Biochemical and morphological characterization of growth and differentiation of normal human neonatal keratinocytes in a serum-free medium

JOURNAL OF CELLULAR PHYSIOLOGY 134:229-237 (1988)

Biochemical and Morphological Characterization of Growth and Differentiation of Normal Human Neonatal Keratinocytes in a

Serum-Free Medium SREEKUMAR PILLAI*, DANIEL D. BIKLE, MARA HINCENBERCS, AND PETER M. ELIAS Endocrine Section and Medicine and Dermatology Services, Veterans Administration

Medical Center, San Francisco, California 94121

Growth and differentiation of keratinocytes in a serum-free medium (keratinocyte growth medium or KGM) was studied and compared to that under conditions in which serum and feeder cell layers were used. Cells were grown in KGM containing 0.1 m M calcium (KGM/low calcium), KGM containing 1.2 m M calcium (KCMlnormal calcium), or Dulbecco’s modified Eagles medium containing 5% fetal calf serum and 1.8 m M calcium in presence of mitomycin treated 3T3 M cells (DMEM/5% FCS). Plating efficiency and rate of growth were similar in the three media till confluence. In postconfluent cultures, protein and DNA content of cells attached to the plate in KGMllow-calcium dishes decreased as an increased number of cells were shed into the medium. Cell shedding was much less evident in the presence of normal calcium. Cells grown in KGM/low calcium had a higher rate of cell proliferation (3H-thymidine incorporation into cellular DNA) than cells grown in normal calcium. Transglutaminase activity, involucrin content, and cornified envelope formation were greatest in cells grown in KGMInormal calcium, intermediate in cells grown in DMEM/5% FCS, and least in cells grown in KGMllow calcium. Keratin profiles from cells grown in KGM/low calcium showed a lower percentage of high molecular weight bands compared to the keratin profiles from cells grown in the presence of normal calcium. Keratinocytes in KGMI low calcium grew as a monolayer of cuboidal cells with few features of differentiation, whereas cells grown in KGMlnormal calcium stratified into multilayered islands (3-5 layers) surmounted by 2-4 layers of enucleated cells with thickened cornified envelopes. Cells grown in KGMlnormal calcium also contained tonofilaments and lamellar bodies unlike cells grown in KGM/low calcium. Cells grown in DMEM/5% FCS also formed stratified layers comparable to cells grown in KGMlnormal calcium but lacked cornified cells, keratohyalin granules, tonofilament bundles, and lamellar bodies. These studies indicate the usefulness of serum-free conditions for the culture of human keratinocytes and confirm the importance of extracellular calcium in keratinocyte differentiation.

Keratinocytes in vi t ro simulate the growth and differentiation of the epidermis w i t h considerable fidelity (Green, 1977; Hennings et al., 1980), making these cells a useful model for the study of growth and differentiation. Both human and murine epidermal cells have been shown to differentiate f rom a basal, proliferating layer in to a flattened, squamous, enucleated upper layer, which i s eventually sloughed off into the medium (Green, 1979). To study the role of specific factors on the growth and differentiation of these cells, keratinocytes should be grown under defined conditions. However, most of cell culture conditions currently in use employ serum in the media and require co-cultivation with fibroblasts or feeder cell layers for normal growth (Rheinwald and

Green, 1975). Because of our interest in the role of cytokines and

calcium in the differentiation of keratinocytes we have sought to optimize growth in more completely defined media. We compared keratinocyte growth in serum-free keratinocyte growth medium (KGM) containing either 0.1 or 1.2 mM calcium to that in Dulbecco’s modified Eagles medium containing 5% fetal calf serum (DMEIW 5% FCS) and 1.8 mM calcium, conditions comparable to those originally described by Rheinwald and Green

Received June 5, 1987; accepted October 23,1987. *To whom reprint requestdcorrespondence should be addressed.

01988 ALAN R. LISS, INC.

PILLAI ET AL. 230 (1975). Our results demonstrate that KGM is a suitable medium for studies of keratinocyte growth and differentiation.

