Morphometric and statistical analyses describing the In utero growth of human epidermis.код для вставкиСкачать
THE ANATOMICAL RECORD 222201-206 (1988) Morphometric and Statistical Analyses Describing the In Utero Growth of Human Epidermis CAROLYN A. FOSTER, JOHN F. BERTRAM, AND KAREN A. HOLBROOK Department of Dermatology 1, Division of Immunobiology, University of Vienna Medical School, Vienna A-1090 Austria (C.A.F.);Department of Anatomy, University of Melbourne, Victoria 3052, Australia (J.F.B.)Departments of Biological Structure and Medicine (Dermatology); University of Washington, School of Medicine, Seattle, WA 98195 (K.A.H.) ABSTRACT Epidermal development of human embryonic and fetal skin from the lower limb was studied using morphometric and statistical methods. Epidermal growth, as defined by an increase in epidermal thickness and the number of cell layers, occurred in three distinct stages during the first and second trimesters. The first growth spurt occurred between 5 and 13 weeks estimated gestational age (EGA)and was followed by a plateau phase with little change in epidermal thickness from 14 to 21 weeks, after which the epidermis began to increase in height again. The periderm reached its maximal height by approximately 13 weeks EGA, and by 25 weeks was shed into the amniotic fluid. Thus, within a five-month period (5 to 25 weeks EGA) the epidermis changed from a single cell layer < 10 pm thick to a 10 to 12-celllayer, keratinized epithelium > 60 pm thick. In contrast, epidermis from adult lower limb consisted of about 25 cell layers and was almost 75 pm in thickness. The age-related differences in epidermal thickness probably reflect changes in cell size and shape more than changes in the directional movement (apically vs. laterally) of proliferating keratinocytes, because the addition of cell layers throughout development was relatively constant. During the plateau phase, when there is a rapid increase in fetal growth rate, the suprabasal keratinocytes become more flattened, thereby allowing for the addition of new cell layers while maintaining a relatively constant epidermal thickness. The loss of glycogen reserves, along with other intrinsic cellular changes, probably contributes to the flattening effect. However, the internal expansion of growing tissues also may exert a mechanical pressure that could stretch the skin passively and influence epidermal structure. The skin is not only the largest organ of the body but is also one of the first to form during development. Within 1 month after zygote formation the human embryo is covered by a bilayered anlage of skin consisting of a simple epithelium (epidermis) overlying a thin, hydrated mesenchyme (dermis). These ectodermal and mesodermal derivatives differentiate into a miniaturized replica of adult skin by 6 months gestation. Numerous investigators have documented the ontogeny of human skin by morphological and biochemical criteria (Moll et al., 1983; Dale et al., 1985; Smith et al., 1986; reviews by Breathnach, 1971; Holbrook, 1979; Holbrook and Hoff, 1984). However, there appears to be no quantitative information on dimension within the skin compartment during development. The density of certain epidermal cells (e.g., Langerhans cells) is usually referenced to a unit area of skin. Except for certain diseases affecting skin thickness (Haftek et al., 19831, this is considered to be an acceptable method of comparing cell numbers in adult skin. Epidermal cell numbers based on surface area can be similarly determined in embryonic and fetal skin, but without three-dimensional measurements these data could be misleading because the cutaneous compartment is continuously expanding and differentiating throughout gestation. 0 1988 ALAN R. LISS, INC. To better understand the Langerhans cell population in developing skin from a two- and three-dimensional perspective (Foster et al., 1986; Foster and Holbrook, in preparation), the thickness of human eipdermis was evaluated a t progressive stages of gestation. Because the surface contours of developing epidermis are highly variable (Holbrook and Odland, 1975), a morphometric (digitizing) technique was chosen to determine epidermal height, and the data were statistically analyzed. This study provides the first mathematical description, in the form of polynomial equations, of human epidermal growth in utero and suggests a pattern of coordinated events that culminate in the formation of a stratified, keratinized epithelium. MATERIALS AND METHODS Specimen Collection Human embryonic and fetal tissue was provided through the courtesy of Dr. Thomas Shepard, Director Received January 12, 1988; accepted March 7, 1988. Address reprint requests to Dr. Karen A. Holbrook, Department of Biological Structure SM-20, University of Washington, School of Medicine, Seattle, WA 98195. 