THE ANATOMICAL RECORD 229:219-226 (1991)
Rat Small Intestinal Mucins: A Quantitative Analysis ABBIE C. KEMPER AND ROBERT D. SPECIAN Department of Cellular Biology and Anatomy, Louisiana State University Medical Center, Shreveport, Louisiana
ABSTRACT Although much is known about the qualitative distribution of mucin-secreting goblet cells in the small intestine, the quantitative distribution of stored mucins remains undefined. The purpose of this study was to determine the distribution of neutral stored mucin in the rat small intestine by using morphometric techniques and once established, to verify that this methodology could detect secretion in animals exposed to a known mucin secretagogue. Twelve male Wistar rats (five baseline, five pilocarpine-treated, and two vehicle controls) were fixed by vascular perfusion. After a brief fixation the intestine was removed, cut into 10 equal segments, sliced, and fixed overnight. Methacrylate sections from each segment were stained with periodic acid-Schiff and toluidine blue. For morphometry, the volume of epithelium per surface area of epithelial basal lamina was calculated with a Merz grid. The volume density of stored mucin per epithelium was determined by point-counting on a square lattice grid. Volumes were related to either surface area of epithelial basal lamina or mucosal surface area. Due mostly to contributions by villus stored mucin, the total amount of product was found to increase proximally to distally in the small bowel, with the most dramatic increases occurring in the first three segments. When subjected to pilocarpine, a massive secretory response was evoked, resulting in a near total depletion of crypt stored mucin a t all levels of the small bowel. Secretion of villus stored mucin also occurred throughout the small intestine, however reaching levels of significance a t only a few points. This study describes the distribution of stored mucin in the small intestine under baseline and accelerated secretory conditions. Intestinal mucins are large, negatively-charged glycoproteins that are synthesized, stored, and secreted by goblet cells. Once secreted into the intestinal lumen, mucins form a gel that serves a s a protective barrier for the underlying epithelium. The function of this gel is multifaceted. It protects the epithelium from physical damage by luminal contents, guards against bacterial invasion (Neutra and Forstner, 19871, regulates epithelial hydration (Lukie, 1977), and interacts with secretory immunoglobulin A to produce antibody and antitoxin effects (Klipstein et al., 1984). Degradation of the mucus gel probably begins in the small intestine by lysozymal enzymes and continues in the large intestine by bacterial neuraminadases. In addition, HC1, pepsin, and pancreatic enzymes play a role in the proximal small bowel (Horowitz, 1967; Allen and Starkey, 1974). Due to luminal proteases, the mucus gel must be constantly replaced to adequately protect the epithelium. This renewal is accomplished by the baseline secretion of mucins. Baseline secretion in goblet cells occurs by an intermittent exocytosis of single mucin granules located a t the upper lateral edge of the theca. Under these conditions it has been observed that peripheral granules migrate along the lateral edges of the theca due to the influence of microtubules (Specian and Neutra, 1984). The exact mechanisms that control normal baseline secretion of mucin are not known; GI 1991 WILEY-LISS, INC
however, the effects of cholinergic stimulation on secretion in the small intestine are fairly well established. In organ culture and in vivo, acetylcholine and other cholinergic analogs, such as pilocarpine and carbachol, elicit a compound exocytosis of mucin (Specian and Neutra, 1980). This cholinergic-mediated accelerated release of mucin is limited to the crypts (Specian and Neutra, 1982) but comparable cellular mechanisms allow the rapid release of mucins from surface or villus cells in response to chemical or physical irritation (Neutra et al., 1982). Much is known about the biochemical nature of mucins and the qualitative distribution of goblet cells in the small intestine. Until now, the quantitative distribution of stored mucin has remained undefined. The purpose of this project was to determine the distribution of neutral stored mucin in the small intestine of the rat, using the methodology of Heidsiek et al. (1987). Once the baseline values were determined, a known mucin secretagogue was employed to verify that this methodology could detect secretion.
Received November 27, 1989; accepted J u n e 27, 1990. Address reprint requests to Dr. Robert D. Specian, Department of Cellular Biology and Anatomy, LSU Medical Center, 1501 Kings Highway, Shreveport, LA 71130.
