Influence of maternal diabetes on basement membranes, type 2 cells, and capillaries in the developing rat lung

DEVELOPMENTAL

BIOLOGY

104,469-476

(1984)

Influence of Maternal Diabetes on Basement Membranes, Type 2 Cells, and Capillaries in the Developing Rat Lung MARGARET M. GRANT, NANCY R. CUTTS,AND JEROMES. BRODY Pulmmarg Center, Department of Medicine, and Department of Biochemistry, Bastwn University Schd of Medicine, Boston, Massachusetts 02118 Received November 15, 1982; accepted in revised

form April 30, 198.4

To determine the effect of maternal diabetes on rat lung development, we studied the ultrastructure of the alveolar wall from the ninteenth day of gestation (term = 22 days) through the eighth postnatal day in fetal and neonatal rats of mothers with streptozotocin-induced diabetes. In normal fetal lung development, epithelial basement membranes develop large discontinuities beneath type 2 cells, through which cytoplasmic foot processes extend into the interstitium. Maternal diabetes delays the appearances of these epithelial basement membrane discontinuities and reduces the number of type 2 cell processes that penetrate it. These alterations in epithelial basement membrane are reversed after birth. There is no ultrastructural evidence of a delay in type 2 cell maturation as assessed by lamellar body volume density morphometry. Endothelial basement membranes, which are not present around the growing pulmonary capillary bed in the pseudoglandular lung, are seen late in normal gestation, primarily around capillaries forming the mature air-blood barrier. This development of endothelial basement membrane may be delayed in the fetuses of diabetic mothers and reflects a significant delay in the expansion of the pulmonary capillary network in these animals as assessed by morphometric volume density measurements. This effect on capillary growth is not reversed in the newborn animals through 8 days after birth. The summation of these effects indicates a generalized slowing of fetal lung development by maternal diabetes, some of which effects persist after birth and may continue to influence lung development during the period of postnatal alveolar septal growth. INTRODUCTION

et aL, 1976). By studying lung mechanical properties, Basement membranes’ separating epithelium and several investigators have demonstrated a functional mesenchyme modulate the interaction of these tissue delay in fetal lung maturation in a rabbit model of macomponents and regulate the early development of the ternal diabetes (Sosenko et d, 1980b; Bose et cd, 1980). bronchial tree in the fetal lung (Alescio and Cassini, These studies suggest that type 2 cell and/or alveolar 1962;Wessels, 1970; Bluemink et al, 1976). In later stages wall maturation might be delayed in fetuses of diabetic of lung development, just before birth and in the new- mothers. We therefore examined whether maternal diabetes born, changes observed in the basement membrane bemight affect the changes in fetal lung basement memneath differentiating type 2 epithelial cells suggest that branes which are associated with normal type 2 cell this structure may also influence the perinatal remoddifferentiation and remodeling of the perinatal lung. In eling of the gas exchange region in the lung (Grant et addition, we quantified alveolar capillary expansion in ~4 1983; Brody et a& 1982). the same lungs to determine whether development of Diabetes mellitus is known to affect basement memthis component of the alveolar wall was retarded in the brane structure and composition in many adult tissues, fetuses of diabetic mothers. Our results show that maincluding the lung (Vracko, 19’78; Vracko et al, 1979). ternal diabetes does delay epithelial basement memHowever, little information exists about diabetes-inbrane development in fetal lungs, that this delay is reduced alterations in basement membranes of fetal lung versed after birth, and that the basement membrane tissues. Preterm infants of diabetic mothers have a 5.6changes do not seem to correlate with the maturity of fold greater risk of developing respiratory distress synthe overlying type 2 cells. Endothelial basement memdrome than do those of nondiabetic mothers (Roberts brane appearance and capillary growth are significantly retarded, and these effects are not reversed in the early ’ The term basement membrane as used in this paper refers to a postnatal period. structure approximately 70- to lO&nm-wide underlying epithelial and endothelial cells which appears, by our staining methods, as a central, electron-dense line (lamina densa) with more electron-lucent zones (laminae rarae) on either side. This structure contains collagen and noncollagen proteins and proteoglycans whose composition cannot be distinguished by simple ultrastructure.

