- Email: [email protected]
Epithelial differentiation in the fetal rat colon
DEVELOPMENTAL
BIOLOGY
97,349-363
(1983)
Epithelial Differentiation
in the Fetal Rat Colon
I. Plasma Membrane PAMELA Pennsylvania
State
University, Received
C. COLONY
Phosphatase
AND MARIAN R. NEUTRA’
Hershey, Pennsylvania October
Activities
Harvard
16802;
8, 1982; accepted
in revised
Medical fmm
School,
December
Boston,
Massachusetts
02115
30, 1982
During the last week of gestation of the fetal rat, the epithelium of the colon is rapidly remodeled. At 16 days a primitive stratified epithelium surrounds a small central lumen. Over the next 3 days, the main lumen extends narrow clefts down to the basal cell layer and small secondary lumina appear within the stratified epithelium between these clefts. At 19 and 20 days, secondary lumina enlarge but remain discrete; an infusion of cationic ferritin into the main lumen does not enter secondary lumina. During the 2 days prior to birth (21-22), the secondary lumina join the main lumen as superficial cells are sloughed, and the epithelium becomes simple columnar. Freeze-fracture replicas indicate that luminal and nonluminal membrane domains of epithelial cell plasma membranes are separated by continuous tight junctions throughout the conversion process. Cytochemical analysis of tissue slices from 16- to 22-day fetal colon demonstrated the appearance and segregation of two phosphatases on apical and hasolateral membrane domains during epithelial conversion. Cysteine-sensitive pH 9.0 (alkaline) phosphatase activity was first detected along the luminal membranes of cells bordering both primary and secondary lumina at 18 days gestation and increased to a maximum at 20-21 days; weaker activity was present on basolateral membranes. Phosphatase activity at pH 8.0 also appeared at 18 days and increased thereafter, but was localized primarily on nonluminal membranes. At pH 8.0, reaction product appeared on both inner and outer sides of the membrane, and was only partially abolished by omission of K’ or addition of ouabain; thus the reaction may be only partially due to K+-dependent ATPase activity. Biochemical analysis of the cytochemical media confirmed the appearance of phosphatase activities at 18 days. Thus, plasma membrane phosphatase activities appear while the epithelium is still stratified, but are segregated to luminal and nonluminal membrane domains at the onset of activity. Segregation is maintained throughout the process of conversion of a simple columnar epithelium. INTRODUCTION
The epithelial lining of the rat small and large intestine is rapidly remodeled during the final week of gestation. At 16 days gestation, 6 days prior to birth, the entire intestine is lined by a primitive stratified epithelium surrounding a single central lumen. Cells bordering the lumen are joined by occluding junctions. By Day 21, the stratified cells have been replaced by a simple columnar epithelium with polarized differentiated cells. In the small intestine, this remodeling process involves the appearance of new secondary lumina deep in the stratified epithelium with the formation of new luminal cell surfaces and new occluding junctions on the deep epithelial cells bordering these lumina (Madara et aL, 1981), as well as invagination of the epithelium by underlying mesenchyme, and sloughing of superficial cells (Mathan et aL, 1976). Epithelial conversion also occurs in the fetal colon (Helander, 1973), but i To whom all correspondence should be addressed: Department of Anatomy, Harvard Medical School, 25 Shattuck Street, Boston, Mass. 02115.
the sequential changes involved have not been described in detail. In the adult small intestine, apical and basolateral membrane domains of columnar epithelial cells have distinct enzymatic profiles that are separated by occluding junctions (Hugon and Borgers, 1968; Fujita et al., 1972, Hopfer et al., 1975; Mircheff and Wright, 1976). Comparable membrane domains are present on epithelial cells of the adult rat colon, but in contrast to the enzyme-rich membranes of small intestine, the enzymatic activities of plasma membranes of colonic cells are relatively simple. Only two membrane-associated enzymes have been revealed by biochemical and EM cytochemical studies of adult rat and rabbit colon: cysteine-sensitive alkaline phosphatase activity (pH 9.0) is restricted to apical (luminal) membranes, whereas potassium-dependent, cardiac glycoside-sensitive phosphatase activity (pH 8.0) is localized on basolateral membranes (Vengesa and Hopfer, 1978; Vengesa and Hopfer, 1979). These two enzymes may thus be used as “markers” for determining the establishment and maintenance of apical and basolateral membrane domains of fetal rat colonic cells. 349 0012-1606/83 $3.00 Copyright All rights
0 1983 by Academic Press, Inc. of reproduction in any form reserved.
350
DEVELOPMENTAL
BIOLOGY
The time of appearance and the cellular localization of phosphatases on epithelial plasma membranes in the fetal rat colon are not known. Alkaline phosphatase activity has been observed on the luminal membranes of late fetal rat and mouse small intestine (Hugon and Borgers, 1969; Shervey, 1973; Calvert et aZ., 1981) and of first trimester human fetuses (Jirasek et aZ., 1975; Lev et all, 19’72; Kelley, 1973). There was no information, however, on the distribution of marker phosphatases on cell surfaces in the fetal rat colon in the early stratified epithelium or during the epithelial conversion process when new luminal surfaces and new junctional complexes appear. In the present study, the time of appearance and ultrastructural localization of two plasma membrane phosphatases, active at pH 8.0 and pH 9.0, respectively, were determined by biochemical and EM cytochemical techniques in fetal rat colon between 16 days gestation and birth (22 days). Our results demonstrate the initial appearance of plasma membrane phosphatase activities in the stratified epithelium of l&day fetuses. Enzymatic differences between luminal and nonluminal membranes are already established at the onset of enzyme activity in the stratified epithelium, and are maintained throughout the complex process of conversion to a simple columnar epithelium. MATERIALS
AND
METHODS
Animals Timed pregnant Sprague-Dawley rats (Charles River Breeding Laboratories, Wilmington, Mass.) between 16 and 22 days gestation (day of birth) were anesthetized either by ether or by ip injection of 25% urethane in phosphate buffered saline (PBS). The fetuses were exposed through a vertical slit in the mother’s abdominal wall and a small incision in the uterine horn. Each fetus was removed, the spinal cord was cut, and the entire colon was dissected intact. The colons were cut into thirds; proximal and distal segments were processed separately for EM cytochemistry, biochemical enzyme assay, or freeze fracture as described below.
EM Cytochemistry Segments of ascending and descending colon was sliced at 0.5 to 1.0 mm with a razor blade and fixed for 15-20 min in a solution of 1% paraformaldehyde and 0.25% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, at 23°C. Slices were rinsed three times for a total of 20 min in 250 mM Tris-HCl buffer, pH 8.5, and then incubated for cytochemistry according to the method of Ernst (1975) as modified for colonic tissue by Vengesa and Hopfer (1979). The incubation media for both enzymes consisted of 20 mM p-nitrophenylphosphate (pNPP; Sigma Chemical Co., St. Louis, MO.), 20 mM
VOLUME
97. 1983
SrClz, and 20 mM MgClz in 250 mM Tris-HCl buffer. Alkaline phosphatase activity was also determined using 20 mM /3-glycerophosphate (/3-GP; Sigma) as substrate in place of pNPP. The pH of the medium was adjusted to 9.0 for alkaline phosphatase and to 8.0 for detection of K+-dependent ATPase. For K+-dependent phosphatase, the medium also contained 30 mM KC1 and 10 mM cysteine. In the pH 8.0 control medium the KC1 was replaced with 10 mM choline chloride, or alternatively, 10 mM ouabain (Sigma) was added. The control medium for alkaline phosphatase (PH. 9.0) contained 10 mM cysteine (Sigma). The tissue slices were incubated in covered-glass petri dishes in a shaking water bath at 37°C for 40 min, rinsed with three changes of 125 mM Tris-HCl buffer at 23°C for a total of 20 min, treated with 2% lead nitrate for 20 min, rinsed twice for 5 min each in 250 mM sucrose to remove the excess lead, and then rinsed for 10 min in 125 mM Tris-HCl buffer. Slices were then rinsed in 0.1 M cacodylate buffer, pH 7.4, for 10 min, postfixed for 30 min with 1% OsOl in 0.1 M cacodylate buffer, pH 7.4, at 4”C, dehydrated in graded ethanol, and embedded in Epon-araldite or Polybed (Ladd).
Biochemical
Analysis
of Cytochemical
Media
Since fixatives inhibit phosphatase activity (Ernst, 1975), parallel experiments were done with fetal colonic slices placed in 125 mMTris-HCl buffer, pH 8.5, without prior fixation. The phosphatase activity of the tissue slices was quantitated by performing the cytochemical incubations at pH 8.0 and 9.0 as described above and then measuring spectrophotometrically the amount of cleaved substrate (PNP) released into the medium. Tissue slices from two to four fetuses were pooled in a single assay tube and were preincubat.ed in 1.4 ml of the appropriate incubation medium (as above) without pNPP for 30-60 min. Substrate was then added to each tube to a final concentration of 20 mM and the tubes were placed in a shaking water bath at 37°C for 40 min. Blanks for each assay tube contained medium plus pNPP, but no tissue. After incubation, the tubes were spun in an IEC clinical centrifuge for 1 min to pellet the tissue and the concentration of PNP in the supernatant was determined by spectrophotometric absorbance at 420 nm. The tissue pellet was resuspended in 0.9% NaCl, sonicated, and the protein concentration was determined according to Lowry (1951) using BSA as a standard. Enzyme activity was expressed as millimoles of PNP liberated per milligram of protein per 40 min.