220
A.C. KEMPEK AND R.D. SPECIAN
MATERIALS AND METHODS Baseline: animals and tissue processing Five male Wistar rats weighing between 450 and 550 g each were fasted 24 hours prior to vascular perfusion fixation. Rats were anesthetized with sodium pentobarbital and fixed with 2% paraformaldehyde, 2% glutaraldehyde, and 2.5% polyvinylpyrolidine in 0.1 M phosphate buffer (Forssmann et al., 1976). Following sacrifice, the small intestine was removed, cut into 10 equal lengths, and photographed (Fig. 1).Tissue was then sliced, processed, and embedded in glycol methacrylate (JB-4; Polysciences, Inc., Warrington, PA). Three micrometer cross sections from each segment were stained with periodic acid-Schiff, to demonstrate neutral mucins and toluidine blue (Humason, 1979). A section from each segment was then photographed on 35 mm film a t 10 randomly selected points using the 16 x objective of a Leitz Orthoplan photomicroscope (R. Leitz Inc., Rockleigh, NJ). Pilocarpine: animals and tissue processing Seven male Wistar rats weighing 450 to 550 g each were fasted 24 hours prior to vascular perfusion fixation. Thirty minutes prior to sacrifice, five animals received a single intraperitoneal injection of pilocarpine suspended in phosphate buffered saline (PBS) at 160 mglkg of animal weight (Specian and Neutra, 1982). Two animals, serving a s vehicle controls, received single intraperitoneal injections of PBS. All remaining methods were identical to those conducted under baseline methods.
Morphornetry Line intercept method The surface area of basal lamina per unit volume of epithelium was calculated with a Merz grid (Weibel, 1979). Counts were performed separately on 8” x 1 0 photographic enlargements for both basal lamina intersections and epithelial points. By use of the following equation:
s v = 2P1, where P1 is the number of intersections with epithelial basal lamina per test line length, the surface area of basal lamina per unit volume of epithelium was determined.
Fig. 1. Schematic representation of the r a t small intestine, depicting the 10 segments used for morphometric analysis. Coarse cross hatching represents the duodenum, unmarked area the jejunum, and fine cross hatching, the ileum
40 -
T
b
30
x
20
i l
7
N
-. E E E
10
E
n 1
Point counting The volume density of stored mucin per epithelium was determined by point counts using a square lattice grid (Weibel, 1979), and the following equation: Vv(muciniepithe1ium) = Pn(mucin)/Pt(epithelium),
where Pn equals the number of points on stored mucin and Pt the number of points on epithelium. To determine the volume of stored mucin per surface area of basal lamina, volume densities were divided by appropriate basal lamina surface areas per volume of epithelium using the following equaton: VvISv = vs
All calculations were performed separately for crypt and villus areas with the demarcation between these
T
T
t
Y
=
2
3
4
5 6 SEGWzNr
7
8
9
10
Em Crypt
Vlllus
Fig. 2. The total amount of stored mucin per mucosal surface area for baseline animals. All volumes are expressed in cubic millimeters per square millimeter of mucosal surface. n = 5.
two sectors being a line drawn under the surface epithelial basal lamina and passing through the mouths of the crypts. The volume of stored mucin could then be expressed in terms of either crypt or villus basal lamina. By using this line of demarcation a s a representation of the mucosal surface, without amplification of either crypt or villus epithelium, a third set of calculations was performed. Through the use of the line in-
A QUANTITATIVE ANALYSIS OF RAT MUCINS DUODENUM
22 1
ILEUM
Fig. 3. The percent contribution oftotal stored mucin per mucosal surface area for duodenum and ileum in baseline animals.
Fig. 4. a: Light micrograph of duodenum from baseline rats. x 265. b Light micrograph of ileum from baseline rats. x 265.
