantly higher synthesis of alkaline phosphatase (7.7 · 10�3 ± 0.8 vs 6.6 · 10�3 ± 0.8 units/liter/cell, respec-

tively, P = 0.012), overall expression of activated caspase-3 following exposure to 5, 10, 50, 100 and 500 parts per billion (ppb)

sodium arsenite (P < 0.0001), and percent development to beating cardiomyocytes at P7 (P = 0.01). Together, our findings suggest

that MatrigelTM shows promise as a substrate for feeder-free propagation of undifferentiated mES cells for embryotoxicity endpoints.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Mouse embryonic stem cells; MatrigelTM; Gelatin; Alkaline phosphatase; SSEA-1 expression; Cardiomyocyte differentiation; Caspase-3;

Chromosome number

0887-2333/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tiv.2004.11.002

Abbreviations: ATCC, American Type Culture Collection; bFGF,

basic fibroblast growth factor; BSA, bovine serum albumin; DMEM,

Dulbecco�s modified Eagle�s medium; EB, embryoid body; EDTA,

ethylenediaminetetraacetic acid; EST, embryonic stem cell test; hES,

human embryonic stem; hSCF, human stem cell factor; LIF, leukemia

inhibitory factor; mEF, murine embryonic fibroblast; mES, murine

embryonic stem; P, passage; PBS, phosphate buffered saline; ppb, parts

per billion; PS, pen/strep; RT-PCR, reverse transcription-polymerase

chain reaction; SSEA-1, stage-specific embryonic antigen-1.* Corresponding author. Tel./fax: +1 541 962 3389.

E-mail address: greenlee@ohsu.edu (A.R. Greenlee).1 Current address: Oregon Health and Science University, School

of Nursing, One University Boulevard, LaGrande, OR 97850, USA.2 Current address: Department of Biostatistics, Forest Research

Institute, Harborside Financial Center, Plaza V, Jersey City, NJ 07303,

USA.

1. Introduction

Undifferentiated murine embryonic stem (mES) cells,

derived from the inner cell mass of blastocyst-stage em-

bryos, have shown promise as indicators of cellular death

and developmental injury following exposure to embryo-

toxic agents (Spielmann et al., 1997). Depending on envi-

ronmental cues, stem cells have the ability to eitherperpetually divide or step through a series of matura-

tional changes that mirror normal embryonic develop-

ment. Cellular death and altered progression of mES

cells to pulsating cardiomyocytes in the presence of chem-

ical toxicants serve as two endpoints for the embryonic

390 A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397

stem cell test (EST) (Wobus et al., 1991; Spielmann et al.,

1997). Validation studies suggest the EST assay reliably

discriminates between three classes of test compounds

with differing toxicity potentials (Scholz et al., 1999;

Spielmann et al., 2001). Improvements to the assay

include the incorporation of molecular endpoints of so-matic cell differentiation using reporter genes and cDNA

microarrays (Rohwedel et al., 2001; Seiler et al., 2004).

The recent discovery that mES cells have the capacity

to develop into male or female gametes presents the

opportunity to examine toxicant effects on the reproduc-

tive cycle at its earliest juncture (Hubner et al., 2003; Ohta

et al., 2004; Geijsen et al., 2004).

Maintaining a homogeneous population of undiffer-entiated mES cells is fundamental to the reliability and

reproducibility of the EST. Co-culturing stem cells with

mitotically inactivated murine embryonic fibroblasts

(mEFs) has been used for this purpose (Heuer et al.,

1993). However, the mixing of cell types makes cytotoxi-

city and other endpoints difficult to measure in the tar-

get mES cells. Strategies for feeder-free maintenance of

undifferentiated mES cells include supplementing med-ium with growth factors (Williams et al., 1988; Pease

et al., 1990) or conditioned medium (Vogel, 1993), coat-

ing growth surfaces with gelatin, or frequent subcultur-

ing of cells on plastic surfaces (Spielmann et al., 1997).

Despite these efforts to maintain a pluripotential pheno-

type over time, changes in morphology and biomarkers

suggest gains in differentiation and declines in the num-

ber of cells undergoing self-renewal.The present study evaluates two substrata, MatrigelTM

and 0.1% gelatin, for their ability to maintain a homoge-

neous population of undifferentiated mES cells without

fibroblast feeder layers. Biomarkers of pluripotentiality,

chromosome number, caspase-3 expression, growth

characteristics, and cardiomyocyte differentiation were

monitored up to passage 7 (P7). Our findings suggest

that MatrigelTM, in combination with growth factors,provides an alternative, unified approach for feeder-free

propagation of undifferentiated mES cells for embryo-

toxic endpoints. A streamlined regimen for maintaining

undifferentiated stem cells may help with large-scale test-

ing strategies aimed at reducing the backlog of xenobio-

tics with uncharacterized effects on early development

(Congress of the US OTA, 1995; Chapin et al., 2004).

2. Materials and methods

2.1. Experimental design

The purpose of these experiments was to determine

the effects of two substrata on the pluripotential state

of mES cells up to seven passages in vitro. MatrigelTM

and 0.1% gelatin substrata, in combination with three

growth factors (leukemia inhibitory factor, LIF; basic

fibroblast growth factor, bFGF; and, human stem cell

factor, hSCF) (Matsui et al., 1992) were compared for

their ability to support mES cell expression of biomark-

ers of differentiation and cytotoxicity. Chromosome

number, cell doubling time, and viability of mES grown

on the two substrata were also evaluated.