Freeze Fracture Ascending and descending portions of fetal colon were separated and fixed for 30 min at 23°C in a solution of 2% formaldehyde, 2.5% glutaraldehyde, and 0.4% CaClz
COLONY
AND
NEUTRA
Phosphatases in Fetal Rat Colon
in 0.1 M Na-cacodylate buffer, pH 7.4. Tissue was then washed in 0.1 M cacodylate buffer, equilibrated with 20% glycerol in the same buffer for 3 hr, sliced at 1 mm, mounted on gold disks, briefly immersed in the liquid phase of partially frozen Freon 22, and stored in liquid nitrogen. Specimens were fractured in a Balzers freeze-etch device (Model BA 360) at a stage temperature of -115”C, replicated with platinum-carbon, cleaned with bleach, and mounted on 300-mesh grids. All luminal borders present in replicas of tissue from 17-, and 19-, and 21-day fetal colons were photographed at 20,000X with a JEOL 100 CX electron microscope. Cat&nixed
Ferritin
To define the main lumen and all of its extensions, cationized ferritin was used as an ultrastructural tracer. Ascending and descending colons were excised from 17to 22-day fetuses, fixed for 45 min at 23°C as described above, rinsed in 0.1 M cacodylate buffer, cut in 4-mm cylindrical segments, and immersed in a solution of 5.5 mg/ml cationized ferritin in 0.015 M NaCl at 4°C for 15 min with intermittent agitation. Segments were then rinsed for 20 min in four changes of cacodylate buffer and returned to the fixative solution for an additional 1 hr at 23°C. The central portion of each tissue cylinder was sliced into 0.5-mm pieces that were then postfixed in 1% OsOl in 0.1 M cacodylate buffer at 4°C for 1 hr, dehydrated, and embedded as described above. For electron microscopy, thin sections were briefly stained with lead citrate. RESULTS
Epithelial
Conversion
in the Fetal Colon
A light microscopic overview of the conversion process is shown in Fig. 1. At 16 days gestation, the colonic epithelium is uniformly stratified and consists of two to four layers of undifferentiated cells lining a narrow flattened lumen (Fig. la). By 17 days, several longitudinal cleft-like extensions of the lumen contact the basal cell layer, giving the lumen a stellate profile in cross section (Fig. lb). Within the thick layer of undifferentiated stratified cells between the clefts, small secondary lumina appear. During the 18th day, both the main lumen and the secondary lumina expand. By 19 days (Fig. lc), some of the apical cell layers form cellular “bridges” over the secondary lumina. These apical cells are often rounded and darkly stained and appear to be shed during the 19th and 20th days (Fig. lc) as ridges of connective tissue invaginate the epithelium from below. By 21 days (Fig. Id) the loss of apical cells has resulted in the joining of the primary lumen with the secondary lumina, leaving a single large
lumen lined by simple columnar epithelium longitudinal ridges of connective tissue.
351 covering
Occluding Junctions At all developmental stages examined (17 to 21 days) primary and secondary luminal surfaces were recognized in freeze-fracture replicas by the presence of typical microvilli. All primary and secondary lumina observed were bordered by continuous occluding junctions (Fig. 2). The continuity of these junctions was confirmed by the observation that cationized ferritin introduced into the main lumen did not enter secondary lumina (Fig. 3). Thus, luminal and nonluminal domains were consistently separated by occluding junctions throughout the conversion process. EM Cytochemistry No phosphatase activity, either at pH 8.0 or pH 9.0, was cytochemically demonstrable in the stratified epithelium of fetal colons at 16 or 17 days (Fig. 4a). Plasma membranes lining the main lumen and the small secondary lumina, as well as nonluminal membranes, were consistently devoid of reaction products. Both pH 8.0 and pH 9.0 phosphatase activities were detectable at 18 days gestation (Figs. 4b and c, 5). In tissues incubated at pH 9.0, phosphatase reaction product was observed along the plasma membranes contacting the main lumen (Fig. 4b) and the secondary lumina (Fig. 4~). A lighter precipitate was observed along basolateral membranes as well (Fig. 4~). After incubation at pH 8.0 in the presence of K+ and cysteine, reaction product appeared along nonluminal membranes of some epithelial cells distributed throughout the stratified epithelium but plasma membranes bordering either the main or the secondary lumina were devoid of dense deposits (Fig. 5). By 19 days gestation the cytochemical reactions were stronger than at 18 days, particularly in the epithelium of the ascending colon (Figs. 6-8). At pH 9.0, cysteinesensitive phosphatase activity was localized along apical membranes bordering the main lumen and its clefts as well as secondary lumina (Fig. 6). Local unreactive cells or groups of cells were common. Small and variable amounts of activity were also observed on basolateral membranes (Fig. 7). Alkaline phosphatase activity was consistently stronger on epithelial cells of the ascending colon, although its distribution was similar in both ascending and descending regions. The reaction product was more uniform and dense when ,&glycerophosphate was used as substrate instead of pNPP. Phosphatase activity at pH 8.0 was distributed on nonluminal membranes of cells at all levels of the epithelium at 19 (Fig. 8) and 20 days gestation. Although
352
DEVELOPMENTALBIOLOGY
V0~~~~97,1983
of fetal rat colon, demonstrating the conversion of the epithelium from stratified to simple columnar. (a) 16 FIG. 1. Light micrographs days gestation. A stratified epithelium, two to three cell layers thick, surrounds a small central lumen (L). X600. (b) 17 days. The main lumen (ML) has expanded to form deep clefts (arrows) that extend down to the basal cell layer. Within the stratified epithelium between the clefts, a small secondary lumen (SL) has formed. X600. (c) 19 days. The secondary lumena (SL) are enlarged. The darkly stained, rounded apical cells (arrows) will be sloughed into the main lumen (ML). The basal side of the epithelium is invaginated by ridges of mesenchyme (*). X600. (d) 22 days. A simple columnar epithelium lines longitudinal ridges of connective tissue. Differentiated goblet cells (GC) are interspersed among absorptive cells. x500.
COLONY
AND NEUTRA
Phosphtases
FIG. 2. Freeze-fracture replicas of occluding junctions on cells bordering (c) 21 days. The organization of occluding junctions is variable and changes X48.000.
luminal membranes generally showed no activity (Fig. 8, inset), reaction product was occasionally observed on the apical membrane of cells contacting a lumen, even in samples incubated at pH 8.0 in the presence of KC1 and cysteine. Cells lacking a luminal surface had reaction product distributed over the entire cell surface. Most deposits were located on the cytoplasmic side of the plasma membrane, although precipitate was also observed on the extracellular side. Addition of cysteine reduced but did not abolish the extracellular reaction product. When ouabain was added to the incubation
in
Fetal
the main lumen with increasing
Rat
Colon
353
of fetal rat colon, at (a) 17 days, (b) 19 days, and age, but they appear to be continuous at all ages.
medium, or when K+ was omitted, the extracellular deposits persisted and the deposits on the cytoplasmic side of the membrane were only partially reduced. In the descending colon, no consistent basolateral membrane phosphatase activity at pH 8.0 was detected at any gestational age. At 21 and 22 days gestation, phosphatase activity at pH 9.0 was localized primarily on the luminal membranes of the simple columnar cells (Fig. 9). At this age phosphatase activity at pH 8.0 in the ascending colon was visualized as distinct crystalline deposits located
354
DEVELOPMENTAL BIOLOGY
VOLUME 97, 1983
chemically detectable levels. A significant portion of the alkaline phosphatase activity detectable at 18 and 19 days was cysteine insensitive (Figs. lla and b). Intestinal-type alkaline phosphatase activity, which is cysteine sensitive, was low at these early ages but increased to a maximum at 20 days in descending colon (Fig. llb) and at 21 days in ascending colon (Fig. lla). The apparent decline in activity just prior to birth (22 days) may reflect the rapid growth in mass of nonepithelial elements of the intestinal wall (such as muscle and connective tissue) at this age. Within 2 hr after birth, cysteine-sensitive alkaline phosphatase activity increased dramatically in the proximal colon (Fig. lla), whereas levels in the distal colon remained low (Fig. llb). Phosphatase activity at pH 8.0 was not significantly different in ascending and descending colon so the data for the two regions were combined (Fig. 12). Activity appeared at 18 days gestation, increased to an apparent maximum at 20 days, and then declined to a level that remained relatively constant through the day of birth. As with alkaline phosphatase, the apparent decline in activity at 21 days coincided with a rapid growth of nonepithelial elements of the intestinal wall, and thus with an increase in total protein. No significant inhibition of pH 8.0 activity was observed when K+ was omitted from the medium (and substituted by choline) or when 10 mM ouabain was added. DISCUSSION
FIG. 3. Electron micrograph of a layer of apical epithelial cells in the colon of a 19-day fetal rat. Cationized ferritin was infused into the main lumen. This cell layer (seen as dark cells in Fig. lc) separates the main lumen from a secondary lumnen. Cationized ferritin (CF) is present on the membranes exposed to the main lumen, but the tracer has not entered the secondary lumen. X44,709.
primarily on the cytoplasmic face of the interdigitated lateral membranes of simple columnar absorptive cells (Fig. 10). Activity on luminal membranes was rarely observed. Addition of ouabain or deletion of K+ ions again reduced but did not eliminate the deposits on the cytoplasmic side of the basolateral membranes. Biochemical Analysis of Cytochemical Media
The onset of phosphatase activity at both pH 8.0 and pH 9.0 was detected biochemically at 18 days gestation, the same age at which enzyme activity reached cyto-
Alkaline phosphatase and K+-dependent ATPase are considered marker enzymes for luminal and contraluminal plasma membrane domains in a variety of transporting epithelia (DiBona and Mills, 1979; Ernst and Mills, 1980; Ernst and Hootman, 1981) including the small intestine (Stirling, 1972; Hopfer et a,!., 1975; Ernst, 1975). The colonic epithelium also transports ions and water, but the histochemical localization of phosphatases in both neonatal and adult colon has been controversial (Deane and Dempsey, 1945; Masnerova et al, 1966; Helander, 1975; Ono, 1976). The contradictory data may be ascribed to the low levels of colonic phosphatase activities, relative to those found in the kidney or small intestine (Vengesa and Hopfer, 1979; Lev and Griffiths, 1982) as well as to the variable sensitivity and specificity of cytochemical methods (Ernst and Hootman, 1981). More recently, both phosphatases have been detected on the plasma membranes of epithelial cells of adult rat and rabbit colon by sensitive ultrastructural cytochemistry and cell fractionation (Vengesa and Hopfer, 1978, 1979). These studies demonstrated mutually exclusive localization of the two enzymes to the two major membrane domains. We have now demonstrated that plasma membrane
FIG. 4. Fetal rat ascending colon, reacted cytochemically to demonstrate alkaline phosphatase activity (pH 9.0) at 1’7 and 18 days gestation. (a) 17 days. This field includes a portion of the main lumen (ML) and a small secondary lumen (SL) in the stratified epithelium. No alkaline phosphatase activity is detectable. Some nonspecific precipitate is present in mitochondria. X14,000. (b) 18 days. Reaction product is localized along the apical plasma membranes of cells bordering the main lumen (ML) Substrate: pNPP. X16500. (c) 18 days. A secondary lumen (SL) deep in the stratified epithelium shows alkaline phosphatase activity on luminal plasma membranes. A smaller amount of reaction product is also seen on nonluminal membranes (arrow). Substrate: @GP. X15,900. 355
356
DEVELOPMENTAL BIOLOGY
VOLUME 9’7, 1983
FIG. 5. Colonic epithelium of an l&day fetus, reacted for phosphatase activity at pH 8.0 in the presence of K+ and cysteine. Reaction product (arrows) is seen along the plasma membranes of some cells deep in the stratified epithelium. No activity is present on membranes bordering the secondary lumen (SL). X10,400.