METHODS

Animals. Dated-pregnant Sprague-Dawley rats (Charles River Breeding Laboratories) were made di-

469

0012-1606m $3.00 Copyright All rights

Q 1994 by Academic Press. Inc. of reproduction in any form reserved.

470

DEVELOPMENTAL BIOLOGY

abetic by a single intravenous injection of streptozotocin (50 mg/kg) given on the sixth day of pregnancy (breeding day = Day 0; term = 22 days). Urinary glucose levels were monitored throughout pregnancy with Ketodiastix and only those animals having levels greater than 2000 mg/dl were used in this study. None of the animals exhibited ketonuria. Fetuses were obtained from these animals and from normal pregnant rats on the nineteenth through the twenty-second day of gestation. Newborn, 2-, 4-, and g-day-old animals were also studied. In total, 38 fetuses from 16 normal mothers and 38 from 14 diabetic mothers were studied. Maternal and fetal blood glucose levels were estimated on all animals at the time of sacrifice using Dextrostix. Normal mothers had blood glucose levels ranging from 90 to 130 mg/dl and their fetuses from 45 to 90 mg/dl. Diabetic mothers’ blood glucose was in excess of 250 mg/dl; their fetuses and newborns also had levels of 250 mg/dl or greater. Within 4 hr after birth these levels dropped to 130 mg/ dl and were identical to normal animals at subsequent times. Serum from selected animals was used to quantify serum glucose and serum insulin (Micromedic Insulin RIA Kit, Micromedic Systems, Horsham, Pa.). Table 1 lists serum glucose and insulin levels measured in two control and two diabetic 22 day pregnant mothers and their fetuses. Serum from each set of fetuses was pooled. Diabetic mothers were hyperglycemic and hypoinsulinemit while diabetic fetuses were hyperglycemic and hyperinsulinemic. Tissue preparation. Fetuses were removed from pentobarbital-anesthetized mothers and their lungs fixed without inflation. Newborn and older rats were anesthetized with pentobarbital (5 mg in 0.1 ml), their tracheas cannulated, and the lungs fixed in situ by inflation with 0.3 to 0.7 ml of fixative before being dissected. For electron microscopic examination, tissue was treated with tannic acid, in order to enhance the contrast of the basement membrane. Minced lung was fixed 3 hr in Karnovsky’s buffered glutaraldehyde-formaldehyde (Karnovsky, 1965), postfixed 1 hr in ferrocyanide-reduced

TABLE 1 GLUCOSE AND INSULIN LEVELSO Serum glucose (mddl) Normal mother Fetuses Diabetic mother Fetuses