MATERIALS AND METHODS Culture media and culture conditions

KGM was purchased from Clonetics Corporation, San Diego, CA. The basal medium is complete MCDB 153 (Boyce and Ham, 19831, with added concentrations of amino acids, as described by Pittelkow and Scott (1986). MCDB 153 is supplemented with epidermal growth factor (10 ng/ml), insulin (5 pglml), hydrocortisone (0.5 pM), phosphatidylethanolamine (0.1 mM), ethanolamine (0.2 mM), bovine pituitary extract (70 pg proteidml media) penicillidstreptomycin (100 pg/ml) fungizone (2.5 mg/ ml), and a variety of trace metals. The added amino acids were: isoleucine (7.5 x lop4 M), histidine (2.4 x lop4 M), methionine (9 x M), phenylalanine (9 x lop5 M), tryptophan (4.5 x MI, and tyrosine (7.5 x

M). For our studies, KGM was either supplemented with low calcium 0.1 mM) or normal calcium (1.2 mM). Human foreskin keratinocytes, grown in this medium, were compared to cells grown according to a modifica- tion of the original method of Rheinwald and Green (1975), in which keratinocytes are cultured on a feeder monolayer of mitomycin C-treated 3T3M cells (5 x lo2/ cm2) and grown in DMEM/5% FCS containing epidermal growth factor (10 ng/ml), cholera toxin (lop9 M), hydrocortisone (0.4 pg/ml), penicillidstreptomycin (100 pg/ml), and fungizone (2.5 pg/ml) (Williams et al., 1987). The calcium concentration of DMEM/5% FCS was measured by atomic adsorption spectrometry and was found to be 1.8 mM.

Human keratinocytes were isolated from newborn human foreskins by incubating with 0.25% trypsin over- night at 4°C. This procedure is selective for keratinocytes since fibroblasts require collagenase for opti- mal isolation. The cells were grown to confluence in DMEM/5% FCS and then passaged into the respective media for the experiments. The cells were seeded into 6- well multiwell plates a t a density of 6 x lo3 cells/cm2 and grown for up to 22 days in KGMAow calcium, KGMI normal calcium, or DMEM/5% FCS. The cells were fed every second or third day with the respective media. Each experiment was performed using the same batch of cells from the same tissue sample. Each data point was the mean of triplicate culture dishes. The relative numbers of 3T3 cells, fibroblasts and keratinocytes were ascertained by fluorescence microscopy of cultures after treatment with 1 pg/ml of acridine orange. Acridine orange labeled the cytosol and nuclei of 3T3 cells and fibroblasts orange to red and the cytosol of keratinocytes green. Contamination of keratinocyte cultures by other cell types (fibroblasts, 3T3M cells) was less than 2% after 3 days of culture in DMEM/5% FCS and was 0% in cultures grown in KGM. Seeding efficiency was determined in triplicate as the percent of cells originally plated that were attached after 24 hr. To determine the number of attached cells the dishes were first treated with 0.1% trypsidO.Ol% EDTA for 5 min; the released cells were collected by centrifugation and subsequently counted with a hemocytometer.

Biochemical markers of growth and differentiation 3H-thymidine incorporation into cellular DNA was de-

termined by incubating the cells with 2 pCi [methyl, 1’ ,2’-3H]-thymidine (Amersham, specific activity 111 Ci/ mol) in 1 ml of the respective media for 2 hr and then quantitating the radioactivity in the TCA precipitate of washed cells.

DNA was assayed as described by Labarca and Pai- gen, using the fluorescent reagent, bisbenzimidazole (1980). Protein was measured by the BCA protien assay reagent, available from Pierce Chemical Company (Pub- lication 23225, 1984).

Cornified envelope formation, a marker for differentiated keratinocytes, was determined by the method described by Sun and Green (1976). Cells from each well were dissolved in 1% SDS, 20 mM DTT, sonicated lightly to reduce viscosity, and the optical density of the turbid solution determined at 340 nm.

Transglutaminase activity was determined by the method of Schmidt et al. (1985). Cells from one 10-cm dish a t different periods of growth were homogenized by sonication in Tris-HC1 buffer, pH 8.0, containing 1 mM EDTA and then centrifuged at 600g for 10 min. The supernatant was used as the enzyme source; 100 p1 of the supernatant was incubated with 600 p1 of 50 mM tris-HC1 buffer, pH 8.0, containing 10 mM calcium chlo- ride, 5 mM DTT, 540 pg dimethyl casein, 1 mM putrescine, and 2.5 uCi 3H-putrescine. The reaction was stopped by addition of 600 pl 10% TCA, and the protein pellet was washed three times with 5% TCA containing 10 mM putrescine. Final washing was done in 95% ethanol. The washed pellet was solubilized in 1 N NaOH, and the radioactivity was determined by scintillation spectroscopy. Blanks were run in parallel by the addition of TCA at the beginning of the incubation. The specific activity was calculated as cpm putrescine incor- porated into caseidpg DNA.