202 C.A. FOSTER, J.F. BERTRAM, AND K.A. HOLBROOK of the Central Laboratory for Human Embryology at the University of Washington. Consent forms and collecting practices were approved by the University of Washington Human Subjects Review Committee. Tissue was obtained from the lower limbs of 38 human embryos and fetuses, ranging in age from 5 to 25 weeks. Estimated gestational age (EGA = fertilization age) was determined from patient histories and fetal measurements (crown-rumpand foot length). Specimens that were 10 weeks EGA or younger were designated as embryonic, while those older than 10 weeks were termed fetal. Skin from three healthy adult volunteers (30-35 yr) was obtained from the ventral thigh by punch biopsy. Tissue Processing Pieces of skin were immersed in half-strength Karnovsky’s fixative (Karnovsky, 1965) overnight at 4”C, rinsed in 0.1 M cacodylate buffer, then post-fixed for an additional hour in 1%OsO, in distilled water at room temperature. After a brief wash in water, the tissue was stained en bloc with 1%aqueous uranyl acetate for 1 hr, dehydrated in a graded series of ethanols into propylene oxide, and embedded in Epon 812 (Luft, 1961). Semi-thin (1 p m ) sections were stained with toluidine blue (Richardson et al., 1960). Ultrathin sections (800 A) were stained with aqueous uranyl acetate and lead citrate (Reynolds, 1963) and examined with a Philips 201 transmission electron microscope. area of the first sections cut from the block face. Linear compression was calculated as the ratio of these two areas. Two blocks analyzed from each specimen at each of three stages (6,10,20 weeks EGA) showed that there was no measurable compression of the sections. Enumeration of Epidermal Cell Layers The mean number of epidermal cell layers was counted in toluidine blue-stained, 1 pm-thick sections of Eponembedded skin from three lower limbs of 34 individuals (5 to 25 weeks EGA and three adults). Ultrathin sections of keratinized epidermis were examined in the transmission electron microscope to delineate the number of layers in the stratum corneum. Calculation of Epidermal Height To estimate epidermal thickness, approximately eight sections of skin per specimen, usually representing two blocks of tissue, were photographed under oil immersion at x 40 on 35 mm film (Panatomic-X) with a Zeiss photomicroscope. Successive frames of film were projected onto the digitizer tablet of a microcomputer-based morphometric facility (Sundsten and Prothero, 1983), and the contours of a constant length of interfollicular epidermis, with and without the peridermal layer, were traced manually with a “mouse.” The coordinates of each contour were stored on floppy disks in a North Star Horizon microcomputer and cross-sectional areas were computed using a software package (MORPHO) Assays for Shrinkage and Compression for morphometric data (Prothero and Prothero, 1982). Epidermal height was determined by dividing the Experiments were conducted to determine if tissue processing and sectioning might affect the epidermal area of each contour by 0.24 mm (the constant length dimensions being measured. As described elsewhere of epidermis measured in each frame of 35 mm film) standard deviation (Foster, 1987), embryonic and fetal skin was removed and calculating the mean value from the lower limb and divided into two groups: 1) for each specimen. whole skin with subcutaneous tissue removed, and 2) Applied Regression Analysis skin with most of the dermis peeled away. Eight pieces The relationship between age and epidermal dimenof skin per group were cut into square or rectangular pieces (3-4 mm2) and measured en face, both before sions was determined using applied regression analysis and after fixation in Karnovsky’s fixative, then again (Kleinbaum and Kupper, 1978). Least-squares analysis after polymerization in Epon. Changes in linear di- was used to minimize the deviation between the obmensions due to tissue processing were calculated from served and fitted points and, thus, to describe the bestthe following equation: postembedment diameterlpre- fitting curve mathematically. The