2.2. Cells and growth factors

D3 mES cells from mouse blastocyst embryos

(Doetschman et al., 1985) were purchased as frozen ali-

quots from American Type Culture Collection (ATCC;

CRL 1934, Manassas, VA, USA). Culture medium con-

sisted of 90% Dulbecco�s modified Eagle�s medium(DMEM; ATCC), 10% VitacellTM fetal bovine serum

(ATCC), 0.1 mM b-mercaptoethanol, 10 U/ml penicillin

and 10 lg/ml streptomycin (Sigma Chemical Co., St.

Louis, MO, USA), 1000 U/ml LIF (Chemicon Interna-

tional, Inc., Temecula, CA, USA), 20 ng/ml recombi-

nant human bFGF (R&D Systems, Minneapolis, MN,

USA) and 40 ng/ml recombinant hSCF (R&D Systems).

LIF was stored as a sterile solution at 4 �C. The hSCFwas received lyophilized and reconstituted with sterile

1· phosphate buffered saline (PBS), 0.4% bovine serum

albumin (BSA), and 1% pen/strep (PS), aliquoted and

stored at �20 �C. The bFGF was reconstituted with

PBS/BSA/PS, as above, plus 100 ng/ml heparin.

2.3. Preparation of MatrigelTM and gelatin substrata

Cells were plated at a density of 5 · 104 cells onto either

a layer of growth factor-reduced MatrigelTM (Becton

Dickinson Biosciences Clontech, Palo Alto, CA, USA)

or 0.1% gelatin (G2500, Porcine Skin Type A, Sigma).

MatrigelTM wells were prepared by spreading 50 ll ali-quots of cold undilutedMatrigelTM into Falcon 35 mm tis-

sue culture wells (Becton Dickinson and Co., Franklin

Lakes, NJ, USA) maintained on ice. Substratum gelledby incubating covered plates 2 h at room temperature.

Gelatin-coated wells were prepared by adding 2.0 ml ali-

quots of 0.1% gelatin in PBS to Falcon 35 mm tissue cul-

ture wells. Gelatin was aspirated after 5 min, leaving a

light film. Culture medium was immediately added to

the gelatin-coated wells to prevent desiccation. Matri-

gelTM- and gelatin-coated plates were used the same day.

mES cells were passaged every 3 or 4 days in 35 mmwells on MatrigelTM or gelatin. Adherent mES cells were

dissociated with 0.25% trypsin-ethylenediaminetetraace-

tic acid (EDTA; Gibco-Invitrogen, Grand Island, NY,

USA) at 37 �C for 10 min and washed in culture medium

without growth factors.

2.4. Control cell populations

Cryopreserved mES cells, previously expanded on

MatrigelTM and characterized for SSEA-1 expression at

A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397 391

P4, were used as a positive control for SSEA-1 activity of

mES cells grown on the two substrata. The cryopreserved

mES cells served as a positive control formonitoring anti-

body activity and standardizing flow cytometer settings.

Differentiated, mEFs (CRL 1503; ATCC) were used

as a negative control for alkaline phosphatase expres-sion. mEFs were cultured in 90% DMEM, 10% Vita-

cellTM fetal bovine serum, 0.1 mM b-mercaptoethanol,

10 U/ml penicillin, and 10 lg/ml streptomycin without

additional growth factors. Cells were plated at low den-

sity in T-25 cm2 flasks (Becton Dickinson Labware, Bed-

ford, MA, USA) without MatrigelTM or gelatin and

passaged at confluency. On the day of biomarker analy-

sis, mEFs were dissociated, washed, and used at cell con-centrations matched to those of mES cells.

2.5. Cell growth and viability

Population doubling time per 24 h and cell viability

were evaluated to determine changes in growth charac-

teristics for cells cultured to P7 on MatrigelTM or 0.1%

gelatin. Cumulative population doublings were deter-mined using the method described in McAteer and

Davis (1994). It was assumed that the population growth

was exponential at the plating density used. The number

of generations was then determined using common loga-

rithms and expressed as population doublings per 24 h.

Cell viability was determined by trypan blue exclusion.

2.6. Chromosome number

The number of chromosomes per metaphase spread

was determined for mES cells cultured on MatrigelTM

or 0.1% gelatin. Spreads were prepared using the meth-

od of MacDonald (1994). Chromosome counts were

performed on 20 spreads from each treatment group at

P1 and P4. Chromosome numbers per spread were deter-

mined using photographs from a Nikon Eclipse 600microscope and digital spotmatic camera at 600·magnification.

2.7. Alkaline phosphatase activity

Two techniques were used to determine alkaline phos-

phatase activity for cells cultured on MatrigelTM or 0.1%

gelatin. A staining protocol described by Donovan et al.(1986) was used to screen for enzyme activity character-

ized as a red reaction product in adherent mES cell colo-

nies. Specific enzyme activitywas inhibited in the presence

of 1 mM levamisole hydrochloride (L9756, Sigma). Cells

were photographed before and 15 min after adding the

stain using 100· magnification on a Nikon Diaphot in-

verted microscope fitted with a Nikon 2000 SLR camera.

An enzyme quantification kit (ALP10, Sigma) wasused to more precisely measure alkaline phosphatase

activity on a per cell basis. Adherent mES cells and

negative control mEF cells were dissociated, washed,

and adjusted to 3 · 106 cells/ml. A total of 2 · 104 cells

in 6 ll was lysed with 1 ll of 0.1% Triton-X (Sigma)

for 5 min before adding to 318 ll of p-nitrophenyl phos-phate. Alkaline phosphatase activity, in enzyme units/

liter/cell, was determined at kabs 405 nm, 37 �C at1 min intervals for 10 min using a Beckman DU-600

spectrophotometer and instrument software. Positive

controls, provided with the kit, and mEF negative con-

trol cells were included with each assay. Results were ac-

cepted when mean control values were within defined

limits.