The content of involucrin and other endogenous substrates of cornified envelope was assessed by a procedure similar to the transglutaminase assay, but in which the exogenous substrate, dimethylcasein, was omitted from the incubation mixture (Rice and Green, 1979). The ra- tionale for this method is based on the fact that labelling of endogenous substrates occurs by the transfer of 3H- putrescine to these substrates by the endogenous transglutaminase present in the extract. In the absence of exogenous substrate the endogenous transglutaminase is in excess such that the endogenous substrate (e.g., involucrin) is rate limiting. This method is not specific for involucrin since at least 6 other proteins are labelled by this method (Simon and Green, 1984). Nevertheless, involucrin is the most abundant component of the cross- linked envelope, and the other proteins may function as anchor to the membrane either directly or by acting as a linker between involucrin and other integral membrane proteins (Simon and Green, 1984).

Keratins from 12- and 20-day-old cultures were extracted according to the procedure of Fuchs and Green (1981) as follows. The plates were washed with PBS; the cells were harvested by scraping into a tris-HC1 buffer, pH 7.4, containing 1 mM EDTA; the cells were sonically disrupted and centrifuged at 12,OOOg; the resulting keratin-containing pellet was incubated at 37°C for 20 min in a Tris-HC1 buffer, pH 6.8, containing 1 mM EDTA, 10 mM DTT, and 2% SDS; the samples then were sonically

GROWTH AND DIFFERENTIATION OF NEONATAL KERATINOCYTES

= 0) . - P 2.0-

P - c a2 .- c

e a 1.0-

231

disrupted, heated at 100°C for 2 min, and centrifuged at 12,OOOg for 20 min to remove the insoluble residue. Aliquots of cellular extract (10 pg proteidane) were electrophoresed through a 10% SDS-polyacrylamide gel using the Laemmli system in a mini Protean I1 dual slab cell system (Bio-Rad, Richmond, CA). Staining was carried out in a solution containing Coomassie blue (0.125%), methanol (50%), and acetic acid (10%). Molec- ular weights of the bands were estimated by comparison with SDS molecular-weight markers (Sigma Chemical Co., St. Louis, MO), run on the same gel. The relative proportion of the different keratin bands was determined by comparing the areas under the curve of a densitometric scan of the gels.

Statistical analyses Multivariable analyses of variance were performed

where appropriate using the software package SYSTAT (Systat, Inc., Evanston, IL).

Morphological markers for differentiation Twenty-day-old cultures were washed two times with

PBS and once with 0.1 M cacodylate buffer. The cultures were fixed in situ with 2 ml glutaraldehyde solution for 2 hr at room temperature. Cultures were then washed twice with cacodylate buffer, postfixed in 1% aqueous osmium tetroxide, dehydrated in graded ethanols, and embedded in an Epon-epoxy mixture for light and electron microscopy (McNutt and Crain, 1981). One-half mi- crometers sections, stained with l% toluidine blue, were studied by light microscopy. Ultrathin sections, addi- tionally stained with lead citrate and uranyl acetate, were examined in a Zeiss 10A electron microscope at 50 kV.

RESULTS Growth characteristics

Second passage keratinocytes, plated at a density of 6 x 103/cm2 into six-well multiwell dishes, were grown in either KGM/low calcium, KGMInormal calcium, or DMEM/5% FCS (the latter plated on 3T3-M cells). The seeding efficiency of the cultured keratinocytes was the same in the three different media (i.e., - 80%). Likewise, the cells reached visual confluence at the same time (day 7) in all three media. However, differences soon became apparent. Figure 1A compares the protein and 1B the DNA content of keratinocytes grown in the different media. Keratinocytes continued to grow postcon- fluence in all three media until day 12, at which point no further increase in DNA was observed. Cultures grown in KGlWlow calcium showed a progressive decrease in the number of attached cells and DNA content after day 12. Whereas the protein content of cultures grown in KGMInormal calcium and DMEM/5% FCS increased up to day 20, it fell in cultures grown in KGMI low calcium, paralleling the decrease in DNA content. Thus, the ratio of proteinDNA remained constant in the cultures grown in KGM/low calcium, while the ratio increased 1.5-fold in cultures grown in KGMInormal calcium and 1.3-fold in cultures grown in DMEM/5% FCS (Table 1).