phosphatases are present in the fetal rat colon, appearing at 18 days gestation while the epithelium is still stratified. These membrane phosphatases remain active during the period of rapid epithelial reorganization (between 18 and 21 days gestation). One day before birth, a simple columnar epithelium is established and the cellular localization of both phosphatases is similar to that seen in the adult colonic epithelium. Our cytochemical data do not rule out the possibility that phosphatases may be present prior to 18 days in an inactive form. Immunohistochemical studies of small intestine have detected immunoreactive alkaline phosphatase in neonatal epithelium and in crypt cells that is not enzymatically active (Etzler and Moog, 1966; Wachsmuth and Torhorst, 1974; Shields et aL, 1982). In the adult small intestinal epithelium, alkaline phosphatase activity is not restricted to the luminal plasmalemma. Membrane fractionation and histochemical studies have consistently revealed low levels of alkaline phosphatase activity on basolateral membranes (Hugon and Borgers, 1968; Ono, 1974, 19’75; Hanna et aL, 1979). Our EM cytochemical data indicate that the distribution of colonic alkaline phosphatase in the fetal rat resembles that of the adult small intestine and of
postnatal rat colon (Helander, 1975; Ono, 1976), as it is located primarily on the apical membrane but is also present to a lesser extent on basolateral membranes. The cytochemical reaction products localized along both apical and basolateral membranes of fetal colonic epithelium after incubation at pH 9.0 reflect, at least in part, intestinal-type alkaline phosphatase activity according to criteria established in previous histochemical and biochemical studies: the precipitate is localized on the outer membrane leaflet (Ono, 1973; Ernst, 1975), is inhibited by cysteine (Van Belle, 1976; Vengesa and Hopfer, 1979), and shows the same distribution with either p-nitrophenylphosphate (pNPP) or P-glycerophosphate (P-GP) as substrate. The effectiveness of 8GP as substrate further confirms the presence of alkaline phosphatase, since other phosphatases such as 5’-nucleotidase, Ca-dependent ATPase and Na/K ATPase do not cleave this substrate. Moog and Yeh (1973) reported that in the suckling rat small intestine and in the crypt of adult intestine P-GP was cleaved more effectively than pNPP, whereas pNPP was the preferred substrate in cells at the adult villus tip. In the fetal colon, we observed a stronger and more consistent cytochemical reaction in tissue incubated at pH
COLONY
AND NEUTRA
Phosphutmes
in
Fetal
Rat
Colon
357
FIG. 6. 19-day fetal ascending colon, alkaline phosphatase activity. The apical plasma membranes of most of the cells that contact the main lumen show strong phosphatase activity at pH 9.0. Although cells lining the luminal clefts in this field are not reactive, many cleft cells did show phosphatase activity (see Fig. ‘7). Occasional cells (see lower center) show strong nonluminal enzyme activity; whether this is a sign of impending cell death is not known. Substrate: @GP. X24,000.
9.0 with P-GP as substrate. These data suggest that an isoenzyme that preferentially cleaves /3-GP may be characteristic of immature cells. The biochemical data confirm the onset of alkaline phosphatase activity at 18 days gestation in fetal colon, Total activity increases to a peak at 20 days in the descending and 21 days in the ascending colon, coincident with the completion of the conversion of the stratified
epithelium to simple columnar. Just after birth, activity in ascending colon increases dramatically, presumably reflecting the morphological differentiation of an alkaline phosphate-positive apical tubulovesicular system in the surface columnar cells (Masnerova et aL, 1966; Ono, 1976). This system does not develop in absorptive cells of the descending colon (Ono, 1976, 1977) and thus, no rise in enzyme activity is noted after birth
358
DEVELOPMENTAL
BIOLOGY
FIG. ‘7. 19-day fetal ascending colon, alkaline phosphatase activity. Luminal plasma membranes of cells lining a cleft of the main lumen are strongly reactive, but weaker phosphatase activity is also seen on lateral membranes below the occluding junctions (OJ). In both locations, dense reaction product is restricted to the extracellular side of the membrane. Substrate: p-GP. X27,000.
in this region. Only a portion of the total activity at any age is cysteine sensitive and thus represents intestinal epithelial cell activity. The cysteine-insensitive activity measured biochemically is presumably due to the cytochemically demonstrable enzyme activity on capillary endothelia in the lamina propria (Lev and Griffiths, 1982). Phosphatase activity at pH 8.0 was detected cytochemically in fetal rat colon under the conditions known to demonstrate the K+-dependent phosphatase com-
VOLUME
97,1983
ponent of the Na+/K+ ATPase enzyme complex in other tissues (Ernst, 1972a,b; 19’75; Ernst and Hootman, 1981). As in other tissues, the reaction product in fetal rat colon was concentrated along nonluminal membranes, was located at least in part on the inner leaflet of the plasma membrane, and was not abolished by cysteine. The positive identification of basolateral Na+/K’ ATPase in fetal colon is seriously complicated, however, by the finding that enzyme activity persisted as determined biochemically and as visualized cytochemically, in the absence of K+ ions or in the presence of ouabain. In previous studies, omission of K+ or addition of cardiac glycosides eliminated detectable activity in most adult tissues (Ernst and Philpott, 1970; Ernst, 1972a,b; Vengesa and Hopfer, 1979). In contrast, K+-dependent pNPPase in rat tissues was found to be relatively insensitive to ouabain (Allen and Schwartz, 1969; Ernst, 1975). Ernst further showed that loss of sensitivity of the ATPase complex to ouabain occurs when the native substrate ATP is substituted by pNPP, and also in the presence of 20 mM strontium. Thus a combination of these effects may have reduced the ouabain sensitivity of the fetal rat enzyme in the present study to undetectable levels. Nevertheless, pNPPase activity at pH 8.0 was completely abolished by the cardiac glycoside scillaren and by omission of K+ ions in a previous study of adult rat and rabbit colon (Vengesa and Hopfer, 1979). The insensitivity of pH 8.0 pNPPase to ouabain and to omission of K+ in the fetal colon could also be due in part to an intrinsically low level of activity that has been further reduced by fixation and by the presence of strontium (Ernst, 1972a,b). In addition, the sensitivity of the cytochemical technique may be reduced by diffusion of reaction product, inefficient trapping of the released phosphate, and solubilization of the SrP04 precipitate at pH 8.0 (Ernst, 1972a,b). Finally, although controls were preincubated without substrate in K+free media, sufficient K+ ions may have been retained in the tissue slices to maintain low levels of pNPPase activity. Thus, the exact identity of the fetal basolateral enzyme activity is not established, but at least part of the cytochemical reaction product on the inner membrane leaflet was decreased under inhibitory conditions and may reflect Na+/K+ ATPase activity. The secondary lumina which form within the stratified epithelium of the fetal colon separate entire layers of superficial cells from the cells below. These cells form barriers between primary and secondary lumina and may have two luminal surfaces, both sealed by tight junctions. One surface contacts the main lumen and is labeled by a luminal infusion of cationized ferritin (Fig. 3), whereas the other contacts a closed intraepithelial space. Since both of these surfaces acquire alkaline phosphatase activity simultaneously, the induction of
F IG. 8. 19-day fetal ascending colon; pH 8.0 phosphatase activity in the presence of K+ and cysteine. Enzyme the entire plasma membrane of basal cells that do not contact a lumen. Reaction product is located principally the membrane. Inset: Cells that border a luminal cleft have reaction product only on nonluminal membranes. prel cipitate is on the extracellular side of the membrane. X14,300.
359
activity is distributed olver on the cytoplasmic side ? of In this field, much of the
360
DEVELOPMENTAL
BIOLOGY
VOLUME 97, 1983
that do not contact a luminal surface at all. The entire plasma membrane of these cells exhibits a pattern of phosphatase activities similar to basolateral membrane of adult intestinal cells; i.e., low levels of cysteine-sensitive pH 9.0 phosphatase on the outer leaflet and pH 8.0 pNPPase on the inner leaflet. The stratified fetal colonic epithelium transiently resembles the stratified transporting epithelium of frog skin (Mills et aL, 1977; DiBona and Mills, 1979) in that pH 8.0 phosphatase activity is distributed over the entire plasma membrane of basal cells. It is generally accepted that the continuity of occluding junctions is crucial for maintaining the separate identity of luminal and contraluminal domains of et aL, 1978; epithelial plasma membranes (Meldolesi studies of fetal Dragsten et al, 1981). Freeze-fracture small intestinal epithelium (Madara et aL, 1981; Teillet
FIG. 9. 22-day fetal ascending colon, alkaline phosphatase activity. Enzyme activity on simple columnar epithelial cells appears principally on the luminal microvillus membrane (mv) hut is also present on basolateral membranes (arrow). Substrate: @-GP. X7700.
enzyme activity apparently does not depend on exposure of the cell surface to the milieu of the main lumen. Deeper in the stratified epithelium, there are cells
FIG. 10. 22-day fetal ascending colon, pH 8.0 phosphatase activity in the presence of K’ ions and cysteine. The dense crystalline precipitate produced by enzyme activity is present in the interdigitated lateral membranes of two adjacent columnar cells. The reaction product is located on the cytoplasmic side of the membranes. X29,100.