124.5 47.6 570.6 202.8

+ t+ +

23.6 23.7 100.8 49.5

Serum insulin NJ/ml) 101.5 117.0 52.0 217.5

+ Ik + +

54.5 7.1 24.3 38.9

VOLUME 104, 1984

1% OsOl (Karnovsky, 1971), and then immersed in 1% tannic acid for 30 min (Simionescu and Simionescu, 1976). The fixed tissue was then washed, dehydrated in acetone, infiltrated, and embedded in Polybed-812 (Polysciences, Warrington, Pa.). Thin sections were stained with saturated aqueous uranyl acetate and with 1% lead citrate. Morphologic and murphometric evaluation, One hundred randomly selected alveolar epithelial cuboidal and type 2 cells were examined at each pre- and postnatal age from litters of both normal and diabetic mothers. Twenty-five cells were analyzed from each of four animals at each age (two animals from two litters). With each cell, we measured whether the basement membrane was continuous beneath the cell, the number of basal cytoplasmic foot processes per cell, and the number of these processes penetrating the basement membrane. The volume fraction of lamellar bodies in these cells was determined by point counting (Weibel, 1973) placing a square lattice system containing test points, 1 cm apart, over micrographs printed at a magnification of 17,000~. Results are expressed as ratios, number of test points falling over lamellar bodies to number of test points falling over cell cytoplasm including lamellar bodies; that is, volume of lamellar bodies to volume of bodies/ Veytoplasm). cell cytoplasm ( Valsmdar Capillary basement membrane continuity was assessed for 100 capillary cross sections in the 22 day fetal lungs of four animals each (one animal from four litters) from both normal and diabetic mothers by following the endothelial cell basement membrane around the entire capillary circumference. Capillary volume density in fetal and neonatal lungs was estimated by point counting. At a magnification of 100X, 0.5-pm-thick sections stained with toluidine blue were examined with a lo-mm-square lattice ocular grid using 100 test points. Two hundred tissue points were counted for each section and the capillary volume fraction was determined as the ratio of number of points falling on capillary lumens to the total number of points falling on all tissue including the capillaries (Vmpin,J Vtissue).Areas around airways and noncapillary blood vessels were excluded. Five animals each (taken from two litters) from both normal and diabetic mothers were used at each age. Statistics. Comparisons between groups were made using Student’s t test. Multiple comparison analyses of morphometry data were performed with a standard computer program, SAS (Statistical Analysis System, SAS Institute, Cary, N. C.) RESULTS

‘Data are expressed as mean f standard deviation. Differences between normal mothers and their fetuses and the corresponding diabetic animals are significant at P c 0.05.

In development of the normal fetal rat lung, discontinuities appear in the previously continuous basement

GRANT, CUTTS, AND BRODY

Diabetes and Lung Lkuelqmnent

471

membrane that underlies differentiating type 2 cells (Grant et d, 1983, and Fig. 1). The basal surface on the maturing type 2 cell has numerous cytoplasmic foot processes, some of which are contained by the basement membrane (arrowheads) while others pass through basement membrane discontinuities into the interstitium (asterisk and inset). The cytoplasmic processes which penetrate the basement membrane frequently approximate closely the underlying lipid-containing interstitial cells. Neither the basement membrane discontinuities nor the type 2 cell foot processes correlate with these cells’ lamellar body content. The basement membrane beneath type 1 cells is continuous and these cells have no basal cytoplasmic foot processes.

domly selected type 2 cells at each age, from the nineteenth day of gestation through the eighth day after birth, comparing the continuity of basement membrane in fetuses of diabetic mothers with those of normal mothers. As shown in Fig. 2, there is a delay in both the timing of appearance and the extent of discontinuity in type 2 cell basement membranes in fetuses of diabetic mothers. This effect is most pronounced in the immediate perinatal period when, in normal fetuses, only 18% of type 2 cells have a continuous basement membrane at 22 days and only ‘7% immediately after birth. In the fetuses of diabetic mothers 40 and 27% of the type 2 cells have continuous basement membrane at 22 days and at birth, respectively.

Epithelial Basement Mmbrane Continuity In order to assess the effects of maternal diabetes in type 2 cell basement membrane, we examined 100 ran-

Type 2 Cell Foot Processes As type 2 cells differentiate in late fetal life, the number of foot processes increases, reaching a plateau just

FIG. 1. Electron micrograph of a type 2 cell from a normal 22-day fetal rat lung. A continuous basement membrane with a distinct lamina densa is seen along the lateral surfaces of the cell (arrowheads), but is discontinuous at the basal surfaces of the cell. Through the discontinuities in the basement membrane, type 2 cell cytoplasmic foot processes (asterisks and inset) pass into the interstitium and approximate lipidcontaining interstitial cells (LIC). lb, Lamellar body. Magnification, X17,ooO,inset, X23,940; bar = 1 pm.