2.8. SSEA-1 immunofluorescence

mES cells were analyzed for SSEA-1 expression after

passaging to P5 on MatrigelTM or 0.1% gelatin. Positive

control mES cells were included in each assay. Adher-

ent cells were dissociated, rinsed in 500 ll Perm/WashTM

buffer (Pharmingen, San Diego, CA, USA) and

counted. A total of 105 cells from each population in

10 ll were added to wells of a microtiter plate (Lab-Source, Chicago, IL, USA), fixed 30 min in 190 llCytofix/CytopermTM solution (Pharmingen), and washed

twice with 200 ll Perm/WashTM buffer, then incubated

30 min on ice with 100 ll mouse SSEA-1 antibody

(Developmental Studies Hybridoma Bank, Iowa City,

IA, USA) diluted 1:4 in Perm/WashTM buffer. Cells were

rinsed twice with 200 ll Perm/WashTM buffer and then

incubated 30 min on ice in the dark with 100 ll goatanti-mouse immunoglobulin-FITC (Pharmingen) di-

luted 1:50 in Perm/WashTM buffer. Cells were then

washed twice and diluted in 400 ll Perm/WashTM buffer.

Analysis was performed on a MoFloTM (DakoCytoma-

tion, Fort Collins, CO, USA) flow cytometer with an

Omnichrome visible laser (Melles Griot, Carlsbad,

CA, USA) at 488 nm and 155 mW. A total of 10,000

cells was analyzed for each specimen. Data acquisitionand analyses were performed using SummitTM software

(DakoCytomation).

2.9. Detection of activated caspase-3

Sodium arsenite was chosen as a model toxicant be-

cause it is a known teratogen and activator of cellular

death by apoptosis (Mirkes and Little, 1998). Arseniccontaminated water is a human concern because of the

increased risk for spontaneous abortion, stillbirth, and

preterm delivery reported for women drinking water

with high levels of arsenic (Ahmad et al., 2001). mES

cells, cultured on either MatrigelTM or 0.1% gelatin, were

incubated 16 h with 0, 5, 10, 50, 100, and 500 parts per

billion (ppb) sodium arsenite diluted in culture medium

(J.T. Baker, Phillipsburg, NJ, USA). At the end of theexposure period, cells were dissociated, washed in PBS

(pH 7.2), counted, and 105 cells from each population

392 A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397

were added to microtiter wells. Cells were pelleted and

fixed 30 min at room temperature with 100 ll Cytofix/CytopermTM. Cells were washed twice in 200 ll Perm/

WashTM buffer and incubated 1 h at room temperature

with 100 ll of a 1:10 dilution of purified phycoery-

thrin-conjugated rabbit anti-activated caspase-3 mono-clonal antibody (Pharmingen). Cells were washed twice

in 200 ll Perm/WashTM buffer and diluted to 400 ll inPerm/WashTM buffer. Ten thousand cells from each pop-

ulation were analyzed on a MoFloTM flow cytometer.

2.10. Differentiation into cardiomyocytes

Differentiation of mES cells into embryoid bodies(EBs) (aggregates of differentiating embryonic stem

cells) and contracting cardiomyocytes was determined

using the ‘‘hanging drop’’ method (Wobus et al.,

1991). mES cells cultured on either MatrigelTM or 0.1%

gelatin were harvested at P1 and P7 and dissociated,

washed, and suspended in EB medium containing 80%

DMEM, 15% fetal bovine serum, 2 mM LL-glutamine,

0.1 mM b-mercaptoethanol, 10 U/ml penicillin/10 lg/ml streptomycin, and no growth factors. Droplets

(20 ll each) containing approximately 1000 cells from

each population were placed on the lids of 100 mm petri

dishes filled with 5 ml 1· PBS. Closed plates were incu-

bated 3 days at 5% CO2, 37 �C at which time aggregates

within the hanging droplets were pooled and transferred

to 100 mm non-adhesive Falcon Optilux tissue culture

plates and incubated an additional 3 days in EB med-ium. Individual EBs were placed into Falcon 24 well tis-

sue culture plates (Becton Dickinson Labware) and

observed daily for 2 weeks for attachment and develop-

ment into pulsating cardiomyocytes.

Fig. 1. Line graphs comparing population doublings per 24 h and

percent viability for mES cells cultured on MatrigelTM or 0.1% gelatin.

Open boxes and triangles correspond to the means of duplicate values

for cells grown on MatrigelTM. Closed boxes and triangles correspond

to means of duplicate values for cells grown on 0.1% gelatin. Choice of

substratum did not affect mean population doublings per 24 h

(P = 0.09) or mean percent cell viability (P = 0.77).

2.11. Statistics

Two-way ANOVA was used to compare the effect of2 substrata (MatrigelTM or gelatin) on population dou-

blings and cell viability over seven passages. Difference

among two substrata and one positive control with re-

spect to SSEA-1 expression was tested using two-way

ANOVA after controlling for experimental conditions.

Three-way ANOVA was employed to analyze caspase-

3 expression with substratum, dose of sodium arsenite

and experimental conditions as the three analysis fac-tors. Post-hoc, pair wise comparisons were conducted

using t-tests. One-way ANOVA was applied to evaluate

the effects of substrata on alkaline phosphatase synthe-

sis, followed by pair wise t-test. Percent development

of beating cardiomyocytes was compared between two

substrata at P1 and P7 using linear mixed models to ac-

count for the correlation among measurements across

days in cultures (Verbeke and Molenberghs, 2000). TheP-values were not adjusted for multiple comparisons.