COLONY AND NEUTRA
Phosphatases
Qa Fetal
Rat
361
Colon
lla FETAL RAT ASCENDING ON PHOSPHATASE ACTIVITY, pH 9.0
0.6
16
19
21
22
Fetal Age (In days)
L 4 nc born Mrth
(2 hrs)
llb FETAL RAT DESCENDING COLON PHOSPHATASE ACTIVITY, pH 9.0
0.6 0.5 L E s c 3 g
0
Total Actlvlty
a Cyatolne
Added
0.4 0.3
,P B 0.2 a =
0.1
I8
I9
!O
21
22
FIG. 11. Alkaline phosphatase activity of fetal colonic slices, measured spectrophotometrically as amount of cleaved substrate (PNP) released into the medium per mg tissue protein, during a 40 min incubation period. Activity is present at 18 days gestation and increases to a peak at 21 days in ascending colon (a) and 20 days in descending colon (b). The apparent decline prior to birth may be due to rapid growth of nonepithelial tissues. The cysteine-sensitive portion of the total activity represents specific intestinal epithelial alkaline phosphatase; the cysteine-insensitive portion is presumably due to alkaline phosphatase on other cell types.
et aL, 1981) and of fetal colonic epithelium in this study have shown that all lumina, whether main or secondary, are bordered by continuous occluding junctions. Although junctions on fetal cells appear more loosely organized than those seen in adult intestinal epithelium,
they display no visible discontinuities. Presumably, this prevents the lateral diffusion of fetal membrane proteins in the plane of the membrane between luminal and contraluminal domains. It is interesting that in the fetal colon these two domains are separated by tight
362
DEVELOPMENTAL
BIOLOGY
VOLUME
97, 1983
12 FETAL
0.6 o 5 .G 0: o 4. . * c -Z
RAT
COIBN (total)
PHOSPHATASE 0
Total Activity
q q
K* Omitted Ouobain
I6
ACTIVITY,
Added
19
21
Fetal Age (in days) FIG. 12. Phosphatase activity at pH 8.0 in fetal reaches a peak at 20 days. The apparent decline phosphatase activity was not reduced by omission
pH .6.0
22
b newborn (2 hrr)
birth
colon slices, measured and expressed as in Fig. 12. Enzyme activity after 20 days is presumably due to rapid growth of nonepithelial at Kt ions from the medium, or by addition of 10 mM ouabain.
junctions prior to the appearance of membrane phosphatase activities, and that when these phosphatases first appear, they are already segregated. The excellent technical assistance of Ms. Teresa Hall is gratefully acknowledged. This work was supported in part by Grant AM 21505 from the National Institutes of Health. Dr. Colony was recipient of a Senior Research Fellowship from the National Foundation for Ileitis and Colitis. Dr. Neutra is recipient of Research Career Development Award AM 00407 from the National Institutes of Health. The JEOL 100 CX electron microscope was funded by National Science Foundation Grant PCM 7816158.
REFERENCES ALLEN, J. C., and SCHWARTZ, A. (1969). A possible biochemical explanation for the insensitivity of the rat to cardiac glycosides. J. Pharmacol. Exp. Ther. 168, 42-46. CALVERT, R., MALKA, D., and MENARD, D. (1981). Establishment of regional differences in brush border enzymatic activities during the development of fetal mouse small intestine. Cell Tissue Res. 214, 97-106. DAHLQUIST, A., and LINDBERG, T. (1966). Fetal development of the small intestinal disaccharidase and alkaline phosphatase activities in the human. Biol. Neonat. 9,24-32. DEANE, H. W., and DEMPSEY, E. W. (1945). The localization of phosphatase in the Golgi region of intestinal and other epithelial cells. Anat. Rec. 93, 401-417. DIBONA, D. R., and MILLS, J. W. (1979). Distribution of Na+-pump sites in transporting epithelia. Fed Proc. 38, 134-143. DRAGSTEN, P. R., BLUMENTHAL, R., and HANDLER, J. S. (1981). Membrane assymetry in epithelia: Is the tight junction a barrier to diffusion in the plasma membrane? Nature (London) 294, 718-722. ERNST, S. A. (1972a). Transport adenosine triphosphatase cytochemistry I. Biochemical characterization of a cytochemical medium for the ultrastructure localization of ouabain-sensitive, potassium-de-
is low at 18 days and tissues. At all ages,
pendent phosphatase activity in the avian salt gland. J. His&hem. Cytochem. 20, 13-22. ERNST, S. A. (1972b). Transport adenosine triphosphatase cytochemistry. II. Cytochemical localization of ouabain-sensitive, potassiumdependent phosphatase activity in the secretory epthelium of the avian salt gland. .J. Histochem. Cytochem. 20,23-38. ERNST, S. A. (1975). Transport ATPase cytochemistry: Ultrastructural localization of potassium-dependent and potassium-independent phosphatase activities in rat kidney cortex. J. Cell Biol. 66,586-608. ERNST, S. A., and HOOTMAN, S. R. (1981). Microscopical methods for the localization of Na+/K’-ATPase. Histochem. J. 13, 397-418. ERNST, S. A., and MILLS, J. W. (1980). Autoradiographic localization of tritiated ouabain sensitive sodium pump sites in ion transporting epithelia. J. Histochem. Cytochem. 28, 72-77. ERNST, S. A., and PHILPOTT, C. W. (1970). Preservation of Na-K-activated and Mg-activated adenosine triphosphate activities of avian salt gland and teleost gill with formaldehyde as fixative. J. Histothem. Cytochem. 18,251-263. ETZLER, M. E., and MOOG, F. (1966). Inactive alkaline phosphatase in duodenum of nursling mouse: Immunological evidence. Science 154, 1037-1038. FUJITA, M., OHTA. H., KAWAI, K., MATSUI, H., and NAKAO, M. (1972). Differential isolation of microvillous and basolateral plasma membranes from intestinal mucosa: Mutually exclusive distribution of digestive enzymes and ouabain-sensitive ATPase. B&him. Biophys. Acta 274, 336-347. HANNA, S. D., MIRCHEFF, A. R., and WRIGHT, E. M. (1979). Alkaline phosphatase of basal-lateral and brush border plasma membranes from rat intestinal epithelium. J. Suprumol Struct. 11,451-466. HELANDER, H. F. (1973). Morphological studies on the development of the rat colonic mucosa. Acta Anat. 85, 153-176. HELANDER, H. F. (1975). Enzyme patterns and protein absorption in rat colon during development. Aeta Anat. 91,330-349. HOPFER, U., SIGRIST-NELSON, K., and MURER, H. (1975). Intestinal sugar transport: Studies with isolated plasma membranes. Ann. N. Y. Acad. SC%. 264,414-426. HUGON, J., and BORGERS, M. (1968). Fine structural localization of
COLONY AND NEUTRA
Phosphatases
acid and alkaline phosphatase activities in the absorbing cells of the duodenum of rodents. Histochemie 12,42-66. HUGON, J. S., and BORGERS,M. (1969). Ultrastructural differentiation and enzymatic localization of phosphatase in the developing duodenal epithelium of the mouse. I. The foetal mouse. Histochemie 19,13-30. JIRASEK, J. E., UHER, J., and KOLDOVSKY, 0. (1975). A histochemical analysis of the development of the small intestine of human fetuses. Actu
Hi&o&em
22,33-39.
KELLEY, R. (1973). An ultrastructural and cytochemical study of developing small intestine in man. J. Embryol Exp. Morph.& 29.411430.
LEV, R., and GRIFFITH& W. C. (1982). Colonic and small intestinal alkaline phosphatase: A histochemical and biochemical study. Gastroenter~ 82,1427-1435. LEV, R., SEIGEL, H. I., and BARTMAN, J. (1972). Hi&chemical studies of developing human fetal small intestine. Histdwmie 29.103-119. LOWRY, 0. H., ROSEBROUGH,N. J., FARR, A. L., and RANDALL, R. J. (1951). Protein measurement with the folin-phenol reagent. J. CeU Bid Chem. 193, 265-275. MADARA, J. L., NEUTRA, M. R., and TRIER, J. S. (1981). Junctional complexes in fetal rat small intestine during morphogenesis. Dew. Bid 86, 170-178. MASNEROVA, M., KOLDOVSKY, O., and KUBAT, K. (1966). Postnatal changes in phosphatase and non-specific esterase activity in the large intestine of the rat. Ewentia (Basel) 22,518-519. MATHAN, M., MOXEY, P. C., and TRIER, J. S. (1976). Morphogenesis of fetal rat duodenal villi. Amer. J. And. 146, 73-92. MELDOLESI, J., CASTIGLIONI, G., PARMA, R., NASSIVERA, N., and CAMILLI, P. (1978). Ca++-dependent disassembly and reassembly of occluding junctions in guinea pig pancreatic acinar cells. J. CeU Bid 79, 156-172. MILLS, J. W., ERNST, S. A., and DIBONA, D. R. (1977). Localization of Na+ pump sites in frog skin. J. CeU Bid 73.88-110. MIRCHEFF, A. K., and WRIGHT, E. M. (1976). Analytical isolation of plasma membranes of intestinal epithelial cells: Identification of Na, K-ATPase rich membrane and the distribution of enzyme activities. J. Menrbr. Bid 28, 309-333. MOOG, F., and YEH, KWO-YIH. (1973). Intestinal alkaline phosphatase of the rat: Development and distribution of activity with phenyl-
in
Fetal
Rat
Colon
363
phosphate and fl-glycerophosphate. Camp. Biochem. Physiol. 44B: 6457-666. ONO,K. (1973). The fine structural localization of alkaline phosphatase activity of intestinal microvilli in the developing chick embryo. Acta Anat
86,71-82.
ONO, K. (1974). The localization of alkaline phosphatase activity in the rat small intestinal epithelium after fat and protein feeding. Acta Hi&&m. 51,124-137. ONO, K. (1975). The fine structure and localization of alkaline phosphatase activity of the small intestinal epithelium in the postnatal developing rat. Acta His&hem+ S2,117-133. ONO, K. (1976). Alkaline phosphatase activity of large intestinal principal cells in postnatal developing rats. Acta His&hem. 57, 312319. ONO, K. (1977). Absorption of horseradish peroxidase by the principal cells of the large intestines of postnatal developing rats. Anat. EmiqoL 151, 53-62. SHERVEY, P. D. (1973). A histochemical study of the inclusion bodies in the small intestine of the developing rat. Ziet Anat. Entticklung. 141,39-54. SHIELDS, H. M., BAIR, F. A., BATES, M. L., YEDLIN, S. T., and ALPERS, D. H. (1982). Localization of immunoreactive alkaline phosphatase in the rat small intestine at the light microscopic level by immunocytochemistry. Gastroenterology 82,32-45. STIRLING, C. E. (1972). Radioautographic localization of sodium pump sites in rabbit intestine. J. Cell Biol. 53, 704-714. TEILLET, M. A., HUGON, J. S., and CALVERT, R. (1981). The occluding junctions of mouse duodenal enterocytes during development. A freeze fracture study. Cell Tissue Res. 217, 65-78. VAN BELLE, H. (1976). Kinetics and inhibition of rat and avian alkaline phosphatase. Ga PharmacoL 7, 53-58. VENGESA, P. B., and HOPFER, U. (1978). A localization of alkaline phosphatase and Na+-pump sites in adult rat and rabbit colon by cytochemical and cell fractionation techniques. J. Cell Biol. 79,238a. VENGESA, P. B., and HOPFER, U. (1979). Cytochemical localization of alkaline phosphatase and Na+ pump sites in adult rat colon. J. Hi&o&em Cytochem. 27,1231-1235. WACHSMUTH, E. D., and TORHORST, A. (1974) Possible precursors of aminopeptidase and alkaline phosphatase in the proximal tubules of kidney and the crypts of small intestine of mice. Histochemistly 38,43-56.