472

DEVELOPMENTALBIOLOGY

VOLUME104,1984

Basement Membrane Continuity and Type .2 Cell L.urnellar Body Volume Fraction

FIG. 2. Effect of maternal diabetes on the percentage of cuboidal and type 2 epithelial cells with continuous basement membrane at different ages of the developing rat lung. In the lungs of fetuses from normal mothers (solid bars) epithelial basement membrane becomes progressively more discontinuous late in gestation, reaching a minimum around birth; this trend reverses during the first postnatal week. In the fetuses of diabetic mothers (hatched bars) both the appearance of diecontinuities is delayed and their extent reduced. Mean values f SEM; asterisk = P < 0.05.

before birth (Fig. 3). The type 2 cells of fetuses of diabetic mothers have significantly fewer foot processes than normal animals. This trend begins before birth and continues in the postnatal period for up to 8 days. The percentage of these foot processes penetrating the basement membrane is the same in the offspring of both normal and diabetic mothers, however, the total number of type 2 cell processes extending into the interstitium is still decreased in maternal diabetes.

5I Iq

Normal Diabetic

mother mother

Because maternal diabetes effected marked differences in fetal type 2 cell basement membrane before birth we wondered whether maturation of the surfactant content of these cells, as judged by lamellar body volume, might also be influenced. Morphometric analysis of the volume fraction of lamellar bodies contained in the type 2 cell cytoplasm at 22 days gestation, in the newborn and on the second postnatal day shows no significant difference from normal in fetuses of diabetic mothers (Table 2). Thus during the perinatal period in which maternal diabetes has a maximal affect on type 2 cell basement membrane structure there appears to be no effect on these same cells’ surfactant content as measured by lamellar body volume fraction. Endothelid Basement M&ane: Nmrnal Development In the normal l&day fetus, little or no basement membrane is present around the endothelial cells of the capillaries in the pulmonary mesenchyme (Fig. 4a) while epithelial basement membrane is well established (arrowheads). Small patches of surface-associated material are present along the margin of the endothelial cells (arrows). During the period of capillary bed expansion from 20 days gestation to birth at 22 days, the layer of cell-associated material becomes more prominent on the abluminal side of capillaries (Fig. 4b) and some areas of distinct capillary basement membrane are seen. As capillaries approach the flattened type 1 cells, the surface-associated material completely covers the endothelial cell on the interstitial side of the pulmonary capillary network. A continuous capillary basement membrane with a distinct lamina densa as seen in the adult lung is generally not found in the fetal lung. Indeed, maturation of the capillary basement membrane probably is completed only during the first several weeks after birth (Brody et al, 1982). TABLE 2 LAMELLAR BODYCONTENTOFTYPE~CEXUY Vhmetlrrr

22

Birth

+2

+4

+a

bodied

&oplasm

Age

Normal mothers

Diabetic mothers

22 day Birth +2 day

0.069 f 0.025 0.107 -+ 0.026 0.171 f 0.026

0.072 t- 0.014 0.119 + 0.026 0.132 ir 0.033

Age

FIG. 3. The number of basal cytoplasmic foot processes per epithelial cell is significantly reduced by maternal diabetes. Mean values f SEM; asterisk = P < 0.05.

“Data are expressed as mean f standard deviation. There is no significant difference between fetuses of normal and diabetic mothers at any age.

GRANT, Gums,

AND BRODY

Diabetes and Lung Development

FIG. 4. Capillary basement membrane development in normal fetal rat lung. (a) 18 day fetal rat lung: Endothelial basement membrane is undeveloped around mesenchymal capillaries (cap), although some surface-associated glycocalix is present (arrows). The epithelial cell (Ep) basement membrane at the upper right (arrowheads) is well defined. Magnification, X15,300, bar = 1 pm. (b) 29 day fetal rat lung: On the interstitial side of a capillary (cap) beneath a flattened type 1 epithelial cell (Ep), the endothelial cell surface-associated material (arrows) is more prominent than at earlier times. It has still not condensed into a mature basement membrane with the distinct lamina densa that is characteristic of the structure in the adult lung. The epithelial basement membrane (arrowhead) is seen on the left of the figure. Magnification, X19,700; bar = 1 pm.