3. Results

3.1. Cell growth, viability and chromosome numbers

Population doublings per 24 h and percent viability

were measured at each passage to determine if mES cellsdisplayed differences in growth characteristics related to

substratum over seven passages (Fig. 1). Cell popula-

tions doubled a mean of 1.5 ± 0.1 and 1.4 ± 0.2 times

every 24 h for cells grown on MatrigelTM or gelatin,

respectively (P = 0.09). Cell viability was not signifi-

cantly different for mES cells grown on MatrigelTM or

gelatin substrata (98.0 ± 1.2% vs 97.7 ± 2.0%, respec-

tively, P = 0.77).Chromosome counts were performed on 20 meta-

phase spreads for P1 and P4 mES cells maintained on

either MatrigelTM or gelatin (Table 1). Of mES cells

grown on MatrigelTM or gelatin, 35–45% contained 40

chromosomes (modal number for mouse), 10–20% con-

tained less than 40 chromosomes, and 35–55% con-

tained more than 40 chromosomes. These findings

suggest that the mES cells used in our study were aneu-ploid at the start of the experiment and that chromo-

some numbers were unaffected by choice substratum.

3.2. Alkaline phosphatase activity and colony morphology

Alkaline phosphatase enzyme activity is a marker of

undifferentiated embryonic cells. An enzyme staining

protocol and a quantification assay were used to com-pare the effects of substrata on enzyme levels in mES

cells at P7. Fig. 2a and b show staining results of mES

cells before and after incubating cells grown on Matri-

gelTM with reagents specific for alkaline phosphatase.

Fig. 2c and d show results for cells grown on 0.1% gel-

atin. The amount of red cytoplasmic staining suggests

both substrata supported alkaline phosphatase synthesis

Fig. 3. Bar graph showing mean alkaline phosphatase enzyme units

per liter per cell ± SD for mES cells cultured on MatrigelTM or gelatin.

Results indicate that mES cells grown on MatrigelTM synthesize greater

amounts of alkaline phosphatase than cells grown on 0.1% gelatin

(P = 0.012). The negative control cells (differentiated mEF cells)

synthesized significantly less alkaline phosphatase than either mES

cell population (both P < 0.0001).

Fig. 2. Photomicrographs of mES cells cultured seven passages on MatrigelTM (2a, 2b) or 0.1% gelatin (2c, 2d). Figs. 2a and c show results before

addition of the stain to detect alkaline phosphatase activity and Figs. 2b and d show results 15 min after staining with reagents. Stem cells on

MatrigelTM grew as uniformly concentric, compact, raised colonies (2a, 2b), whereas colonies on gelatin (2c, 2d) grew as spreading, flattened colonies

with an occasionally round, raised colony. Cells negative for enzymatic activity appear colorless after staining (region with black-filled arrow, 2d).

Table 1

Distribution of 20 metaphase spreads for mES cells cultured on MatrigelTM or gelatin evaluated at P1 and P4

Substratum Number of chromosomes per metaphase spread

638 39 40 41 42 P43 Total spreads examined

MatrigelTM P1 2 0 7 7 1 3 20

Gelatin P1 2 1 8 7 0 2 20

MatrigelTM P4 1 1 8 8 0 2 20

Gelatin P4 2 2 9 5 0 2 20

A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397 393

to P7. The staining reaction was inhibited by the addi-

tion of enzyme-specific 1 mM levamisole hydrochloride

(data not shown).Fig. 2 demonstrates colony morphologies representa-

tive of mES cells grown on the two substrata. Stem cells

growing on MatrigelTM (Fig. 2a and b) appeared as uni-

formly concentric, compact, raised colonies. mES cells

growing on gelatin (Fig. 2c and d) appeared as spread-

ing, flattened colonies with an occasionally round and

raised appearance.

A quantification assay was used to more preciselymeasure alkaline phosphatase enzyme activity in cells

grown on the two substrata. Fig. 3 shows significantly

higher amounts of enzyme activity in mES cells grown

on MatrigelTM when compared to results of cells on gel-

atin (7.7 · 10�3 ± 0.8 vs 6.6 · 10�3 ± 0.8 units/liter/cell,

respectively, P = 0.012). In comparison to mEF differen-

tiated control cells (1.09 · 10�3 ± 0.07 units/liter/cell),

undifferentiated mES cells grown on MatrigelTM or gela-tin synthesized significantly greater amounts of alkaline

phosphatase (both P < 0.0001).

3.3. SSEA-1 activity

The monoclonal antibody SSEA-1 defines a stage-

specific mouse embryonic antigen and is useful as a

Fig. 4. Bar graphs showing mean percent SSEA-1 ± SD for mES cells

cultured on MatrigelTM or 0.1% gelatin. Approximately 67% and 65%

of mES cells grown on MatrigelTM and 0.1% gelatin express SSEA-1,

respectively (P = 0.64). Data points represent triplicate observations

from two independent experiments. Comparisons with results obtained

for the positive control cells suggest that SSEA-1 expression was not

significantly different from that obtained for cells grown on the two

substrata (all P P 0.56).

Fig. 6. Comparison of beating cardiomyocyte formation over 14 days

of observation using EBs from mES cells cultured on MatrigelTM (P1,

open squares; P7, open triangles) or 0.1% gelatin (P1, closed squares;

P7, closed triangles). Percent beating cardiomyocyte formation was not

significantly different for P1 cells grown on MatrigelTM or gelatin

substrata (P = 0.26). However, MatrigelTM supported significantly

higher cardiomyocyte formation when substrata were compared at

P7 (P = 0.01).