appropriate order fixation diameter x 100 = % of original linear dimen- (first, second, or third) for the polynomial model was determined by calculating the test statistic (F = MS sion. Although it is known that tissue processing through regressiodMS residual) and testing the null hypothesis primary fixation and Epon embedment causes differ- H, = no significant lack of fit of the assumed model. ential, age-dependent shrinkage of skin during ontogRESULTS eny (Foster, 1987), it was difficult to correct for these Morphometric and Statistical Measurements of Epidermal effects on measurements of epidermal height without Thickness the risk of grossly distorting the data. For instance, the Epidermal growth in developing human skin, as dedermis had a profound influence on differential shrinkage in Karnovsky-fixed skin, as determined by changes fined by the rate of increase (i.e., slope) in epidermal in the linear dimensions of skin pieces that were meas- height, was measured by two approaches: calculating ured en face; however, it was technically impractical to epidermal thickness utilizing a morphometric, digitizmeasure the depth of skin in these assays. In contrast, ing technique and counting the number of cell layers. postfxation processing through Epon accounted for about Within a 5-month period (5 to 25 weeks EGA) the ep4% of the total net shrinkage (Foster, unpublished ob- idermis changed from a single cell layer < 10 pm-thick servations), and these changes were not age-dependent. to a 10 to 12-cell layered, keratinized epithelium > 60Therefore, measurements of epidermal height were ad- pm thick (Figs. 1-3). In contrast, epidermis from adult justed for 4% linear shrinkage because it was reason- lower limb had about 25 cell layers and was almost 75 able to assume that the epidermal compartment might ym in thickness (Figs. lg, 2). Epidermal growth of the lower limb occurred in three also change by this amount. Tissue compression due to sectioning was assayed for different stages during embryonic and fetal developby comparing the area of the Epon block face to the ment. The first growth spurt occurred between 5 and * 203 IN U T E R O GROWTH OF HUMAN EPIDERMIS 14 wks 1 17 wks rc Adult Fig. 1. Photomicrographs (left panel) of toluidine blue-stained sections (1 km-thick) of Epon-embedded embryonic fetal and adult skin, with corresponding profiles of digitized epidermis (right panel.) P, periderm a-g X320. 204 C.A. FOSTER, J.F. BERTRAM, AND K.A. HOLBROOK u - Epidermis w i t h periderrn t Epidermis w i t h o u t periderrn 70 Iy = -137.67 + x E 5 2 . 8 2 6 ~- 5 . 9 5 9 ~+ ~0 . 2 4 3 ~ ~ ( R = 0.991) 60- I- I W 0 5040- -I 2 30- Ly I g I r/u -L w - *-A&' 8 12 10 14 18 16 20 -+IAd"lt 2 4 22 Y a 5 W E k 7 111 y = -48.832 53 W 6-I ' I 6 * I I ' 9 8 I . ' I 10 11 I ' 12 ' I 13 Fig. 4. Least-squares plot and equation describing the relationship between epidermal height (pm) and EGA in weeks (6 to 13). Fig. 2. Plot of epidermal height (pm) 2 standard deviation showing measurements with the periderm layer, without periderm, and periderm only (see Table 1) against EGA in weeks (6 to 23). Adult values (with the stratum corneum) are included. I . r ' ESTlPlATED G E S T A T I O N A L AGE ( w k s ) ESTIMATED GESTATIONAL AGE (Wks) W I / 10, 5 ---&- 0 6 20- 501 - I 14 . I 16 + * . 1 9 . 0 3 6 ~- 1 , 184x2 ( R = 0.931) I 18 . + 0.024~ 0 I 20 8 22 1 7 24 - 26 ESTIMATED GESTATIONAL AGE (wks) 4- - Fig. 5. Least-squares plot and equation describing the relationship between epidermal height (pm) and EGA in weeks (14 to 25). 2- rt 0 7 ' 4 6 1 ' I 8 . I 10 ' . 12 I 14 ' 9 16 . I ' 8 18 20 1 , ' 8 ' ' 22 2 4 26 E S T I H A T E D G E S T A T I O N A L AGE (wks) Fig. 3. Least-squares plot and equation describing the relationship between number of epidermal cell layers and EGA in weeks (5 to 25). 