BIOLOGY
97,349-363
(1983)
Epithelial Differentiation
in the Fetal Rat Colon
I. Plasma Membrane PAMELA Pennsylvania
State
University, Received
C. COLONY
Phosphatase
AND MARIAN R. NEUTRA’
Hershey, Pennsylvania October
Activities
Harvard
16802;
8, 1982; accepted
in revised
Medical fmm
School,
December
Boston,
Massachusetts
02115
30, 1982
During the last week of gestation of the fetal rat, the epithelium of the colon is rapidly remodeled. At 16 days a primitive stratified epithelium surrounds a small central lumen. Over the next 3 days, the main lumen extends narrow clefts down to the basal cell layer and small secondary lumina appear within the stratified epithelium between these clefts. At 19 and 20 days, secondary lumina enlarge but remain discrete; an infusion of cationic ferritin into the main lumen does not enter secondary lumina. During the 2 days prior to birth (21-22), the secondary lumina join the main lumen as superficial cells are sloughed, and the epithelium becomes simple columnar. Freeze-fracture replicas indicate that luminal and nonluminal membrane domains of epithelial cell plasma membranes are separated by continuous tight junctions throughout the conversion process. Cytochemical analysis of tissue slices from 16- to 22-day fetal colon demonstrated the appearance and segregation of two phosphatases on apical and hasolateral membrane domains during epithelial conversion. Cysteine-sensitive pH 9.0 (alkaline) phosphatase activity was first detected along the luminal membranes of cells bordering both primary and secondary lumina at 18 days gestation and increased to a maximum at 20-21 days; weaker activity was present on basolateral membranes. Phosphatase activity at pH 8.0 also appeared at 18 days and increased thereafter, but was localized primarily on nonluminal membranes. At pH 8.0, reaction product appeared on both inner and outer sides of the membrane, and was only partially abolished by omission of K’ or addition of ouabain; thus the reaction may be only partially due to K+-dependent ATPase activity. Biochemical analysis of the cytochemical media confirmed the appearance of phosphatase activities at 18 days. Thus, plasma membrane phosphatase activities appear while the epithelium is still stratified, but are segregated to luminal and nonluminal membrane domains at the onset of activity. Segregation is maintained throughout the process of conversion of a simple columnar epithelium. INTRODUCTION
The epithelial lining of the rat small and large intestine is rapidly remodeled during the final week of gestation. At 16 days gestation, 6 days prior to birth, the entire intestine is lined by a primitive stratified epithelium surrounding a single central lumen. Cells bordering the lumen are joined by occluding junctions. By Day 21, the stratified cells have been replaced by a simple columnar epithelium with polarized differentiated cells. In the small intestine, this remodeling process involves the appearance of new secondary lumina deep in the stratified epithelium with the formation of new luminal cell surfaces and new occluding junctions on the deep epithelial cells bordering these lumina (Madara et aL, 1981), as well as invagination of the epithelium by underlying mesenchyme, and sloughing of superficial cells (Mathan et aL, 1976). Epithelial conversion also occurs in the fetal colon (Helander, 1973), but i To whom all correspondence should be addressed: Department of Anatomy, Harvard Medical School, 25 Shattuck Street, Boston, Mass. 02115.
the sequential changes involved have not been described in detail. In the adult small intestine, apical and basolateral membrane domains of columnar epithelial cells have distinct enzymatic profiles that are separated by occluding junctions (Hugon and Borgers, 1968; Fujita et al., 1972, Hopfer et al., 1975; Mircheff and Wright, 1976). Comparable membrane domains are present on epithelial cells of the adult rat colon, but in contrast to the enzyme-rich membranes of small intestine, the enzymatic activities of plasma membranes of colonic cells are relatively simple. Only two membrane-associated enzymes have been revealed by biochemical and EM cytochemical studies of adult rat and rabbit colon: cysteine-sensitive alkaline phosphatase activity (pH 9.0) is restricted to apical (luminal) membranes, whereas potassium-dependent, cardiac glycoside-sensitive phosphatase activity (pH 8.0) is localized on basolateral membranes (Vengesa and Hopfer, 1978; Vengesa and Hopfer, 1979). These two enzymes may thus be used as “markers” for determining the establishment and maintenance of apical and basolateral membrane domains of fetal rat colonic cells. 349 0012-1606/83 $3.00 Copyright All rights
0 1983 by Academic Press, Inc. of reproduction in any form reserved.
350
DEVELOPMENTAL
BIOLOGY
The time of appearance and the cellular localization of phosphatases on epithelial plasma membranes in the fetal rat colon are not known. Alkaline phosphatase activity has been observed on the luminal membranes of late fetal rat and mouse small intestine (Hugon and Borgers, 1969; Shervey, 1973; Calvert et aZ., 1981) and of first trimester human fetuses (Jirasek et aZ., 1975; Lev et all, 19’72; Kelley, 1973). There was no information, however, on the distribution of marker phosphatases on cell surfaces in the fetal rat colon in the early stratified epithelium or during the epithelial conversion process when new luminal surfaces and new junctional complexes appear. In the present study, the time of appearance and ultrastructural localization of two plasma membrane phosphatases, active at pH 8.0 and pH 9.0, respectively, were determined by biochemical and EM cytochemical techniques in fetal rat colon between 16 days gestation and birth (22 days). Our results demonstrate the initial appearance of plasma membrane phosphatase activities in the stratified epithelium of l&day fetuses. Enzymatic differences between luminal and nonluminal membranes are already established at the onset of enzyme activity in the stratified epithelium, and are maintained throughout the complex process of conversion to a simple columnar epithelium. MATERIALS
AND
METHODS
Animals Timed pregnant Sprague-Dawley rats (Charles River Breeding Laboratories, Wilmington, Mass.) between 16 and 22 days gestation (day of birth) were anesthetized either by ether or by ip injection of 25% urethane in phosphate buffered saline (PBS). The fetuses were exposed through a vertical slit in the mother’s abdominal wall and a small incision in the uterine horn. Each fetus was removed, the spinal cord was cut, and the entire colon was dissected intact. The colons were cut into thirds; proximal and distal segments were processed separately for EM cytochemistry, biochemical enzyme assay, or freeze fracture as described below.
EM Cytochemistry Segments of ascending and descending colon was sliced at 0.5 to 1.0 mm with a razor blade and fixed for 15-20 min in a solution of 1% paraformaldehyde and 0.25% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, at 23°C. Slices were rinsed three times for a total of 20 min in 250 mM Tris-HCl buffer, pH 8.5, and then incubated for cytochemistry according to the method of Ernst (1975) as modified for colonic tissue by Vengesa and Hopfer (1979). The incubation media for both enzymes consisted of 20 mM p-nitrophenylphosphate (pNPP; Sigma Chemical Co., St. Louis, MO.), 20 mM
VOLUME
97. 1983
SrClz, and 20 mM MgClz in 250 mM Tris-HCl buffer. Alkaline phosphatase activity was also determined using 20 mM /3-glycerophosphate (/3-GP; Sigma) as substrate in place of pNPP. The pH of the medium was adjusted to 9.0 for alkaline phosphatase and to 8.0 for detection of K+-dependent ATPase. For K+-dependent phosphatase, the medium also contained 30 mM KC1 and 10 mM cysteine. In the pH 8.0 control medium the KC1 was replaced with 10 mM choline chloride, or alternatively, 10 mM ouabain (Sigma) was added. The control medium for alkaline phosphatase (PH. 9.0) contained 10 mM cysteine (Sigma). The tissue slices were incubated in covered-glass petri dishes in a shaking water bath at 37°C for 40 min, rinsed with three changes of 125 mM Tris-HCl buffer at 23°C for a total of 20 min, treated with 2% lead nitrate for 20 min, rinsed twice for 5 min each in 250 mM sucrose to remove the excess lead, and then rinsed for 10 min in 125 mM Tris-HCl buffer. Slices were then rinsed in 0.1 M cacodylate buffer, pH 7.4, for 10 min, postfixed for 30 min with 1% OsOl in 0.1 M cacodylate buffer, pH 7.4, at 4”C, dehydrated in graded ethanol, and embedded in Epon-araldite or Polybed (Ladd).
Biochemical
Analysis
of Cytochemical
Media
Since fixatives inhibit phosphatase activity (Ernst, 1975), parallel experiments were done with fetal colonic slices placed in 125 mMTris-HCl buffer, pH 8.5, without prior fixation. The phosphatase activity of the tissue slices was quantitated by performing the cytochemical incubations at pH 8.0 and 9.0 as described above and then measuring spectrophotometrically the amount of cleaved substrate (PNP) released into the medium. Tissue slices from two to four fetuses were pooled in a single assay tube and were preincubat.ed in 1.4 ml of the appropriate incubation medium (as above) without pNPP for 30-60 min. Substrate was then added to each tube to a final concentration of 20 mM and the tubes were placed in a shaking water bath at 37°C for 40 min. Blanks for each assay tube contained medium plus pNPP, but no tissue. After incubation, the tubes were spun in an IEC clinical centrifuge for 1 min to pellet the tissue and the concentration of PNP in the supernatant was determined by spectrophotometric absorbance at 420 nm. The tissue pellet was resuspended in 0.9% NaCl, sonicated, and the protein concentration was determined according to Lowry (1951) using BSA as a standard. Enzyme activity was expressed as millimoles of PNP liberated per milligram of protein per 40 min.