474

DEVELOPMENTAL BIOLOGY

VOLUME 104,1984

Continuity of Endothelial Basement Membrane We measured the continuity of basement membrane and surface-associated material from complete cross sections of 100 capillaries in 22 day fetal lungs of animals from 4 normal mothers and 4 diabetic mothers. In the lungs of 22 day fetuses of normal mothers, 92.7 + 0.4% (mean + standard deviation) of all the capillaries’ interstitial surface had continuous basement membrane or surface-associated material while in the capillaries of 22 day fetuses of diabetic mothers, significantly less basal surface was covered, 34.0 f 3.7% (P < 0.05). The difference between fetuses from normal and diabetic mothers was most marked in capillaries that had not yet formed the mature alveolar-capillary interface (Fig. 5). The capillaries located beneath type 1 cells are those forming the functional air-blood barrier and represent the most mature group. In these, 95.3 +- 0.8% of the circumference of the capillaries is surrounded by either basement membrane or endothelial cell surfaceassociated material in normal fetuses. Although not statistically significant, there is a slight reduction in continuity of basement membrane material to 92.2 -t 2.7% of the capillary circumference in fetuses of diabetic mothers. Capillaries located adjacent to cuboidal cells or in the center of the interstitium probably represent actively growing vessels and normally have less basement membrane-related material surrounding them (92.8 + 2.6 and 86.0 + 2.3%, respectively). In these groups of capillaries, maternal diabetes significantly (P < 0.05) retards the appearance of basement membrane material (reduced to 77.4 f 10.1 and 80.3 f 3.6%, respectively).

: : 1; . b ; :: E

FIG. 6. Morphometric measurement of capillary growth in fetuses of normal and diabetic mothers. Development of pulmonary capillaries is retarded in fetuses of diabetic mothers (hatched bars) when compared with normal fetuses @lid bars). This effect of maternal diabetes is not reversed up to the eighth postnatal day. Mean values + SEW, asterisk = P < 0.05.

The apparent delay of pulmonary capillary basement membrane development in the 22 day fetuses of the diabetic mothers led us to question whether overall growth of the capillary network was delayed by maternal diabetes, and, if so, whether such a delay might be reversible after birth. The lungs of 22 day fetuses, newborn, 2-, 4-, and 8-day-old offspring from normal and diabetic mothers were examined at the light microscopic level to determine the volume fraction occupied by capillary lumens ( Veapillaw) within the total lung parenchymal volume

5 z

0 4

100

22 DAY

NORMAL MOTHER

FETUS

1

Type

under I cell

under Cuba,dal

cell

q

center I”terstit,“m

DIABETIC MOTHER

01

( Vtissue)-

There is a dramatic increase in capillary density over the first week of postnatal life in normal rat pups. At all ages examined, from the twenty-second day of gestation through the eighth day after birth, pulmonary capillary density is significantly reduced in the offspring of diabetic mothers (Fig. 6). There is no increase in the capillary volume fraction in the diabetic group between 2 and 8 days of age. At no age does the capillary density in the lungs of offspring of the diabetic mothers reach that of normal animals. DISCUSSION

In this study of the influence of maternal diabetes on late fetal and early postnatal lung growth and develFIG. 5. Effect of maternal diabetes on continuity of capillary base- opment, we have found effects, not only on epithelial ment membrane in 22 day fetal rat lungs. Immature capillaries, those basement membranes and type 2 cells, but also on ennot underlying type 1 cells, have less complete basement membrane and surface-associated glycocalix in fetuses of diabetic mothers dothelial basement membranes and on capillary growth. in the basement (hatched bars) than do normal fetuses (solid bars). Mean values + SEW, The development of discontinuities asterisk = P < 0.05. membranes underlying type 2 cells is delayed and the CAPILLARY