13 weeks EGA when the epidermis increased in height from < 10 pm to about 66 pm, with a slope of 7.5 (Fig. 2); the number of cell layers increased from one to four (Fig. 3). The periderm (the outermost epidermal layer) reached its maximal height by 12.5 to 13.5 weeks (Figs. lc, 2, Table l),then began to decline as the cells entered the complex bleb stage described by Holbrook and Odland (1975)and expanded laterally. Periderm regression, marked by a decrease in periderm thickness, loss of glygogen, and formation of a cornified cell envelope, was most dramatic between 13 and 14 weeks (Fig. lc, d), as reflected by the abrupt decrease in epidermal thickness from about 66 pm to 52 pm (Fig. 2, Table 1). Between 14 and 21 weeks, designated the "plateau phase," the epidermis increased from four to six layers (Fig. 3) but essentially remained the same thickness (about 50 pm). The second growth spurt occurred after 21 weeks when the epidermis began to increase in height (Fig. 2), undergo interfollicular keratinization (Fig. 10, and accumulate more cell layers in the stratum intermediudspinosum and stratum corneum. As indicated by a slope of 3.5, the increase in epidermal thickness between 21 and 25 weeks was more gradual than during the first growth phase before 13 weeks (Fig. 2). These data were tested by applied regression analysis and plotted as a function of age. The cubic equation y = -137.67 + 52.826~- 5.959~' ,0.243x3(R = 0.991), where y = epidermal height in micrometers and x = EGA in weeks (Fig. 4), best described the relationship between age and epidermal thickness during the first trimester (6 to 13 weeks). Epidermal growth during the second trimester, from 14 to 25 weeks (Fig. 5), was defined by the equation y = -48.832 + 19.036~1.184~2+ 0.024~3(R = 0.931). The relationship between EGA and number of cell layers also was curvilinear. The equations y = 3.81 - 0 . 5 4 6 ~+ 0.047~' 0 . 0 0 1 (R ~ ~= 0.989) and y = - 3.944 + 1 . 4 0 5 ~ - 0.094~~ 0.002~ (R~ = 0.988) where y = number of cell layers and x = EGA in weeks, best described the relationship between 5 and 25 weeks EGA, with (plot not shown) + + 205 IN UTERO GROWTH OF HUMAN EPIDERMIS TABLE 1. Epidermal height measurementsat progressive stages of human gestation Estimates (wk) Gestational age no. of specimens 6 7 8 9 10 11 13 14 16 17 21 23 Adult 4 2 3 3 3 2 3 4 Epidermal thickness + S.D. (pm) With periderm Without periderm Periderm only 18.14 f 0.67 20.88 2 3.56 28.65 f 4.68 33.64 2 1.87 36.89 f 1.88 44.32 f 4.47 65.83 f 8.63 51.68 i 4.16 52.83 f 4.88 51.79 f 2.80 50.96 -t 3.12 60.11 f 6.06 74.46 f 8.94l 12.85 f 1.04 16.03 f 3.29 21.34 f 4.71 25.15 f 4.47 27.19 ? 2.70 31.36 -t 5.22 48.20 f 4.75 42.32 f 5.09 45.76 -t 2.34 47.32 _t 2.08 48.10 ? 4.16 56.13 ? 6.86 NA 5.28 2 0.36 4.88 c_ 1.41 7.28 f 1.56 8.52 f 2.91 9.77 f 2.39 13.72 f 0.83 16.69 f 3.64 7.86 f 2.12 7.12 f 2.60 4.47 -t 2.91 2.86 -t 1.04 1.72 ? 0.93 NA 'Including a stratum corneum. and without adult values, respectively (Fig. 3). It should be emphasized that the above equations do not necessarily predict outside of the designated age range for a given regression plot, e.g., 6 to 13 weeks EGA in Figure 4. Epidermal measurements excluding the periderm also were calculated from digitized contours (Fig. 2). The same basic pattern of increasing epidermal thickness was apparent. However, unlike the trend including periderm measurements, there was no statistically signficant decrease in epidermal thickness after 12 to 13 weeks. The periderm steadily increased in thickness, from 5 pm at 6 weeks to almost 17 pm by 12.5 to 13.5 weeks, then gradually decreased in height (Fig. 3, Table 1). By 23 weeks the peridermal layer was less than 2 pm in thickness, and by 25 weeks the periderm cells were shed into the amniotic fluid. DISCUSSION Quantitative, morphological techniques were used to analyze epidermal development of the lower limb. Development was defined as an increase in epidermal thickness and accumulation of cell layers. Three distinct growth stages between 5 and 25 weeks EGA were identified. Although previous investigators have described human embryonic and fetal development both morphologically and biochemically (Moll et al., 1983; Dale et al., 1985; Smith et al., 1986; reviews by Breathnach, 1971; Holbrook, 1979; Holbrook and Hoff, 1984), this is the first quantitative study in humans of epidermal dimensions at progressive stages during the first two trimesters of embryonic and fetal development. The observed trends in epidermal growth seem to correlate with the two main phases of gestational maturation (Moore, 1982), i.e., organogenesis (embryonic period) and rapid body growth along with differentiation of established organs (fetal period). Most of the first growth spurt (5 to 13 weeks), when the epidermis undergoes its greatest increase in thickness, coincides with the initial development of the main organ systems (organogenesis) with little body growth. Conversely, the epidermal height remains essentially unchanged for almost 2 months (14 to 21 weeks) during the plateau phase when the rate of body growth is very rapid (Moore, 1982), including 5-fold