394 A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397

marker of undifferentiated mES cells (Solter and

Knowles, 1978). As shown in Fig. 4, expression of

SSEA-1 in mES cells grown on MatrigelTM or gelatin

to P5 was not significantly different for the two treat-

ments (66.7 ± 2.7% vs 65.5 ± 4.3%, respectively, P =

0.64). Mean percent SSEA-1 expression by positive

control cells (thawed aliquots of P4 mES cells previouslycharacterized for SSEA-1) was 65.6 ± 6.3% which was

not significantly different from results of mES cells

grown on either of the two substrata (both comparisons

P P 0.56).

3.4. Expression of activated caspase-3

Activated caspase-3 is a biomarker for the apoptoticcellular death pathway (Woo et al., 1998). As shown in

Fig. 5. Line graphs comparing percent activated caspase-3 expres-

sion ± SD for mES cells cultured on MatrigelTM or 0.1% gelatin

following 16 h incubation with 0, 5, 10, 50, 100 and 500 ppb sodium

arsenite. In comparison to mES cells grown on gelatin (closed squares),

overall expression of activated caspase-3 is significantly higher for mES

cells grown on MatrigelTM (open squares) (P < 0.0001). Data points

(n = 12) for cells grown on MatrigelTM represent triplicate observations

at each dose from four experiments. Data points for cells grown on

gelatin (n = 6) represent triplicate observations at each dose from two

experiments.

Fig. 5, mES cells cultured on MatrigelTM or gelatin to P5

were compared for expression of activated caspase-3 fol-

lowing a 16 h exposure to 0, 5, 10, 50, 100, and 500 ppb

sodium arsenite. In comparison to mES cells grown on

gelatin, expression of activated caspase-3 in response

to arsenite was significantly higher for mES cells grown

on MatrigelTM (P < 0.0001). These results suggest that

cells cultured to P5 on MatrigelTM were better able to

activate caspase-3 in response to increasing doses ofarsenite.

3.5. Differentiation assay

Percent development to beating cardiomyocytes is a

primary endpoint for the EST and measures stem cell

capacity for terminal differentiation. MatrigelTM effects

on mES cell differentiation were not known. Therefore,mES cells cultured on MatrigelTM or 0.1% gelatin were

harvested at P1 and P7 for comparisons of cardiomyo-

cyte differentiation. As shown in Fig. 6, percent develop-

ment to pulsating cardiomyocytes was not significantly

different for cells cultured on MatrigelTM or gelatin at

P1 (P = 0.26). However, at P7, significantly higher per-

centages of beating cardiomyocytes formed using popu-

lations of mES cells cultured on MatrigelTM (P = 0.01).

4. Discussion

EST is the only validated in vitro assay for develop-

mental toxicology that does not require pregnant ani-

mals (Spielmann et al., 1997). The test provides an

excellent starting point for creating automated, high-throughput screening systems for testing drugs and

chemicals. In addition to the classical endpoints of cyto-

A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397 395

toxicity and somatic differentiation, the assay now

includes quantitative assessment of gene expression

by flow cytofluorimetry (Seiler et al., 2004), reverse

transcription-polymerase chain reaction (RT-PCR)

(Schmidt et al., 2001) and DNA microarrays (Kelly

and Rizzino, 2000). Maintaining a homogeneous, well-characterized population of undifferentiated mES cells

is fundamental to these toxicity assays. Currently,

pluripotent mES cells for cytotoxicity studies are main-

tained without feeders by frequent subculturing on plas-

tic or gelatin-coated surfaces (Spielmann et al., 1997),

while pluripotent mES cells for terminal differentiation

studies are maintained on mouse fibroblast feeder layers

(Heuer et al., 1994). Our findings suggest that Matri-gelTM, without feeder cells, in combination with three

growth factors (LIF, hSCF and bFGF), may offer an

alternative strategy for unifying and standardizing

conditions for long-term culture of undifferentiated

mES cells.

The decision to evaluate MatrigelTM for feeder-free

propagation of mES cells was prompted by the findings

of Xu et al. (2001, 2002) that demonstrated humanembryonic stem (hES) cells grown on MatrigelTM with

conditioned medium maintained a normal karyotype,

stable proliferation rate, and pluripotential biomarker

profile for more than 200 population doublings. With-

drawal of conditioned medium and substratum resulted

in transformation of the stem cells to EBs and beat-

ing cardiomyocytes. Our comparison of MatrigelTM and

gelatin revealed no substratum differences regardingcell viability, population doubling times or expression

of SSEA-1 during the period of observation. However,

MatrigelTM supported significantly higher mES cell

alkaline phosphatase enzymatic activity, caspase-3 acti-

vation, and cardiomyocyte formation at the later pas-

sage. Morphologically, mES cells on MatrigelTM

retained an undifferentiated appearance with predomi-

nantly raised, compact, circular colonies. Subsequentstudies on the combined effects of MatrigelTM and growth

factors on long-term maintenance of undifferentiated

mES cells (119 population doublings) were recently re-

ported (Greenlee et al., 2004). Biomarker findings over

10 weeks of continuous culture suggested MatrigelTM

combined with growth factors sustained a population

of undifferentiated mES cells. However, increased aneu-

ploidy, reduced caspase-3 activation, and stem cellinability to terminally differentiate suggest further mod-

ifications to the culture system may be required.