POSITION

GRANT,

CUTE,

AND BRODY

numbers of type 2 cell foot processes which normally penetrate into the interstitial compartment are reduced in fetuses of diabetic mothers. Maturation of the surfactant content of type 2 cells judged by ultrastructure does not appear to be delayed by diabetes since type 2 cell lamellar body volume density was similar in fetuses of normal and diabetic mothers. The appearance of endothelial basement membrane is also delayed, and overall capillary growth is retarded. The summation of all of these effects indicates that maternal diabetes causes a more generalized slowing of fetal lung development than has been recognized previously. Moreover, the effects on the basilar surface of type 2 cells (Fig. 3) and on capillary growth (Fig. 6) are not completely reversible up to 8 days after birth, even though the blood glucose levels of the newborns drop to normal within hours of birth. Thus the effects of maternal diabetes may continue to influence lung development during the period of postnatal alveolar septal growth. The nutritional status of the pregnant rat can influence fetal lung growth (Faridy, 1975). Maternal malnutrition, in particular protein rather than caloric deprivation, reduces fetal lung dry weight and affects phospholipid content. In our studies, the diabetic mothers, while having free access to food and water at all times, did loose up to 15% of their initial body weight during their pregnancy and we cannot rule out effects on fetal growth and development by other nutritional alterations in addition to those in blood glucose and insulin levels. The retardation of lung development in the offspring of diabetic mothers might be the result either of hyperglycemia or of secondary hyperinsulinemia. In this study, the fetuses of the diabetic mothers had increased blood levels of both glucose and insulin (Table 1). High levels of insulin in vitro have been reported to retard biochemical activity and morphological maturation of fetal rabbit and rat lung type 2 cells in organ culture (Neufield, 1979; Gross et aL, 1980) and to inhibit cortisol stimulation of type 2 cell differentiation in mixed fetal lung cell culture (Smith et aL, 1975). Effects of insulin on other cell types and components of the developing lung have not been reported. Fetal hyperglycemia alone may retard lung growth. High levels of glucose in organ cultures of rat embryo cause neural tube abnormalities which appear to be due to delaying normal developmental processes (Cockroft and Coppola, 1977; Sadler, 1980b). Serum from diabetic adult rats also will cause growth inhibition and cell necrosis in cultured rat embryos (Sadler, 1980a) and this effect can be duplicated by the addition of glucose to normal rat serum (Sadler, 1980b). Elevated serum glucose levels in the fetuses of diabetic mothers may result in nonenzymatic glycosylation of

Diabetes and Lung Develqvment

475

proteins in the fetal lungs and thereby influence basement membrane connective tissue turnover as well as cell function in the developing organ. Dixon and Jersild (1982) have reported differences in alveolar epithelial cell surface glycoprotein sugars as demonstrated by lectin binding in fetuses and newborns of diabetic mothers. These changes are similar to those seen in the lungs of diabetic adult animals. Williams et al. (1982) have demonstrated that nonenzymatic glycosylation of tubulin decreases GTP-dependent polymerization and that the glycosylated tubulin isolated from brains of diabetic rats also exhibits this polymerization defect. Thus hyperglycemia may affect any cell function which is dependent upon tubulin polymerization and depolymerization, such as cell movement and division. In diabetic human adults, glomerular basement membrane collagen has increased hexose content in ketoamine linkages characteristic of nonenzymatic glycosylation while enzymatic glycosylation is unaffected (Uitto et ak, 1982). Increased glycosylation also occurs in vitro in rat glomerular and lens capsule basement membranes incubated in media containing elevated glucose concentrations (Cohen et uL, 1981) and in wivo in streptozotocin-induced diabetic adult rats (le Pape et aL, 1981). This increased nonenzymatic glycosylation, together with increased collagen cross-links reported in collagen from diabetic adult rats (Golub et aL, 1978) could reduce the basement membrane’s susceptibility to degradation, and account for the failure of the epithelial basement membrane to become discontinuous beneath type 2 cells in the fetuses of diabetic mothers. Glycosylation and increased cross-linking of collagen in the lung interstitium might also influence restructuring of that portion of the extracellular matrix. The short- and long-term consequences of the generalized slowing of lung growth and development in fetuses and newborns of diabetic mothers are many. Although type 2 cells of offspring of both normal and diabetic mothers have similar surfactant content as judged by lamellar body volume fraction (Table 2), the reduced numbers of cytoplasmic foot processes and basement membrane discontinuities (Figs. 2 and 3) may result in decreased interaction between the epithelial cells and the underlying interstitial cells and the connective tissue framework. Slowing the growth and the thinning of the interstitial walls of the alveolar septa might reduce the distensibility of the newborn lung and thus contribute to the increased risk of respiratory distress syndrome. The delayed appearance of mature endothelial basement membrane (Fig. 5) is a reflection of the slowed growth and expansion of the pulmonary capillary network (Fig. 6) which is evident in the fetuses of diabetic mothers and which persists in the postnatal period. Throughout the fetal and early postnatal development