increase in surface area of the lower limb (calculated from Klein and Scammon, 1930). During this latter stage, some of the keratinocytes may move laterally to accomodate the increased surface area, but many are also growing in a downward direction as they give rise to the epidermal appendages (review by Holbrook, 1979). Changes in body surface area, cell structure, and mitotic activity are among the factors that could influence the pattern of epidermal growth during embryonic and fetal development. The differences in epidermal thickness would appear to reflect changes in cell size and shape more than changes in the directional movement (apically versus laterally) of proliferating keratinocytes because the addition of cell layers is relatively constant between 5 and 21 weeks. In general, embryonic and early fetal keratinocytes are plump and balloon-shaped with large amounts of glycogen (Holbrook and Odland, 1975). By the second trimester, suprabasal keratinocytes have expanded and flattened in parallel with cells of the periderm, while basal cells have become less cuboidal, more columnar, and largely depleted of their glycogen content. Lateral expansion and regression of periderm cells after about 13 weeks accounts for a considerable decrease in epidermal height; however, the basic pattern of increasing epidermal thickness is still reflected in measurements excluding the peridermal layer. The flattening of epidermal cells during the plateau phase may be a mechanical response, in part, to the rapid internal expansion of growing tissues, thereby allowing for the addition of new cell layers while maintaining a relatively constant epidermal thickness. However, intrinsic cellular changes, particularly the loss of glycogen reserves from the suprabasal cells, would probably contribute more to the flattening of keratinocytes. Although mitotic activity must play an important role in determining the pattern of epidermal development, little is known about the dynamics of keratinocyte proliferation during ontogeny. A few investigators have observed a steady decrease in the labeling index of human embryonic and fetal epidermal cells from 26% at 8 weeks EGA to 4% by 16 to 19 weeks (calculated from Gerstein, 1971; Stern, 1974; Bickenbach and Holbrook, 1987). The high labeling index in humans between 8 and 12 weeks EGA is likely due to thymidine incorporation by cells in basal, intermediate, and periderm 206 C.A. FOSTER, J.F. BERTRAM, AND K.A. HOLBROOK layers (Stern 1974; Bickenbach and Holbrook, 1987). By 4 to 5 months labeling of fetal keratinocytes is restricted to the basal layers, and the labeling index appears to have stabilized to a value similar to that of the adult (5.5%)(calculatedfrom Epstein and Maibach, 1965; Lachapelle and Gillman, 1969; Weinstein and Frost, 1969; Heenan and Galand, 1971; Flaxman and Chopra, 1972; Allegra and Panfilis, 1974; Gelfant, 1982; Stern, 1974). Thus, based on published data and the present observations, we conclude that an inverse relationship exists between the mitotic index and the increase in both epidermal thickness and number of cell layers. The trends that have emerged from this study of epidermal development may help to provide a better understanding of the complex, interrelated events that culminate in the formation of a stratified and keratinized epithelium. Although the data reported herein are limited to one region of the body, it has been demonstrated that the thickness of the epidermis is reasonably consistent among most body regions at developmental ages ranging from 45 days EGA to 22 weeks. The head, palms, and soles, however, are exceptions showing accelerated differentiationcompared with the other regions, particularly during the second trimester (Holbrook and Odland, 1980).Thus the measurements for the thigh may be considered representative for the body in general. Flaxman, B.A., and D.P. Chopra 1972 Cell cycle of normal and psoriatic epidermis in vitro. J . Invest. Dermatol., 59t102-105. Foster, C.A. 1987 Differential effects of tissue processing on human embryonic and fetal skin. Anat. Rec., 218;355-358. Foster, C.A., RA. Holbrook, and A.G. Farr 1986 Ontogenyof Langerhans cells in human embryonic and fetal skin: Expression of HLA-DR and OKT-6 determinants. J. Invest. Dermatol., 86:240-243. Gelfant, S. 1982 “Of mice and men”: The cell cycle in human epidermis in oivo. J. Invest. Dermatol., 78t296-299. Gerstein, W. 1971 Cell proliferation in human fetal