Freeze Fracture Ascending and descending portions of fetal colon were separated and fixed for 30 min at 23°C in a solution of 2% formaldehyde, 2.5% glutaraldehyde, and 0.4% CaClz
COLONY
AND
NEUTRA
Phosphatases in Fetal Rat Colon
in 0.1 M Na-cacodylate buffer, pH 7.4. Tissue was then washed in 0.1 M cacodylate buffer, equilibrated with 20% glycerol in the same buffer for 3 hr, sliced at 1 mm, mounted on gold disks, briefly immersed in the liquid phase of partially frozen Freon 22, and stored in liquid nitrogen. Specimens were fractured in a Balzers freeze-etch device (Model BA 360) at a stage temperature of -115”C, replicated with platinum-carbon, cleaned with bleach, and mounted on 300-mesh grids. All luminal borders present in replicas of tissue from 17-, and 19-, and 21-day fetal colons were photographed at 20,000X with a JEOL 100 CX electron microscope. Cat&nixed
Ferritin
To define the main lumen and all of its extensions, cationized ferritin was used as an ultrastructural tracer. Ascending and descending colons were excised from 17to 22-day fetuses, fixed for 45 min at 23°C as described above, rinsed in 0.1 M cacodylate buffer, cut in 4-mm cylindrical segments, and immersed in a solution of 5.5 mg/ml cationized ferritin in 0.015 M NaCl at 4°C for 15 min with intermittent agitation. Segments were then rinsed for 20 min in four changes of cacodylate buffer and returned to the fixative solution for an additional 1 hr at 23°C. The central portion of each tissue cylinder was sliced into 0.5-mm pieces that were then postfixed in 1% OsOl in 0.1 M cacodylate buffer at 4°C for 1 hr, dehydrated, and embedded as described above. For electron microscopy, thin sections were briefly stained with lead citrate. RESULTS
Epithelial
Conversion
in the Fetal Colon
A light microscopic overview of the conversion process is shown in Fig. 1. At 16 days gestation, the colonic epithelium is uniformly stratified and consists of two to four layers of undifferentiated cells lining a narrow flattened lumen (Fig. la). By 17 days, several longitudinal cleft-like extensions of the lumen contact the basal cell layer, giving the lumen a stellate profile in cross section (Fig. lb). Within the thick layer of undifferentiated stratified cells between the clefts, small secondary lumina appear. During the 18th day, both the main lumen and the secondary lumina expand. By 19 days (Fig. lc), some of the apical cell layers form cellular “bridges” over the secondary lumina. These apical cells are often rounded and darkly stained and appear to be shed during the 19th and 20th days (Fig. lc) as ridges of connective tissue invaginate the epithelium from below. By 21 days (Fig. Id) the loss of apical cells has resulted in the joining of the primary lumen with the secondary lumina, leaving a single large
lumen lined by simple columnar epithelium longitudinal ridges of connective tissue.
351 covering
Occluding Junctions At all developmental stages examined (17 to 21 days) primary and secondary luminal surfaces were recognized in freeze-fracture replicas by the presence of typical microvilli. All primary and secondary lumina observed were bordered by continuous occluding junctions (Fig. 2). The continuity of these junctions was confirmed by the observation that cationized ferritin introduced into the main lumen did not enter secondary lumina (Fig. 3). Thus, luminal and nonluminal domains were consistently separated by occluding junctions throughout the conversion process. EM Cytochemistry No phosphatase activity, either at pH 8.0 or pH 9.0, was cytochemically demonstrable in the stratified epithelium of fetal colons at 16 or 17 days (Fig. 4a). Plasma membranes lining the main lumen and the small secondary lumina, as well as nonluminal membranes, were consistently devoid of reaction products. Both pH 8.0 and pH 9.0 phosphatase activities were detectable at 18 days gestation (Figs. 4b and c, 5). In tissues incubated at pH 9.0, phosphatase reaction product was observed along the plasma membranes contacting the main lumen (Fig. 4b) and the secondary lumina (Fig. 4~). A lighter precipitate was observed along basolateral membranes as well (Fig. 4~). After incubation at pH 8.0 in the presence of K+ and cysteine, reaction product appeared along nonluminal membranes of some epithelial cells distributed throughout the stratified epithelium but plasma membranes bordering either the main or the secondary lumina were devoid of dense deposits (Fig. 5). By 19 days gestation the cytochemical reactions were stronger than at 18 days, particularly in the epithelium of the ascending colon (Figs. 6-8). At pH 9.0, cysteinesensitive phosphatase activity was localized along apical membranes bordering the main lumen and its clefts as well as secondary lumina (Fig. 6). Local unreactive cells or groups of cells were common. Small and variable amounts of activity were also observed on basolateral membranes (Fig. 7). Alkaline phosphatase activity was consistently stronger on epithelial cells of the ascending colon, although its distribution was similar in both ascending and descending regions. The reaction product was more uniform and dense when ,&glycerophosphate was used as substrate instead of pNPP. Phosphatase activity at pH 8.0 was distributed on nonluminal membranes of cells at all levels of the epithelium at 19 (Fig. 8) and 20 days gestation. Although
352
DEVELOPMENTALBIOLOGY
V0~~~~97,1983
of fetal rat colon, demonstrating the conversion of the epithelium from stratified to simple columnar. (a) 16 FIG. 1. Light micrographs days gestation. A stratified epithelium, two to three cell layers thick, surrounds a small central lumen (L). X600. (b) 17 days. The main lumen (ML) has expanded to form deep clefts (arrows) that extend down to the basal cell layer. Within the stratified epithelium between the clefts, a small secondary lumen (SL) has formed. X600. (c) 19 days. The secondary lumena (SL) are enlarged. The darkly stained, rounded apical cells (arrows) will be sloughed into the main lumen (ML). The basal side of the epithelium is invaginated by ridges of mesenchyme (*). X600. (d) 22 days. A simple columnar epithelium lines longitudinal ridges of connective tissue. Differentiated goblet cells (GC) are interspersed among absorptive cells. x500.
COLONY
AND NEUTRA
Phosphtases
FIG. 2. Freeze-fracture replicas of occluding junctions on cells bordering (c) 21 days. The organization of occluding junctions is variable and changes X48.000.
luminal membranes generally showed no activity (Fig. 8, inset), reaction product was occasionally observed on the apical membrane of cells contacting a lumen, even in samples incubated at pH 8.0 in the presence of KC1 and cysteine. Cells lacking a luminal surface had reaction product distributed over the entire cell surface. Most deposits were located on the cytoplasmic side of the plasma membrane, although precipitate was also observed on the extracellular side. Addition of cysteine reduced but did not abolish the extracellular reaction product. When ouabain was added to the incubation
in
Fetal
the main lumen with increasing
Rat
Colon
353
of fetal rat colon, at (a) 17 days, (b) 19 days, and age, but they appear to be continuous at all ages.
medium, or when K+ was omitted, the extracellular deposits persisted and the deposits on the cytoplasmic side of the membrane were only partially reduced. In the descending colon, no consistent basolateral membrane phosphatase activity at pH 8.0 was detected at any gestational age. At 21 and 22 days gestation, phosphatase activity at pH 9.0 was localized primarily on the luminal membranes of the simple columnar cells (Fig. 9). At this age phosphatase activity at pH 8.0 in the ascending colon was visualized as distinct crystalline deposits located
354
DEVELOPMENTAL BIOLOGY
VOLUME 97, 1983
chemically detectable levels. A significant portion of the alkaline phosphatase activity detectable at 18 and 19 days was cysteine insensitive (Figs. lla and b). Intestinal-type alkaline phosphatase activity, which is cysteine sensitive, was low at these early ages but increased to a maximum at 20 days in descending colon (Fig. llb) and at 21 days in ascending colon (Fig. lla). The apparent decline in activity just prior to birth (22 days) may reflect the rapid growth in mass of nonepithelial elements of the intestinal wall (such as muscle and connective tissue) at this age. Within 2 hr after birth, cysteine-sensitive alkaline phosphatase activity increased dramatically in the proximal colon (Fig. lla), whereas levels in the distal colon remained low (Fig. llb). Phosphatase activity at pH 8.0 was not significantly different in ascending and descending colon so the data for the two regions were combined (Fig. 12). Activity appeared at 18 days gestation, increased to an apparent maximum at 20 days, and then declined to a level that remained relatively constant through the day of birth. As with alkaline phosphatase, the apparent decline in activity at 21 days coincided with a rapid growth of nonepithelial elements of the intestinal wall, and thus with an increase in total protein. No significant inhibition of pH 8.0 activity was observed when K+ was omitted from the medium (and substituted by choline) or when 10 mM ouabain was added. DISCUSSION
FIG. 3. Electron micrograph of a layer of apical epithelial cells in the colon of a 19-day fetal rat. Cationized ferritin was infused into the main lumen. This cell layer (seen as dark cells in Fig. lc) separates the main lumen from a secondary lumnen. Cationized ferritin (CF) is present on the membranes exposed to the main lumen, but the tracer has not entered the secondary lumen. X44,709.
primarily on the cytoplasmic face of the interdigitated lateral membranes of simple columnar absorptive cells (Fig. 10). Activity on luminal membranes was rarely observed. Addition of ouabain or deletion of K+ ions again reduced but did not eliminate the deposits on the cytoplasmic side of the basolateral membranes. Biochemical Analysis of Cytochemical Media
The onset of phosphatase activity at both pH 8.0 and pH 9.0 was detected biochemically at 18 days gestation, the same age at which enzyme activity reached cyto-
Alkaline phosphatase and K+-dependent ATPase are considered marker enzymes for luminal and contraluminal plasma membrane domains in a variety of transporting epithelia (DiBona and Mills, 1979; Ernst and Mills, 1980; Ernst and Hootman, 1981) including the small intestine (Stirling, 1972; Hopfer et a,!., 1975; Ernst, 1975). The colonic epithelium also transports ions and water, but the histochemical localization of phosphatases in both neonatal and adult colon has been controversial (Deane and Dempsey, 1945; Masnerova et al, 1966; Helander, 1975; Ono, 1976). The contradictory data may be ascribed to the low levels of colonic phosphatase activities, relative to those found in the kidney or small intestine (Vengesa and Hopfer, 1979; Lev and Griffiths, 1982) as well as to the variable sensitivity and specificity of cytochemical methods (Ernst and Hootman, 1981). More recently, both phosphatases have been detected on the plasma membranes of epithelial cells of adult rat and rabbit colon by sensitive ultrastructural cytochemistry and cell fractionation (Vengesa and Hopfer, 1978, 1979). These studies demonstrated mutually exclusive localization of the two enzymes to the two major membrane domains. We have now demonstrated that plasma membrane
FIG. 4. Fetal rat ascending colon, reacted cytochemically to demonstrate alkaline phosphatase activity (pH 9.0) at 1’7 and 18 days gestation. (a) 17 days. This field includes a portion of the main lumen (ML) and a small secondary lumen (SL) in the stratified epithelium. No alkaline phosphatase activity is detectable. Some nonspecific precipitate is present in mitochondria. X14,000. (b) 18 days. Reaction product is localized along the apical plasma membranes of cells bordering the main lumen (ML) Substrate: pNPP. X16500. (c) 18 days. A secondary lumen (SL) deep in the stratified epithelium shows alkaline phosphatase activity on luminal plasma membranes. A smaller amount of reaction product is also seen on nonluminal membranes (arrow). Substrate: @GP. X15,900. 355
356
DEVELOPMENTAL BIOLOGY
VOLUME 9’7, 1983
FIG. 5. Colonic epithelium of an l&day fetus, reacted for phosphatase activity at pH 8.0 in the presence of K+ and cysteine. Reaction product (arrows) is seen along the plasma membranes of some cells deep in the stratified epithelium. No activity is present on membranes bordering the secondary lumen (SL). X10,400.