3H-thymidine incorporation into cellular DNA was studied on day 12, the point a t which cell numbers began to stabilize. Cells grown in KGMAow calcium exhibited the greatest rate of proliferation with a 3-fold higher

3.0 1

I I I l l I I l l

3 5 8 10 12 15 18 20 22

A Days in Culture

3 5 8 10 12 15 18 20 22

B Days in Culture

Fig. 1. Effect of different media on keratinocyte proliferation. Kerati- nocytes were grown for up to 22 days in KGMAow calcium (O), KGW normal calcium (0) or DMEW5% FCS (A). At different days, cells from triplicate wells were harvested and assayed for protein (A) and DNA (J3) content. Data are plotted as mean values + SD. Protein and DNA significantly (p < 0.001) increased in all media with time in culture. These increases were significantly (P < 0.001) greater in cells grown in the KGWnormal calcium and DMEM/5% FCS than in cells grown in the KGMflowcalcium medium.

TABLE 1. ProteinDNA ratio of keratinocytes grown in different medial

ProteinDNA ratio Days in KGM KGM DMEW5% culture low Ca normal Ca FCS

5 42.1 k 0.92 35.0 f 3.11 35.2 k 2.0 8 47.6 k 0.28 39.3 & 0 39.2 k 4.2

10 42.9 & 0.92 40.0 k 2.83 40.5 * 1.2 15 41.3 & 4.2 41.3 & 2.2 43.1 f 1.9 18 44.3 * 1.76 43.1 * 3.46 40.5 + 3.8 20 48.0 k 2.5 50.0 + 6.5 44.3 * 12.0 22 45.3 rt 0.84 54.5 + 5.8 47.0 * 4.8 'The ratio of proteinDNA was calculated from data presented in Figures 1A and 1B. The values are expressed as mean & SD of triplicate dishes. The change in ratio of proteinDNA with time was significant (P=O.O19) in cells grown in KGWnormal calcium and was significantly greater (P=O.O28) than the change in ratio observed in cells grown in KGhUlow calcium.

232 PILLAI ET AL.

TABLE 2. Biochemical markers of keratinocyte growth and differentiation'

12-day cultures 20-day cultures DMEM/FCS KGMnC KGM/NC DMEM/FCS KGMnC KGM/NC

3H-thymidine incorporation (cpdpg DNA)

Protein (mg/dish)

DNA (pg/dish)

Transglutaminase (cpm putrescinel pg DNA)

Involucrin (cpm putrescine/ pg DNA)

Cornified envelope (OD34dpg DNA x

Keratins (high MWIlow MW)

Percent of attached cells

378.6 + 28.5 108.4 + 11.5 186.5 + 28.5 nd nd nd

1.38 rf: 0.014

45.9 + 0.55

298.5 * 14.8

1.71 * 0.106

61.2 * 0.15

1250.5 + 101

1.82 rf: 0.113

59.1 k 2.26

531.5 + 17.6

1.41 + 0.00

29.32 + 3.50

97.5 rf: 11.4

3.07 + 0.396

61.00 + 0.05

409.3 rf: 17.9

2.79 + 0.34

62.99 + 9.03

130.5 + 14.8

nd nd 5.06 38.5 29.0 nd

2.60

2.58

75

2.85

3.36

97

2.96

3.80

95

8.20

3.00

46

12.56

4.00

95

6.45

3.34

94

'Biochemical markers were determined in triplicate wells and the values were expressed as mean +SD. Involucrin, cornified envelope, and keratins were determined in pooled samples from triplicated dishes. nd, not determined.

1 3 1 7- rate of DNA synthesis than cells grown in KGWnormal calcium and a 2-fold higher rate than cells grown in DMEM/5% FCS (Table 2).