Cell death is an important endpoint of the EST. Cells

engaged in apoptotic cell death express increased levels

of activated caspase-3 (Sarkar and Sharma, 2002). We

compared populations of mES cells grown on MatrigelTM

or gelatin for dose-responsive expression of acti-

vated caspase-3 following exposure to sodium arseniteand found that significantly higher levels of activated

caspase-3 were achieved in mES cells maintained on

MatrigelTM. Higher levels of activated caspase-3 expres-

sion may be associated with improved capacity for dis-

criminating toxicant effects. Laschinski et al. (1991)

noted that cytotoxic responses to teratogens varied be-

tween embryonal stem cells and adult fibroblasts with

undifferentiated embryonic stem cells showing increasedsensitivity to the toxicants. In our experiments, it is pos-

sible that mES cell populations on MatrigelTM were more

uniformly undifferentiated than those maintained on

gelatin and, therefore, at greater risk of succumbing to

the cytotoxic effects of arsenite. This may be an advan-

tage when using a substrate to identify embryotoxic

challenges. However, false positive responses may result

and require caution when interpreting data.The ability of mES cells to terminally differentiate is a

key prerequisite for in vitro studies investigating the

embryotoxic effects of xenobiotics. We, therefore, mea-

sured substratum effects on the capacity of mES cells

to transform into beating cardiomyocytes at P1 and

P7. Significantly higher percentages of beating cardio-

myocytes were obtained at P1 when compared to mES

cells harvested at P7, independent of substratum. Incontrast, significantly higher percentages of beating

cardiomyocytes were obtained at P7 with mES cells

propagated on MatrigelTM compared to gelatin. The

overall decline in capacity of mES cells to terminally dif-

ferentiate by P7 may be explained by extracellular matrix

components lacking in a feeder-free culture system. Nor-

mally, cells in developing tissues are surrounded by a

fiber-composite extracellular matrix that transmitsmechanical stimuli, maintains the shape of developing

tissues, and functions as a scaffold for cell migration

and attachment (Kadler, 2004). The fact that MatrigelTM

supported a higher proportion of mES cells with poten-

tial for somatic differentiation to P7 suggests that this

substratum provided some, but not all, components

needed for long-term, feeder-free maintenance of mES

cells. It will be important to determine if mES cells cul-tured in feeder-free conditions for longer periods of time

result in further declines in somatic differentiation.

We noted that mES cells cultured from P1 to P4 on

MatrigelTM or gelatin yielded metaphase spreads in

which only 35–45% contained 40 chromosomes, with

the remaining 55–65% spreads containing more or less

than the modal number. Doetschman et al. (1985) also

reported considerable aneuploidy in D3 mES cell lines,as up to 60% of metaphase spreads prepared from D3

subcultures contained more or less than 40 chromo-

somes. Karyotypic changes have been noted for hES

cells maintained for long periods of feeder-free continu-

ous culture. Gains in chromosomes 8 and 11 have been

noted in mouse cultures and gains in 17 and 12 have

been observed in fresh and frozen subcultures of hES

cells (Draper et al., 2004). Karyotypic changes may cor-relate with a reduced ability to colonize the germ line in

chimeric mice (Mitalipov et al., 1994; Longo et al., 1997)

396 A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397

or provide a selective advantage for the long-term prop-

agation of undifferentiated embryonic stem cells (Chiou

et al., 2003). Future comparisons may be well served by

earlier passages of embryonic stem cell lines or by clonal

derivatives selected for a normal karyotype (Amit et al.,

2000).Currently, the EST consists of two in vitro embryo-

toxic endpoints, cytotoxicity and percent formation of

beating cardiomyocytes. Additional efforts are needed

to validate the assay for purposes of human risk assess-

ment and to determine the relevance of the in vitro expo-

sures to in vivo pregnancy outcomes (NRC, 2000; Pryor

et al., 2000). Steps towards this objective might include

evaluating toxicant effects on mES cell differentiationinto cell types other than cardiomyocytes, e.g., neurons,

chondrocytes, or osteoblasts (Schmidt et al., 2001).

MatrigelTM is particularly well suited for this purpose,

as stem cells from rodents and humans grown on this

substratum will develop into various somatic lineages

by interchanging soluble signals (Asakura et al., 2001;

Ruhnke et al., 2003).

In summary, our findings suggest that MatrigelTM, incombination with growth factors, provides an alterna-

tive, unified approach for feeder-free propagation of

undifferentiated mES cells for embryotoxic endpoints.

Acknowledgments

The authors appreciate the assistance of Alice Star-gardt and Linda Weis for manuscript preparation,

Tom Brunette and Amy Wilhelmi for assistance with

graphic arts, and Tammy Ellis BS, Jim Burmester

Ph.D., and Deanna Cross Ph.D. for reviewing the manu-

script. The SSEA-1 monoclonal antibody was developed

by Solter and Knowles (1978) and was obtained from

the Developmental Studies Hybridoma Bank main-

tained by The University of Iowa, Department of Bio-logical Sciences, Iowa City, IA, under the auspices of

NICHD. The careful reading of this manuscript by un-

known reviewers is gratefully acknowledged.

Funding for this project was provided in part by a

grant from the Marshfield Clinic Research Foundation.

References

Ahmad, S.A., Sayed, M.H., Barua, S., Khan, M.H., Faruquee, M.H.,

Jalil, A., Hadi, S.A., Talukder, H.K., 2001. Arsenic in drinking

water and pregnancy outcomes. Environmental Health Perspec-

tives 109, 629–631.