phosphatases are present in the fetal rat colon, appearing at 18 days gestation while the epithelium is still stratified. These membrane phosphatases remain active during the period of rapid epithelial reorganization (between 18 and 21 days gestation). One day before birth, a simple columnar epithelium is established and the cellular localization of both phosphatases is similar to that seen in the adult colonic epithelium. Our cytochemical data do not rule out the possibility that phosphatases may be present prior to 18 days in an inactive form. Immunohistochemical studies of small intestine have detected immunoreactive alkaline phosphatase in neonatal epithelium and in crypt cells that is not enzymatically active (Etzler and Moog, 1966; Wachsmuth and Torhorst, 1974; Shields et aL, 1982). In the adult small intestinal epithelium, alkaline phosphatase activity is not restricted to the luminal plasmalemma. Membrane fractionation and histochemical studies have consistently revealed low levels of alkaline phosphatase activity on basolateral membranes (Hugon and Borgers, 1968; Ono, 1974, 19’75; Hanna et aL, 1979). Our EM cytochemical data indicate that the distribution of colonic alkaline phosphatase in the fetal rat resembles that of the adult small intestine and of
postnatal rat colon (Helander, 1975; Ono, 1976), as it is located primarily on the apical membrane but is also present to a lesser extent on basolateral membranes. The cytochemical reaction products localized along both apical and basolateral membranes of fetal colonic epithelium after incubation at pH 9.0 reflect, at least in part, intestinal-type alkaline phosphatase activity according to criteria established in previous histochemical and biochemical studies: the precipitate is localized on the outer membrane leaflet (Ono, 1973; Ernst, 1975), is inhibited by cysteine (Van Belle, 1976; Vengesa and Hopfer, 1979), and shows the same distribution with either p-nitrophenylphosphate (pNPP) or P-glycerophosphate (P-GP) as substrate. The effectiveness of 8GP as substrate further confirms the presence of alkaline phosphatase, since other phosphatases such as 5’-nucleotidase, Ca-dependent ATPase and Na/K ATPase do not cleave this substrate. Moog and Yeh (1973) reported that in the suckling rat small intestine and in the crypt of adult intestine P-GP was cleaved more effectively than pNPP, whereas pNPP was the preferred substrate in cells at the adult villus tip. In the fetal colon, we observed a stronger and more consistent cytochemical reaction in tissue incubated at pH
COLONY
AND NEUTRA
Phosphutmes
in
Fetal
Rat
Colon
357
FIG. 6. 19-day fetal ascending colon, alkaline phosphatase activity. The apical plasma membranes of most of the cells that contact the main lumen show strong phosphatase activity at pH 9.0. Although cells lining the luminal clefts in this field are not reactive, many cleft cells did show phosphatase activity (see Fig. ‘7). Occasional cells (see lower center) show strong nonluminal enzyme activity; whether this is a sign of impending cell death is not known. Substrate: @GP. X24,000.
9.0 with P-GP as substrate. These data suggest that an isoenzyme that preferentially cleaves /3-GP may be characteristic of immature cells. The biochemical data confirm the onset of alkaline phosphatase activity at 18 days gestation in fetal colon, Total activity increases to a peak at 20 days in the descending and 21 days in the ascending colon, coincident with the completion of the conversion of the stratified
epithelium to simple columnar. Just after birth, activity in ascending colon increases dramatically, presumably reflecting the morphological differentiation of an alkaline phosphate-positive apical tubulovesicular system in the surface columnar cells (Masnerova et aL, 1966; Ono, 1976). This system does not develop in absorptive cells of the descending colon (Ono, 1976, 1977) and thus, no rise in enzyme activity is noted after birth
358
DEVELOPMENTAL
BIOLOGY
FIG. ‘7. 19-day fetal ascending colon, alkaline phosphatase activity. Luminal plasma membranes of cells lining a cleft of the main lumen are strongly reactive, but weaker phosphatase activity is also seen on lateral membranes below the occluding junctions (OJ). In both locations, dense reaction product is restricted to the extracellular side of the membrane. Substrate: p-GP. X27,000.
in this region. Only a portion of the total activity at any age is cysteine sensitive and thus represents intestinal epithelial cell activity. The cysteine-insensitive activity measured biochemically is presumably due to the cytochemically demonstrable enzyme activity on capillary endothelia in the lamina propria (Lev and Griffiths, 1982). Phosphatase activity at pH 8.0 was detected cytochemically in fetal rat colon under the conditions known to demonstrate the K+-dependent phosphatase com-
VOLUME
97,1983
ponent of the Na+/K+ ATPase enzyme complex in other tissues (Ernst, 1972a,b; 19’75; Ernst and Hootman, 1981). As in other tissues, the reaction product in fetal rat colon was concentrated along nonluminal membranes, was located at least in part on the inner leaflet of the plasma membrane, and was not abolished by cysteine. The positive identification of basolateral Na+/K’ ATPase in fetal colon is seriously complicated, however, by the finding that enzyme activity persisted as determined biochemically and as visualized cytochemically, in the absence of K+ ions or in the presence of ouabain. In previous studies, omission of K+ or addition of cardiac glycosides eliminated detectable activity in most adult tissues (Ernst and Philpott, 1970; Ernst, 1972a,b; Vengesa and Hopfer, 1979). In contrast, K+-dependent pNPPase in rat tissues was found to be relatively insensitive to ouabain (Allen and Schwartz, 1969; Ernst, 1975). Ernst further showed that loss of sensitivity of the ATPase complex to ouabain occurs when the native substrate ATP is substituted by pNPP, and also in the presence of 20 mM strontium. Thus a combination of these effects may have reduced the ouabain sensitivity of the fetal rat enzyme in the present study to undetectable levels. Nevertheless, pNPPase activity at pH 8.0 was completely abolished by the cardiac glycoside scillaren and by omission of K+ ions in a previous study of adult rat and rabbit colon (Vengesa and Hopfer, 1979). The insensitivity of pH 8.0 pNPPase to ouabain and to omission of K+ in the fetal colon could also be due in part to an intrinsically low level of activity that has been further reduced by fixation and by the presence of strontium (Ernst, 1972a,b). In addition, the sensitivity of the cytochemical technique may be reduced by diffusion of reaction product, inefficient trapping of the released phosphate, and solubilization of the SrP04 precipitate at pH 8.0 (Ernst, 1972a,b). Finally, although controls were preincubated without substrate in K+free media, sufficient K+ ions may have been retained in the tissue slices to maintain low levels of pNPPase activity. Thus, the exact identity of the fetal basolateral enzyme activity is not established, but at least part of the cytochemical reaction product on the inner membrane leaflet was decreased under inhibitory conditions and may reflect Na+/K+ ATPase activity. The secondary lumina which form within the stratified epithelium of the fetal colon separate entire layers of superficial cells from the cells below. These cells form barriers between primary and secondary lumina and may have two luminal surfaces, both sealed by tight junctions. One surface contacts the main lumen and is labeled by a luminal infusion of cationized ferritin (Fig. 3), whereas the other contacts a closed intraepithelial space. Since both of these surfaces acquire alkaline phosphatase activity simultaneously, the induction of
F IG. 8. 19-day fetal ascending colon; pH 8.0 phosphatase activity in the presence of K+ and cysteine. Enzyme the entire plasma membrane of basal cells that do not contact a lumen. Reaction product is located principally the membrane. Inset: Cells that border a luminal cleft have reaction product only on nonluminal membranes. prel cipitate is on the extracellular side of the membrane. X14,300.
359
activity is distributed olver on the cytoplasmic side ? of In this field, much of the
360
DEVELOPMENTAL
BIOLOGY
VOLUME 97, 1983
that do not contact a luminal surface at all. The entire plasma membrane of these cells exhibits a pattern of phosphatase activities similar to basolateral membrane of adult intestinal cells; i.e., low levels of cysteine-sensitive pH 9.0 phosphatase on the outer leaflet and pH 8.0 pNPPase on the inner leaflet. The stratified fetal colonic epithelium transiently resembles the stratified transporting epithelium of frog skin (Mills et aL, 1977; DiBona and Mills, 1979) in that pH 8.0 phosphatase activity is distributed over the entire plasma membrane of basal cells. It is generally accepted that the continuity of occluding junctions is crucial for maintaining the separate identity of luminal and contraluminal domains of et aL, 1978; epithelial plasma membranes (Meldolesi studies of fetal Dragsten et al, 1981). Freeze-fracture small intestinal epithelium (Madara et aL, 1981; Teillet
FIG. 9. 22-day fetal ascending colon, alkaline phosphatase activity. Enzyme activity on simple columnar epithelial cells appears principally on the luminal microvillus membrane (mv) hut is also present on basolateral membranes (arrow). Substrate: @-GP. X7700.
enzyme activity apparently does not depend on exposure of the cell surface to the milieu of the main lumen. Deeper in the stratified epithelium, there are cells
FIG. 10. 22-day fetal ascending colon, pH 8.0 phosphatase activity in the presence of K’ ions and cysteine. The dense crystalline precipitate produced by enzyme activity is present in the interdigitated lateral membranes of two adjacent columnar cells. The reaction product is located on the cytoplasmic side of the membranes. X29,100.