epidermis. J. Invest. Dermatol., 57t262-265. Haftek, M., M. Faure, D. Schmitt, and J. Thivolet 1983 Langerhans cells in skin from patients with psoriasis: Quantitative and qualitative study of T6 and HLA-DR antigenexpressing cells and changes with aromatic retinoid adminstration. J. Invest. Dermatol., 81:lO14. Heenan, M.A.H., and P. Galand 1971 Cell population kinetics in human epidermis: In vitro autoradiographic study by double l a b e h g method. J . Invest Dermatol., 56t425-429. Holbrook,K.A. 1979 Human epidermal embryogenesis. Int. J. Dermatol., 18:329-356. Holbrook, KA., and M. Hoff 1984 Structure of the developing human embryonic and fetal skin. Semin. Dermatol., 3t185-202. Holbrook, K.A., and G.F. Odland 1975 The fme structure of developing human epidermis: light, scanning, and transmission electron microscopy of the periderm. J . Invest. Dermatol., 65t16-38. Holbrook, KA., and G.F. Odland 1980 Regional development of the human epidermis in the first trimester embryo and the second trimester fetus (ages related to the timing of amniocentesis and fetal skin biopsy). J. Invest. Dermatol., 8Ut161-168. Karnovsky, M.J. 1965 A formaldehyde-glutaraldehyde furative of high osmolarity for use in electron microscopy. J. Cell Biol., 27t137A. Klein, A.D., and R.E. Scammon 1930 the regional growth in surface area of the human body in prenatal life. Proc. SOC.Exp. Biol. Med., 27:463-466. Kleinbaum, D.G., and L.L. Kupper 1978 Applied regression analysis and ACKNOWLEDGMENTS other multivariable methods. Duxbury Press, North Scituate, Mass. pp. 113-130. We are grateful to Drs. John Sundsten, Paul Samp- Lachapelle, J.M., and T. Gillman 1969 Tritiated thymidine labelling of son (Department of Statistics), Mr. James Holbrook normal human epidermal cell nuclei: A comparison, in the same subjects, of in vivo and in vitro techniques. Br. J . Dermatol., (NOAA),and Ms. Mary Pat Larson for generously shar81t603-616. ing their expertise in computers and statistical analyJ.H. 1961 Improvements in epoxy resin embedding methods. J. sis. We wish to thank Mr. Robert Underwood and Mrs. LUR,Biophys. Biochem. Cytol., 9:409-414. Mary Hoff for their invaluable technical assistance, and Moll, R., I. Moll, and W. Wiest 1983 Changes in the pattern ofcytokeratin polypeptides in epidermis and hair follicles during skin development Drs. John Prothero and Pritinder Kaur (Fred Hutchin human fetuses. Differentiation, 23:170-178. inson Cancer Research Center) for helpful suggestions. K.L. 1982 The Developing Human. W.B. Saunders Co., PhilaThis study was suported by grants HD 17664 (K.A.H.), Moore, delphia, pp. 366-374. AR 21557 (K.A.H.)from the National Institutes of Health Prothero, J., and J . Prothero 1982 Three-dimensional reconstruction from serial sections. I. A portable microcomputer-based soRware and Public Health Service National Research Service package in Fortran. Comput. Biomed. Res., 15:598-604. Award 2 T32 GMO7270-09 (C.A.F.) Reynolds, E.S. 1963 The use of lead citrate at high pH as an electron opaque stain in electron microscopy. J. Cell. Biol., 17:208-213. Richardson, K.C., L. Jarret, and E.H. Finke 1960 Embedding in epoxy LITERATURE CITED resins for ultrathin sectioning in electron microscopy. Stain Technol., Allegra, F., and G. Panfiiis 1974 An in vivo method of studying the 35t313-323. kinetics of cell proliferation in normal human epidermis. Acta Derm. Smith, L.T., K.A. Holbrook, and J.A. Madri 1986 Collagen types I, 111, Venereol. (Stockh) 54:87-90. and V in human embryonic and fetal skin. Am. J. Anat., 175.507Bickenbach, J.R., and K.A. Holbrook 1987 Label-retaining cells (LRCs) 521. in human embryonic and fetal epidermis. J. Invest. Dermatol., Stem, LB. 1974 The uptake of tritiated thymidine by human fetal epi88:42-46. dermis. J . Invest. Dermatol., 63:268-272. Breathnach, A.S. 1971 An Atlas of the Ultrastructure of Human S h . Sundsten, J.W., and J.W. Prothero 1983 Three-dimensional reconstrucChurchill, London. tion from serial sections. 11.A microcomputer-based facility for rapid Dale, B.A., K A Holbrook, J.R. Kimball, M. Hoff, and T.-T Sun 1985 data collection. Anat. Rec., 207:665-671. Expression of epidermal keratins and filaggrin during human fetal Weinstein, G.S., and P. Frost 1969 Cell proliferation kinetics in benign skin development. J. Cell Biol., 101:1257-1269. and malignant skin diseases in humans. Natl. Cancer Inst. Monogr., Epstein, W.L., and H.I. Maibach 1965 Cell renewal in human epider3Ot209-246. mis. Arch. Dermatol., 92t462-468.