476

DEVELOPMENTALBIOLOGY

of the lung the endothelial cells of the pulmonary capillaries are committed to growth and to angiogenesis. Under these conditions, endothelial basement membrane synthesis probably does not occur and thus is not directly affected by the maternal diabetes, although that of preexisting epithelial basement membrane can be. Capillary growth may be the controlling force which initiates the postnatal restructuring of the newborn lung and directs the alveolar septal growth that occurs during this period. The slowing of pulmonary capillary angiogenesis by maternal diabetes may retard postnatal alveolar growth in infants of diabetic mothers, and if this effect is not completely reversible with time, it could result in reduced numbers of alveolar septa and a reduction in the lung’s gas exchange capacity. In summary we have shown that maternal diabetes has widespread effects on fetal and postnatal lung development, altering extracellular matrix of both epithelial and endothelial cells and influencing capillary growth and development. The generalized slowing of lung development may contribute a tissue component to the increased incidence of respiratory distress syndrome in newborns of diabetic mothers and may also have prolonged effects upon postnatal lung growth and alveolar development.

VOLUME104,1984

GOLUB,L. M., GREENWALD.R. A., ZEBROWSICI, E. J., and RAMAMURTHY, N. S. (1978). The effects of experimental diabetes on the molecular characteristics of soluble rat-tail tendon collagen. Bimhim Biophvs A& 534, 73-81. GRANT, M. M., CUTTS,N. R., and BRODY,J. S. (1983). Alterations in lung basement membrane during fetal growth and type 2 cell development. Dev. BioL 97.173-183. GROSS,J., WALKER SMITH, G. J., WILSON,C. M., MANISCALO,W. M., INGELSON,L. D., BREHIER,A., and ROONEY,S. A. (1986).The in!luence of hormones on the biochemical development of fetal rat lung in organ culture. II. Insulin. Pediatr. Rex 14,834~838. KARNOVSKY,M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolarity for use in electron microscopy. J. Ceu Bid 27,137. KARNOVSKY,M. J. (1971). Use of ferrocyanide-reduced 0~0, in electron microscopy. Abstracts of papers (Number 284), 11th Annual Meeting, Am. Sot. for Cell Biology, New Orleans, p. 146. LE PAPE, A., GUIT~ON,J-D., and MUH, J. P. (1981). Modifications of glomerular basement membrane cross-links in experimental diabetic rats. Bidem. Biophys. Res. Commun 100, 1214-1221. NEUFELD, N. D., SEVANIAN, A., BARRETT, C. T., and KAPLAN, S. A. (1979). Inhibition of surfactant production by insulin in fetal rabbit lung slices. Pediatr. Rex 13, 752-754. ROBERT,M. F., NEFF, R. K., HUBBELL, J. P., TAEUSCH,H. W., and AVERY,M. E. (1976). Maternal diabetes and the respiratory-distress syndrome. N. EngL J. MecL 294, 357-360. SADLER, T. W. (1986a). Effects of maternal diabetes on early embryogenesis: I. The teratogenic potential of diabetic serum. Teratdogy 21,339-347.

SADLER, T. W. (1980b). Effects of maternal diabetes on early embryogenesis: II. Hyperglycemic-induced exencephaly. Teratdopy 21, 349-356.