When the number of attached versus unattached cells in the dishes was quantitated on day 12, cultures grown in KGM/low calcium showed a higher percentage of unattached cells than did cultures grown in either KGW normal calcium or DMEW5% FCS (25%, 3%, and 5% for KGMAow calcium, KGWnormal calcium, and DMEM/ 5% FCS, respectively) (Table 2). On day 20 more than 50% of cells were unattached in cultures grown in KGW low calcium, whereas only 5 and 6% of the cells were unattached in the cultures grown in KGWnormal calcium and DMEW5% FCS, respectively (Table 2). These measurements reflect only the number of cells that had been shed during the preceding 48 hours, the time at which the medium was last changed. Thus, the apparent paradox of the higher rate of DNA synthesis but lower DNA content of the cultures grown in KGMAow calcium is resolved by the extensive shedding of cells into the medium under these conditions. The unattached cells in the KGWnormal calcium and DMEW5% FCS cultures were flat, large, enucleated squamous cells. The unattached cells in the KGMAow calcium cultures were smaller, cuboidal cells that did not differ in morphology from the attached cells. More than 95% of the unattached cells in KGMAow calcium excluded trypan blue, incorported 3H-thymidine into cellular DNA in comparable amounts to the attached cells (data not shown), and when plated into fresh KGWnormal calcium readily attached to the culture dishes.

Differentiation markers Various markers of keratinocyte differentiation were

measured in cells grown for varying periods of time in the three different media (Table 2, Fig. 3). Table 2 sum- marizes the changes in biochemical markers of keratinocyte growth and differentiation measured at the time when the cultures are post confluent and spontaneously differentiating (day 12 and day 20). Cornified envelope

,+-r- 12 15 18 20 22

A Days in culture

1 2 5 0 i

:,il I

I I l l 1 1 1 1 3 5 8 10 12 15 18 20 22

B Oays in Culture

Fig. 2. Effect of different media on differentiation of keratinocytes. Keratinocytes were grown in KGWIow calcium (01, KGWnormal calcium (O), or DMEM/5% FCS (A) for different lengths of time. On indicated days, cells from duplicate wells were harvested, sonicated, and the homogenate used for assays. The amount of cornified envelope (A) was determined in pooled duplicate wells. Transglutaminase activity (B) was determined in each of duplicate wells. KGMllow calcium (0), KGWnormal calcium (O), and DMEM/5% FCS (A).

GROWTH AND DIFFERENTIATION OF NEONATAL KERATINOCYTES 233

Fig. 3. Keratin profiles of keratinocytes grown in the different media; 20-day-old-keratinocytes were harvested and the keratins extracted as

formation, a marker of terminal keratinocyte differentiation, was increased two-fold by day 20 in cells grown in KGWnormal calcium compared to cells grown in DMEM/5% FCS (Table 2, Fig. 2A). Keratinocytes grown in KGM/low calcium produced nearly the same amount of cornified envelope as cells grown in DMEW5% FCS medium. The activity of transglutaminase, the enzyme which crosslinks the envelope precursor protein, involucrin, was maximal on day 12 in cells grown in all three media, and then decreased in postconfluent cultures (Ta- ble 2, Fig. 2B). At 12 days the activity of this enzyme was approximately 5-fold and 2.2-fold higher in cells grown in KGWnormal calcium than in cells grown in KGM/low calcium and DMEW5% FCS, respectively. The involucrin content of these cells was also maximal on day 12 (time course not shown), with the highest levels present in cells grown in KGWnormal calcium (Table 2). Involucrin levels in cells grown in DMEW5% FCS were 76% of those in cells grown in KGWnormal calcium, whereas involucrin levels in cells grown in KGW low calcium were only 14% of those in cells grown in KGWnormal calcium (Table 2).

The keratin profile of 12-day- and 20-day-old cells contained four distinct bands with molecular weights of 58 Kd, 54 Kd, 50 Kd, and 48 Kd. Figure 3 shows the keratin profile of 20-day-old cells grown in the three different media. In cells grown for 20 days in KGW

described in Materials and Methods. The protein profiles on 10% SDS- PAGE gels are shown.

normal calcium the ratio of high molecular weight keratins (58 Kd and 54 Kd) to low molecular weight keratins (50 Kd and 48 Kd) as generated by densitometry of the bands was 3.00, a ratio 33% and 20% higher than that from cells grown in KGM/low calcium and DMEW 5% FCS, respectively (Table 2).