Amit, M., Carpenter, M.K., Inokuma, M.S., Chiu, C.P., Harris, C.P.,

Waknitz, M.A., Itskovitz-Eldor, J., Thomson, J.A., 2000. Clonally

derived human embryonic stem cell lines maintain pluripotency

and proliferative potential for prolonged periods of culture.

Developmental Biology 227, 271–278.

Asakura, A., Komaki, M., Rudnicki, M.A., 2001. Muscle satellite cells

are multipotential stem cells that exhibit myogenic, osteogenic and

adipogenic differentiation. Differentiation 68, 245–253.

Chapin, R.E., Robbins, W.A., Schieve, L.A., Sweeney, A.M., Taba-

cova, S.A., Tomashek, K.M., 2004. Off to a good start: the

influence of pre- and periconceptional exposures, parental fertility,

and nutrition on children�s health. Environmental Health Perspec-

tives 112, 69–78.

Chiou, S.K., Jones, M.K., Tarnawski, A.S., 2003. Survivin-an anti-

apoptosis protein: its biological roles and implications for cancer

and beyond. Medical Science Monitor 9, P143–P147.

Congress of the United States Office of Technology Assessment.,

(1995). Screening and Testing Chemicals in Commerce. OTA-BP-

ENV-166. Office of Technology Assessment, Washington. Avail-

able at: www.wws.princeton.edu/cgi-bin/byteserv.prl/~ota/disk1/

1995/9553/9553.PDF. Accessed 27 April, 2004.

Doetschman, T.C., Eistetter, H., Katz, M., Schmidt, W., Kemler, R.,

1985. The in vitro development of blastocyst-derived embryonic

stem cell lines: formation of visceral yolk sac, blood islands and

myocardium. Journal of Embryology and Experimental Morphol-

ogy 87, 27–45.

Donovan, P.J., Stott, D., Cairns, L.A., Heasman, J., Wylie, C.C.,

1986. Migratory and postmigratory mouse primordial germ cells

behave differently in culture. Cell 44, 831–838.

Draper, J.S., Smith, K., Gokhale, P., Moore, H.D., Maltby, E.,

Johnson, J., Meisner, L., Zwaka, T.P., Thomson, J.A., Andrews,

P.W., 2004. Recurrent gain of chromosomes 17q and 12 in cultured

human embryonic stem cells. Nature Biotechnology 22, 53–54.

Geijsen, N., Horoschak, M., Kim, K., Gribnau, J., Eggan, K., Daley,

G.Q., 2004. Derivation of embryonic germ cells and male gametes

from embryonic stem cells. Nature 427, 148–154.

Greenlee, A.R., Kronenwetter-Koepel, T.A., Kaiser, S.J., Ellis, T.M.,

Liu, K., 2004. Combined effects of Matrigel and growth factors on

maintaining undifferentiated murine embryonic stem cells for

embryotoxicity testing. Toxicology in Vitro 18, 543–553.

Heuer, J., Bremer, S., Pohl, I., Spielmann, H., 1993. Development of

an in vitro embryotoxicity test using murine embryonic stem cell

cultures. Toxicology in Vitro 7, 551–556.

Heuer, J., Graeber, I., Pohl, I., Spielmann, H., 1994. Culture system

for the differentiation of murine embryonic stem cells—a new

approach to in vitro testing for embryotoxicity and for develop-

mental immunotoxicology. In: Fracchia, G.N. (Ed.), European

Medicines Research. IOS Press, Amsterdam, pp. 134–145.

Hubner, K., Fuhrmann, G., Christenson, L.K., Kehler, J., Reinbold,

R., De La Fuente, R., Wood, J., Strauss III, J.F., Boiani, M.,

Scholer, H.R., 2003. Derivation of oocytes from mouse embryonic

stem cells. Science 300, 1251–1256.

Kadler, K., 2004. Matrix loading: assembly of extracellular matrix

collagen fibrils during embryogenesis. Birth Defects Research Part

C Embryo Today 72, 1–11.

Kelly, D.L., Rizzino, A., 2000. DNA microarray analyses of genes

regulated during the differentiation of embryonic stem cells.

Molecular Reproduction and Development 56, 113–123.

Laschinski, G., Vogel, R., Spielmann, H., 1991. Cytotoxicity test using

blastocyst-derived euploid embryonal stem cells: a new approach to

in vitro teratogenesis screening. Reproductive Toxicology 5, 57–64.

Longo, L., Bygrave, A., Grosveld, F.G., Pandolfi, P.P., 1997. The

chromosome make-up of mouse embryonic stem cells is predictive

of somatic and germ cell chimaerism. Transgenic Research 6, 321–

328.

MacDonald, C., 1994. Primary culture and the establishment of cell

lines. In: Davis, J.M. (Ed.), Basic Cell Culture. A Practical

Approach. IRL Press, Oxford, pp. 160–161.

Matsui, Y., Zsebo, K., Hogan, B.L., 1992. Derivation of pluripotential

embryonic stem cells from murine primordial germ cells in culture.

Cell 70, 841–847.

A.R. Greenlee et al. / Toxicology in Vitro 19 (2005) 389–397 397

McAteer, J.A., Davis, J., 1994. Basic cell culture technique and the

maintenance of cell lines. In: Davis, J.M. (Ed.), Basic Cell Culture.

A Practical Approach. IRL Press, Oxford, pp. 124–126.

Mirkes, P.E., Little, S.A., 1998. Teratogen-induced cell death in

postimplantation mouse embryos: Differential tissue sensitivity and

hallmarks of apoptosis. Cell Death and Differentiation 5, 592–600.