SIMIONESCU,N., and SIMIONESCU,M. (1976). Galloylglucoses of low The authors thank Eva Belur for performing the insulin determimolecular weight as mordants in electron microscopy. I. Procedure nations and Laurie Beardsley for preparation of the manuscript. This and evidence for mordanting effect. J. Cell BioL 70, 608-621. work was supported by National Institutes of Health Grant I-IL 19717. SMITH, B. T., GIROUD,C. J. P., ROBERT,M., and AVERY, M. E. (1975). Insulin antagonism of cortisol action on lecithin synthesis by cultured fetal lung cells. .Z.Pedintr. 87, 953-955. REFERENCES SOSENKO,I. R. S., LAWSON,E. E., DEMOT~AG,V., and FRANTZ, I. D. (1980). Functional delay in lung maturation in fetuses of diabetic ALESCIO,T., and CASSINI, A. (1962). A quantitative assessment of rabbits. J. AppL PhysioL 48, 643-647. mesenchymal contribution to epithelial growth rate in mouse embryonic lung developing in vitro. J. EmbrgoL Exp. Morph& 17,213- UITTO,J., PEREJDA,A. J., GRANT,G. A., ROWOLD,E. A., KITO, C., and 227. WILLIAMSON,J. R. (1982). Glycosylation of human glomerular baseBLUEMINK,J. G., VAN MAURIK, P., and LAWSON,K. A. (1976). Intimate ment membrane collagen: Increased content of hexose in ketoamine cell contacts at the epithelial/mesenchymal interface in embryonic linkage and unaltered hydroxylysine-0-glycosides in patients with mouse lung. J. Ultmstmct Res. 66.257-270. diabetes. Connec Tiss. Res. 10,287~296. BOSE,C. L., MANNE, D. N., D’ERCOLE,A. J., and LAWSON,E. E. (1986). VRACKO,R. (1978). Effect of aging and diabetes on basal lamina thickDelayed fetal pulmonary maturation in a rabbit model of the diabetic ness of six cell types. In “Biology and Chemistry of Basement Mempregnancy. J. C&L Invest. 66,226-226. branes” (N. Kefalides, ed.), pp. 483-493. Academic Press, New York. BRODY,J. S., VACCARO,C. A., GILL. P. J., and SILBER, J. E. (1982). VRACKO,R. (1979). Basal lamina of alveolar epithelium and capillaries: Alterations in alveolar basement membranes during postnatal lung Quantitative changes with aging and in diabetes mellitus. Amer. growth. J. CeU BioL 95.394-402. Rev. Req. Dk. 120.973-983. COCKROFT, D. L., and COPPOLA,P. T. (1977). Teratogenic effects of WEIBEL,E. R. (1973). Stereological techniques for electron microscopic excess glucose on head-fold rat embryos in culture. Teru~ 16, morphometry. In “Principles and Techniques of Electron Micros141-146. copy” (M. A. Hayat, ed.), Vol. 3, pp. 237-296. Van Nostrand Reinhold, COHEN,M. P., URDANIVIA,E., SURMA,M., and CIBOROWSKI, C. J. (1981). New York. Non-enzymatic glycosylation of basement membranes. In vitro WESSELLS,N. K. (1970). Mammalian lung development interactions studies. Diabetes 30.367-371. in formation and morphogenesis of tracheal buds. .Z.Exp ZooL 175, DIXON, M. T., and JERSILD, R. A., JR. (1982). Influence of maternal 455-466. diabetes on lectin binding to the surface of alveolar epithelial cells. WILLIAMS, S. K., HORWARTH,N. L., DEVERING,J. J., and BITENSKY, J. CeU Bid 95,llOa. M. W. (1982). Structural and functional consequences of increased FARIDY,E. E. (1975).Effect of maternal malnutrition on surface activity tubulin glycosylation in diabetes mellitus. Proc NatL Acad sci of fetal lungs in rats. J. Appl PhysioL 39,535~540. USA 79,6546-6550.