Ultrastructural observation Microscopic observations of cells grown in the three

different media were carried out on day 20. Keratino- cytes grown in KGM/low calcium (Fig. 4) formed a monolayer of cuboidal cells, with gaps separating individual cells. Where cells contacted their neighbors, they did not form well developed intercellular junctions, even in sites of close apposition. Occasionally, rounded cells could be seen positioned above the monolayer, poised as if they were about to slough into the medium. Neither basally attached nor exfoliating cells exhibited signs of toxicity, and none displayed morphologic features of epidermal differentiation; keratohyalin granules, tonofilament bundles, lamellar bodies, and cornified envelopes were lacking in these cells.

Cells grown in KGWnormal calcium (Fig. 5) stratified into multilayered islands. The total number of nucleated cell layers ranged from three-to-five, never exceeding seven. Surmounting the nucleated layers were two-to- four geometrically packed layers of enucleated cells with

234 PILLAI ET AL.

Fig. 4. Human foreskin keratinwytes grown in KGMflow calcium. Cells grew as a monolayer and desquamated individually into the culture medium (A). Even where adjacent cells made contact (arrows),

well formed desmosomes and other membrane junctional structures were not observed. (A) x250; (B) ~32,000.

Fig. 5. Human foreskin keratinocytes grown in KGWnormal calcium. Cells stratified and several layers of cornified cells (c) were present (A, arrows and €4). In intermediate cell layers (C) multiple lamellar bodies can be seen in various stages of assembly (C, mows).

These organelles possess a small radius of curvature and an internal lamellar structure comparable to these organelles in vivo @,El. (A) x 1,000; B ~25,000; (C) ~30,000; (D and E) ~45,000.


Fig. 6. Human foreskin keratinocytes grown in DMEM/5% FCS. Cells stratified, but very little cornification was apparent (A). Three popula- tions of cells were noticed in these cultures; basaloid (B), intermediate (I), and apical (A). The apical cells were surmounted by microvilli, but

were not cornified, although variable amounts of cornified cells are present in the medium in post confluent cultures (not shown), (A) X 1,000; (B) X22,500.22.

thickened cornified envelopes. Nucleated cells contained tonofilament bundles but did not contain keratohyalin granules. In addition, suprabasally positioned nucleated cells contained abundant elements of smooth endo- plasmic reticulum associated with organelles that clearly resembled lamellar bodies found in intact epidermis. Although only a few of these organelles displayed the typical packed disks of lamellar bodies, most pos- sessed some internal lamellar structures, and all were of the correct size (0.15-0.25 pm). Despite the occasional occurrence of large numbers of these organelles, there was no evidence of secretion of lamellar contents into apical intercellular domains, as occurs in epidermis.

Keratinocytes, grown in DMEW5% FCS (Fig. 6) also

formed stratified islands, but certain noteworthy features distinguished these cultures from the KGWnor- ma1 calcium cultures: 1) cornified cells were not routinely observed in the DMEW5% FCS cultures, and when present they appeared as isolated, loosely attached cells above the most apical nucleated cell layer; 2) the keratinocytes in the DMEM/5% FCS cultures lacked keratohyalin granules, tonofilament bundles, and lamellar bodies.

DISCUSSION The most extensively used system for the study of

human keratinocytes in culture utilizes serum-supplemented medium supported by a nonreplicating 3T3

236 PILLAI ET AL.

feeder cell layer (Rheinwald and Green, 1975). Cells grown in this fashion can be clonally passaged and differentiate with many morphological and biochemical features of epidermis. However, this system is not well defined because of the requirement for serum and the 3T3 cells. Consequently, in recent years, several systems have been developed to grow human keratinocytes in serum-free media without the use of feeder cells. Boyce and Ham (1983), using a nutrient-supplemented medium, MCDB 153, showed that cells grown at 0.03 mM Ca concentration proliferated rapidly, remaining as “prokeratinocytes”. When 1.0 mM calcium was added, the cells appeared to differentiate. Further increases in the calcium concentration to 2.0 mM did not show any appreciable increase in the differentiation of these cells. Immunofluoresence studies revealed increased intermediate filament bundles and filaggrin protein levels with the increase in calcium concentration. Wille et al. (1984) used the same medium to study the influence of calcium and a variety of growth factors such as EGF, insulin, and bovine pituitary extract on keratinocyte growth kinetics. In more recent studies, these investigators used this medium in a two-phase culture system to culture human keratinocytes in vitro for autografting of the denuded epidermis of burn victims (Pittelkow and Scott, 1986). Low calcium medium was used in phase 1 to cultivate proliferative keratinocytes which were then transferred to a serum-containing medium to induce the formation of cohesive, stratified sheets of differentiated keratinocytes resembling normal epidermis that could then be used for grafting.