COLONY AND NEUTRA
Phosphatases
Qa Fetal
Rat
361
Colon
lla FETAL RAT ASCENDING ON PHOSPHATASE ACTIVITY, pH 9.0
0.6
16
19
21
22
Fetal Age (In days)
L 4 nc born Mrth
(2 hrs)
llb FETAL RAT DESCENDING COLON PHOSPHATASE ACTIVITY, pH 9.0
0.6 0.5 L E s c 3 g
0
Total Actlvlty
a Cyatolne
Added
0.4 0.3
,P B 0.2 a =
0.1
I8
I9
!O
21
22
FIG. 11. Alkaline phosphatase activity of fetal colonic slices, measured spectrophotometrically as amount of cleaved substrate (PNP) released into the medium per mg tissue protein, during a 40 min incubation period. Activity is present at 18 days gestation and increases to a peak at 21 days in ascending colon (a) and 20 days in descending colon (b). The apparent decline prior to birth may be due to rapid growth of nonepithelial tissues. The cysteine-sensitive portion of the total activity represents specific intestinal epithelial alkaline phosphatase; the cysteine-insensitive portion is presumably due to alkaline phosphatase on other cell types.
et aL, 1981) and of fetal colonic epithelium in this study have shown that all lumina, whether main or secondary, are bordered by continuous occluding junctions. Although junctions on fetal cells appear more loosely organized than those seen in adult intestinal epithelium,
they display no visible discontinuities. Presumably, this prevents the lateral diffusion of fetal membrane proteins in the plane of the membrane between luminal and contraluminal domains. It is interesting that in the fetal colon these two domains are separated by tight
362
DEVELOPMENTAL
BIOLOGY
VOLUME
97, 1983
12 FETAL
0.6 o 5 .G 0: o 4. . * c -Z
RAT
COIBN (total)
PHOSPHATASE 0
Total Activity
q q
K* Omitted Ouobain
I6
ACTIVITY,
Added
19
21
Fetal Age (in days) FIG. 12. Phosphatase activity at pH 8.0 in fetal reaches a peak at 20 days. The apparent decline phosphatase activity was not reduced by omission
pH .6.0
22
b newborn (2 hrr)
birth
colon slices, measured and expressed as in Fig. 12. Enzyme activity after 20 days is presumably due to rapid growth of nonepithelial at Kt ions from the medium, or by addition of 10 mM ouabain.
junctions prior to the appearance of membrane phosphatase activities, and that when these phosphatases first appear, they are already segregated. The excellent technical assistance of Ms. Teresa Hall is gratefully acknowledged. This work was supported in part by Grant AM 21505 from the National Institutes of Health. Dr. Colony was recipient of a Senior Research Fellowship from the National Foundation for Ileitis and Colitis. Dr. Neutra is recipient of Research Career Development Award AM 00407 from the National Institutes of Health. The JEOL 100 CX electron microscope was funded by National Science Foundation Grant PCM 7816158.
REFERENCES ALLEN, J. C., and SCHWARTZ, A. (1969). A possible biochemical explanation for the insensitivity of the rat to cardiac glycosides. J. Pharmacol. Exp. Ther. 168, 42-46. CALVERT, R., MALKA, D., and MENARD, D. (1981). Establishment of regional differences in brush border enzymatic activities during the development of fetal mouse small intestine. Cell Tissue Res. 214, 97-106. DAHLQUIST, A., and LINDBERG, T. (1966). Fetal development of the small intestinal disaccharidase and alkaline phosphatase activities in the human. Biol. Neonat. 9,24-32. DEANE, H. W., and DEMPSEY, E. W. (1945). The localization of phosphatase in the Golgi region of intestinal and other epithelial cells. Anat. Rec. 93, 401-417. DIBONA, D. R., and MILLS, J. W. (1979). Distribution of Na+-pump sites in transporting epithelia. Fed Proc. 38, 134-143. DRAGSTEN, P. R., BLUMENTHAL, R., and HANDLER, J. S. (1981). Membrane assymetry in epithelia: Is the tight junction a barrier to diffusion in the plasma membrane? Nature (London) 294, 718-722. ERNST, S. A. (1972a). Transport adenosine triphosphatase cytochemistry I. Biochemical characterization of a cytochemical medium for the ultrastructure localization of ouabain-sensitive, potassium-de-
is low at 18 days and tissues. At all ages,
pendent phosphatase activity in the avian salt gland. J. His&hem. Cytochem. 20, 13-22. ERNST, S. A. (1972b). Transport adenosine triphosphatase cytochemistry. II. Cytochemical localization of ouabain-sensitive, potassiumdependent phosphatase activity in the secretory epthelium of the avian salt gland. .J. Histochem. Cytochem. 20,23-38. ERNST, S. A. (1975). Transport ATPase cytochemistry: Ultrastructural localization of potassium-dependent and potassium-independent phosphatase activities in rat kidney cortex. J. Cell Biol. 66,586-608. ERNST, S. A., and HOOTMAN, S. R. (1981). Microscopical methods for the localization of Na+/K’-ATPase. Histochem. J. 13, 397-418. ERNST, S. A., and MILLS, J. W. (1980). Autoradiographic localization of tritiated ouabain sensitive sodium pump sites in ion transporting epithelia. J. Histochem. Cytochem. 28, 72-77. ERNST, S. A., and PHILPOTT, C. W. (1970). Preservation of Na-K-activated and Mg-activated adenosine triphosphate activities of avian salt gland and teleost gill with formaldehyde as fixative. J. Histothem. Cytochem. 18,251-263. ETZLER, M. E., and MOOG, F. (1966). Inactive alkaline phosphatase in duodenum of nursling mouse: Immunological evidence. Science 154, 1037-1038. FUJITA, M., OHTA. H., KAWAI, K., MATSUI, H., and NAKAO, M. (1972). Differential isolation of microvillous and basolateral plasma membranes from intestinal mucosa: Mutually exclusive distribution of digestive enzymes and ouabain-sensitive ATPase. B&him. Biophys. Acta 274, 336-347. HANNA, S. D., MIRCHEFF, A. R., and WRIGHT, E. M. (1979). Alkaline phosphatase of basal-lateral and brush border plasma membranes from rat intestinal epithelium. J. Suprumol Struct. 11,451-466. HELANDER, H. F. (1973). Morphological studies on the development of the rat colonic mucosa. Acta Anat. 85, 153-176. HELANDER, H. F. (1975). Enzyme patterns and protein absorption in rat colon during development. Aeta Anat. 91,330-349. HOPFER, U., SIGRIST-NELSON, K., and MURER, H. (1975). Intestinal sugar transport: Studies with isolated plasma membranes. Ann. N. Y. Acad. SC%. 264,414-426. HUGON, J., and BORGERS, M. (1968). Fine structural localization of
COLONY AND NEUTRA
Phosphatases
acid and alkaline phosphatase activities in the absorbing cells of the duodenum of rodents. Histochemie 12,42-66. HUGON, J. S., and BORGERS,M. (1969). Ultrastructural differentiation and enzymatic localization of phosphatase in the developing duodenal epithelium of the mouse. I. The foetal mouse. Histochemie 19,13-30. JIRASEK, J. E., UHER, J., and KOLDOVSKY, 0. (1975). A histochemical analysis of the development of the small intestine of human fetuses. Actu
Hi&o&em
22,33-39.
KELLEY, R. (1973). An ultrastructural and cytochemical study of developing small intestine in man. J. Embryol Exp. Morph.& 29.411430.
LEV, R., and GRIFFITH& W. C. (1982). Colonic and small intestinal alkaline phosphatase: A histochemical and biochemical study. Gastroenter~ 82,1427-1435. LEV, R., SEIGEL, H. I., and BARTMAN, J. (1972). Hi&chemical studies of developing human fetal small intestine. Histdwmie 29.103-119. LOWRY, 0. H., ROSEBROUGH,N. J., FARR, A. L., and RANDALL, R. J. (1951). Protein measurement with the folin-phenol reagent. J. CeU Bid Chem. 193, 265-275. MADARA, J. L., NEUTRA, M. R., and TRIER, J. S. (1981). Junctional complexes in fetal rat small intestine during morphogenesis. Dew. Bid 86, 170-178. MASNEROVA, M., KOLDOVSKY, O., and KUBAT, K. (1966). Postnatal changes in phosphatase and non-specific esterase activity in the large intestine of the rat. Ewentia (Basel) 22,518-519. MATHAN, M., MOXEY, P. C., and TRIER, J. S. (1976). Morphogenesis of fetal rat duodenal villi. Amer. J. And. 146, 73-92. MELDOLESI, J., CASTIGLIONI, G., PARMA, R., NASSIVERA, N., and CAMILLI, P. (1978). Ca++-dependent disassembly and reassembly of occluding junctions in guinea pig pancreatic acinar cells. J. CeU Bid 79, 156-172. MILLS, J. W., ERNST, S. A., and DIBONA, D. R. (1977). Localization of Na+ pump sites in frog skin. J. CeU Bid 73.88-110. MIRCHEFF, A. K., and WRIGHT, E. M. (1976). Analytical isolation of plasma membranes of intestinal epithelial cells: Identification of Na, K-ATPase rich membrane and the distribution of enzyme activities. J. Menrbr. Bid 28, 309-333. MOOG, F., and YEH, KWO-YIH. (1973). Intestinal alkaline phosphatase of the rat: Development and distribution of activity with phenyl-
in
Fetal
Rat
Colon
363
phosphate and fl-glycerophosphate. Camp. Biochem. Physiol. 44B: 6457-666. ONO,K. (1973). The fine structural localization of alkaline phosphatase activity of intestinal microvilli in the developing chick embryo. Acta Anat
86,71-82.
ONO, K. (1974). The localization of alkaline phosphatase activity in the rat small intestinal epithelium after fat and protein feeding. Acta Hi&&m. 51,124-137. ONO, K. (1975). The fine structure and localization of alkaline phosphatase activity of the small intestinal epithelium in the postnatal developing rat. Acta His&hem+ S2,117-133. ONO, K. (1976). Alkaline phosphatase activity of large intestinal principal cells in postnatal developing rats. Acta His&hem. 57, 312319. ONO, K. (1977). Absorption of horseradish peroxidase by the principal cells of the large intestines of postnatal developing rats. Anat. EmiqoL 151, 53-62. SHERVEY, P. D. (1973). A histochemical study of the inclusion bodies in the small intestine of the developing rat. Ziet Anat. Entticklung. 141,39-54. SHIELDS, H. M., BAIR, F. A., BATES, M. L., YEDLIN, S. T., and ALPERS, D. H. (1982). Localization of immunoreactive alkaline phosphatase in the rat small intestine at the light microscopic level by immunocytochemistry. Gastroenterology 82,32-45. STIRLING, C. E. (1972). Radioautographic localization of sodium pump sites in rabbit intestine. J. Cell Biol. 53, 704-714. TEILLET, M. A., HUGON, J. S., and CALVERT, R. (1981). The occluding junctions of mouse duodenal enterocytes during development. A freeze fracture study. Cell Tissue Res. 217, 65-78. VAN BELLE, H. (1976). Kinetics and inhibition of rat and avian alkaline phosphatase. Ga PharmacoL 7, 53-58. VENGESA, P. B., and HOPFER, U. (1978). A localization of alkaline phosphatase and Na+-pump sites in adult rat and rabbit colon by cytochemical and cell fractionation techniques. J. Cell Biol. 79,238a. VENGESA, P. B., and HOPFER, U. (1979). Cytochemical localization of alkaline phosphatase and Na+ pump sites in adult rat colon. J. Hi&o&em Cytochem. 27,1231-1235. WACHSMUTH, E. D., and TORHORST, A. (1974) Possible precursors of aminopeptidase and alkaline phosphatase in the proximal tubules of kidney and the crypts of small intestine of mice. Histochemistly 38,43-56.