Mitalipov, S.M., Mitalipova, M.M., Ivanov, V.I., 1994. Effect of long-

term culturing on the potential of mouse embryonic stem cells for

in vitro and in vivo development. Russian Journal of Develop-

mental Biology 25, 317–323.

National Research Council, 2000. Current practices for assessing risk

for developmental defects and their limitations. Scientific Frontiers

in Developmental Toxicology and Risk Assessment. National

Academy Press, Washington, DC, pp. 26–57.

Ohta, H., Wakayama, T., Nishimune, Y., 2004. Commitment of fetal

male germ cells to spermatogonial stem cells during mouse

embryonic development. Biology of Reproduction 70, 1286–1291.

Pease, S., Braghetta, P., Gearing, D., Grail, D., Williams, R.L., 1990.

Isolation of embryonic stem (ES) cells in media supplemented with

recombinant leukemia inhibitory factor (LIF). Developmental

Biology 141, 344–352.

Pryor, J.L., Hughes, C., Foster, W., Hales, B.F., Robaire, B., 2000.

Critical windows of exposure for children�s health: the reproductivesystem in animals and humans. Environmental Health Perspectives

108, 491–503.

Rohwedel, J., Guan, K., Hegert, C., Wobus, A.M., 2001. Embryonic

stem cells as an in vitro model for mutagenicity, cytotoxicity and

embryotoxicity studies: present state and future prospects. Toxi-

cology in Vitro 15, 741–753.

Ruhnke, M., Ungefroren, H., Zehle, G., Bader, M., Kremer, B.,

Fandrich, F., 2003. Long-term culture and differentiation of rat

embryonic stem cell-like cells into neuronal, glial, endothelial and

hepatic lineages. Stem Cells 21, 428–436.

Sarkar, S.A., Sharma, R.P., 2002. Expression of selected apoptosis

related genes, MIF, IGIF, and TNF alpha, during retinoic acid-

induced neural differentiation in murine embryonic stem cells. Cell

Structure Function 27, 99–107.

Schmidt, M.M., Guan, K., Wobus, A.M., 2001. Lithium influences

differentiation and tissue-specific gene expression of mouse embry-

onic stem (ES) cells in vitro. International Journal of Develop-

mental Biology 45, 421–429.

Scholz, G., Pohl, I., Genschow, E., Klemm, M., Spielmann, H., 1999.

Embryotoxicity screening using embryonic stem cells in vitro:

correlation to in vivo teratogenicity. Cells Tissues Organs 165, 203–

211.

Seiler, A., Visan, A., Buesen, R., Genschow, E., Spielmann, H., 2004.

Improvement of an in vitro stem cell assay for developmental

toxicity: the use of molecular endpoints in the embryonic stem cell

test. Reproductive Toxicology 18, 231–240.

Solter, D., Knowles, B.B., 1978. Monoclonal antibody defining a

stage-specific mouse embryonic antigen (SSEA-1). Proceedings of

the National Academy of Sciences USA 75, 5565–5569.

Spielmann, H., Pohl, I., Doring, B., Liebsch, M., Moldenhauer, F.,

1997. The embryonic stem cell test, an in vitro embryotoxicity test

using two permanent mouse cell lines: 3T3 fibroblasts and

embryonic stem cells. In Vitro Toxicology 10, 119–127.

Spielmann, H., Genschow, E., Scholz, G., Brown, N.A., Piersma,

A.H., Brady, M., Clemann, N., Huuskonen, H., Paillard, F.,

Bremer, S., Becker, K., 2001. Preliminary results of the ECVAM

validation study on three in vitro embryotoxicity tests. ATLA 29,

301–303.

Verbeke, G., Molenberghs, G., 2000. In: Bickel, P., Diggle, P.,

Fienberg, S., Krickeberg, K., Olkin, I., Wermuth, N., Zeger, S.

(Eds.), Linear Mixed Models for Longitudinal Data. Springer-

Verlag, New York, pp. 23–24.

Vogel, R., 1993. In vitro approach to fertility research: genotoxicity

tests on primordial germ cells and embryonic stem cells. Repro-

ductive Toxicology 7, 69–73.

Williams, R.L., Hilton, D.J., Pease, S., Wilson, T.A., Stewart, C.L.,

Gearing, D.P., Wagner, E.F., Metcalf, D., Nicola, N.A., Gough,

N.M., 1988. Myeloid leukaemia inhibitory factor maintains the

developmental potential of embryonic stem cells. Nature 336, 684–

687.

Wobus, A.M., Wallukat, G., Hescheler, J., 1991. Pluripotent

mouse embryonic stem cells are able to differentiate into cardio-

myocytes expressing chronotropic responses to adrenergic and

cholinergic agents and Ca2+ channel blockers. Differentiation 48,

173–182.

Woo, M., Hakem, R., Soengas, M.S., Duncan, G.S., Shahinian, A.,

Kagi, D., Hakem, A., McCurrach, M., Khoo, W., Kaufman, S.A.,

Senaldi, G., Howard, T., Lowe, S.W., Mak, T.W., 1998. Essential

contribution of caspase-3/CPP32 to apoptosis and its associated

nuclear changes. Genes and Development 12, 806–819.

Xu, C., Inokuma, M.S., Denham, J., Golds, K., Kundu, P., Gold,

J.D., Carpenter, M.K., 2001. Feeder-free growth of undifferenti-

ated human embryonic stem cells. Nature Biotechnology 19, 971–

974.

Xu, C., Police, S., Rao, N., Carpenter, M.K., 2002. Characterization

and enrichment of cardiomyocytes derived from human embryonic

stem cells. Circulation Research 91, 501–508.