Extracellular calcium concentrations markedly affect the proliferation and differentiation of epidermal cells (Hennings et al., 1980). The precise molecular mecha- nisms involved in the calcium regulated cellular changes have not been elucidated. The role of calcium in activation of several enzyme systems such as protein kinases (Schulman and Greengard, 1978), phosphodiesterases (Schulman and Greengard, 19781, adenylate cyclases (Cheung et al., 1978), and phospholipases (Daniel, 1985) are well established. These enzymes may be involved in cell proliferation and differentiation. Other calcium re- quiring enzymes such as thymidylate synthetase (Whit- field et al., 1976) and transglutaminases (Buxman and Wuepper, 1976) are clearly involved in cellular proliferation and differentiation. The latter enzyme, transglutaminase, is essential for the formation of (y-glutamyl)E- lysine cross links during cornified envelope development.

Few studies have been done to evaluate the biochemical events underlying calcium induced differentiation of keratinocytes. As a first step in this direction we have compared the proliferation and differentiation of cells grown in serum-free media under low and normal calcium conditions to that of cells grown in the presence of serum, feeder cells, and calcium. We quantitated different biochemical markers of differentiation and assessed the morphological characteristics of differentiation. We observed that basaloid cells in 0.1 mM calcium main- tained a higher rate of proliferation than did cells in 1.2 mM calcium. In contrast, in the presence of 1.2 mM calcium the cells developed higher levels of involucrin, transglutaminase and cornified envelopes. These results correlated with morphological observations in that cells grown in 1.2 mM calcium showed more stratification and had a higher number of cornified cells with kerato-

hyalin granules, tonofilaments, and lamellar bodies than did cells grown in 0.1 mM calcium. Cells grown in the serum containing medium also stratified to a comparable degree but were less differentiated by morphologic and biochemical criteria than were cells grown in KGM with 1.2 mM calcium.

Our data demonstrate the usefulness of KGM in the study of keratinocyte growth and differentiation. They confirm the role of calcium in differentiation and reveal that keratinocytes grown under serum free conditions with 1.2 mM calcium differentiate more completely than those grown by more traditional methods. The observation that cells grown in serum containing medium with feeder cells do not differentiate as completely as cells grown in the absence of serum and feeder cells suggests that serum andor feeder cells may contain or produce a substance(s), which retards differentiation. Factors present in the serum such as vitamin A (Fuchs and Green, 19811, 1,25(OH)zD3 (Hosomi et al., 1983) and TGF-B (Bertolero et al., 1986) are all known to modulate the growth and differentation of kertinocytes. This conclu- sion must be tempered by the fact that these media differed in constituents other than serum. Nevertheless, KGM should prove useful in studies involving cytokines, growth factors, hormones, and other substances that may be present in serum or elaborated by feeder cells when the specific effects of these agents on keratinocyte growth and differentiation are to be determined.

ACKNOWLEDGMENTS The authors wish to thank Dr. Grayson and Ms. D.

Placzek for excellent technical assistance with electron microscopy, and Mr. B. Chapman for preparing the man- uscript. This study was supported by the Medical Re- search Service, Veterans Administration and the National Institute of Health AR 38386.

LITERATURE CITED Bertolero, F., Kaighan, M.E., Camalier, R.F., and Safiotti, U. (1986)

Effects of serum and serum derived growth factors on growth and differentiation of mouse keratinocytes. In Vitro Cell Dev. Biol., 22t423-428.

Boyce, S.T., and Ham, R.G. (1983) Calcium regulated differentiation of normal human epidermal keratinocytes in Ghemically-defined clonal cultures and serum-free serial cultures. J . Invest. Dermatol., (suppl).,

Buxman, N.M., and Wuepper, K.D. (1976) Isolation, purification and characterization of bovine epidermal transglutaminases. Biochem., Biophys. Acta, 452356-369.

Cheung, W.Y., Lynch, T.J., and Wallace, R.W. (1978) An endogenous calcium dependent activator protein of brain adenylate cyclase and cyclic nucleotide phosphodiesterase. Adv. Cyclic Nucl. Res., 9t233- 251.

Daniel, L.W. (1985) Phospholipases. In: Biochemistry of Arachidonic Acid Metabolism. W.E.M. Lands, ed. Martinus Nijhoff Publishing, Boston, pp. 175-189.

Fuchs, E., and Green, H. (1981) Regulation of terminal differentiation of cultured human keratinocytes by vitamin A. Cell, 25:617-625.

Green, H. (1977) Terminal differentiation of cultured human epidermal cells. Cell, 11t405-416.

Green, H. (1979) The keratinocyte as a differentiated cell type. The Harvey Lecture Series, 74t101-139.

Hennings, H., Michael, D., Cheng, C., Steinart, P., Holbrook, K., and Yuspa, H. (1980) Calcium regulation of growth and differentiation of mouse epidermal cells in culture. Cell, 19:245-254.

Hosomi, J., Hosoi, J., Abe, E. Suda, T., and Kuroki, T. (1983) Regulation of terminal differentiation of cultured mouse epidermal cells by 1,25(OH),D. Endocrinology, 113:1950-1957.

81:33-40.


Labarca, C., and Paigen, K. (1980) A simple, rapid and sensitive DNA assay procedure. Anal, Biochem., 102:344-352.

McNutt, N.S., and Crain, W.L. (1981) Quantitative electron microscopic comparison of lymphatic nuclear contours in mycosis fun- goides and in benign inflitrates in skin. Cancer, 47:698-709.

Pierce Chemical Company (1985) BCS protein assay reagent. Publica- tion Number 23225.

Pittelkow, M.R, and Scott, R.E. (1986) New techniques for the in vitro culture of human skin keratinocytes and perspectives on their use for grafting of patients with extensive burns. Mayo Clin. Proc., 61 :771-777.

Rheinwald, J.G., and Green, H. (1975) Serial cultivation of strains of human epidermal keratinocytes: The formation of keratinizing colo- nies from single cells. Cell, 6:331-343.

Rice, R.H., and Green, H (1979) Presence in human epidermal cells of a soluble protein precursor of the cross linked envelope: Activation of the cross linking by calcium. Cell, 18:681-694.

Schmidt, R., Reichert, U., Michel, S., Shroot, B., and Bouclier, M. (1985) Plasma membrane transglutaminase and cornified envelope compe- tence in cultured human keratinocytes. FEBS letters, 186:201-204.

Schulman, H., and Greengard, P. (1978) Calcium dependent protein

phosphorylating system in membranes from various tissues, and its activation by “calcium dependent regulator.” Proc. Natl. Acad. Sci.

Simon, M., and Green, H. (1984) Participation of membrane-associated proteins in the formation of the cross-linked envelope of the keratinocyte. Cell, 36:827-834.

Sun, T.-T., and Green, H. (1976) Differentiation of the epidermal keratinocyte in cell culture: Formation of the cornified envelope. Cell, 9.511-521.

Wille, J.J., Pittelkow, M.R., Shipley, G.D., and Scott, R.E. (1984) Inte- grated control of growth and differentiation of normal human prokeratinocytes cultured in serum-free medium: Clonal analyses, growth, kinetics, and cell cycle studies. J. Cell. Physiol., 121:31-44.

Williams, M.L., Rutherford, S.L., Mommas-Kienhuis, A.M., Grayson, S., Vermeer, B.J., and Elias, P.M. (1988) Free sterol metabolism in LDL-receptor expression as differentiation markers in cultured human keratinocytes. J. Cell. Physiol., (in press).

Whitefield, J.F., Mac Manus, J.P., Rixon, R.H., Boynton, A.L., Youdale, T., and Swierenga, S., (1976) The postive control of cell proliferation by the interplay of calcium ions and cuclic nucleotide. In Vitro, 121- 18.

, USA, 75;5432-5436.

Documents

Biochemical and morphological characterization of growth and differentiation of normal human neonatal keratinocytes in a serum-free medium