tercept method, the mucosal surface area per unit volume of epithelium was calculated. By dividing this into either crypt of villus volume density, the amount of crypt or villus stored mucin per mucosal surface area was deteremined. RESULTS Baseline animals
The total amount of stored mucin per mucosal surface area of the small intestine was found to increase
linearly from segments one to three; thereafter, the volume of stored mucin remained relatively constant throughout the remainder of the small bowel (Fig. 2). From these data, it was observed that villus stored mucin accounts for the majority of total stored glycoprotein throughout the length of the small intestine. To determine the relative contribution of both crypt and villus goblet cells to the entire mucosa, the percentage stored in each region was determined for each segment and is presented graphically for segments 1 and 10
222
A.C. KEMPEK A N D K.D. S P E C I A N
(Fig. 3). In duodenum (segment 1, Fig. 4a), crypt stored mucin constituted 46.0% and villus stored mucin 54.0% of the total amount of product present (Fig. 3). As the villus surface area increases upon leaving the duodenum, the relative contribution of the villus increases. In terminal ileum (segment 10, Fig. 4b), crypt stored mucin represents only 33.4% of the total, whereas villus stored mucin represents 66.6% (Fig. 3). The volume of stored mucin per crypt basal lamina revealed a pattern similar to that for crypt stored mucin per mucosal surface area, increasing from 2.30 x mm3/mm2in segment 1 to 3.53 x lop3 mm3/mm2in segment 3 and thereafter remaining relatively constant (Fig. 5). The volume of villus stored mucin per basal lamina follows a comparable pattern, increasing from 4.22 x lo-" mm3/mm2 in segment 1 to 13.71 x lop3 mm3/mm2 in segment 3 and leveling off (Fig. 6). In contrast to the values for total stored mucin per mucosal surface area which were additive for crypt and villus volumes, the volume of stored mucin per total basal lamina is weighted by the relative contribution of surface area of both regions. These values in segments 1, 5, and 10 were 3.07 x mm3/mm2, 8.91 x mm3/mm2, mm3/mm2, respectively (Fig. 7). and 7.83 x Vehicle control animals
Statistical evaluations between baseline (n = 5 ) and vehicle control animals (n = 2) were performed through the use of a n unpaired T-Test, where P 5 0.05 was considered significant. Of 60 tests, only segment 9 of villus stored mucin per villus basal lamina demonstrated a significant difference. No differences were found between the two groups for total stored mucin per mucosal surface area (Fig. 8). Pilocarpine- treated animals
Pilocarpine, as expected, results in the near total depletion of the stored intracellular mucin in crypt goblet cells throughout the small intestine. When quantitatively evaluated, volumes of crypt stored mucin for both the mucosal surface area and crypt basal lamina were found to be decreased significantly in all regions of the small intestine (Figs. 9, 10). Decreases in villus stored mucin were unexpected but did occur consistently throughout the length of the small intestine, although only reaching significant levels in some segments (Fig. 11). Villus stored mucin volume per mucosal surface area was decreased significantly in segments 1-6 and in segment 9, whereas the stored mucin volume per villus basal lamina decreased significantly in segments 2 and 3 (Figs. 12, 13).Total stored mucin per mucosal surface area decreased significantly a t all levels except segment 8 (Fig. 14). Similarly, quantities of total stored mucin per total basal lamina decreased significantly in all segments except 8 and 9, due to the contribution of the crypt epithelium (Fig. 15). From these results we were able to determine the percentage of stored mucin per mucosal surface area that was secreted by both crypts and villi. In the duodenum (segment l ) , 46% (7.8 mm3/mm2) of the total quantity of mucin was found in the crypts (Fig. 3). Of this amount, over 93% (7.3 mm3/mm2)was secreted (Fig. 16). Villus stored mucin in the same segment accounted for 54% (9.1 mms'/mm2) of the total amount of product present (Fig. 3).After cholinergic stimulation, 50.6% (4.6 mm3/
CRYPT MUCIN/CRYPT BASAL LAMINA EASELlhE
20,
.
n
4
E E
0 1
2
3
4
5
6
7
8
9
10
SEGMNT
Fig. 5. The volume of crypt stored mucin per crypt basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5. VILLUS MUCIN/VILLUS BASAL LAMINA EASELIfE
2
20
W
CD
-t D
b r
10
X
.
N
E
E
E
E a 1
2
3
4
5 6 sEav€NT
7
8
9
10
Fig. 6. The volume of villus stored mucin per villus basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5. TOTAL MUCIN/TOTAL BASAL LAMINA BASELINE
20,
T X n
. E E
E
E 0
1
2
3
4
5 6 sE(;MNT
7
8
9
10
Fig. 7 . The total amount of stored mucin per total basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5.
223
A QUANTITATIVE ANALYSIS OF RAT MUCINS
CRYPT MUCIN/MUCOSAL SURFACE AREA
TOTAL MUCIN/MUCOSAL SURFACE AREA
5 -t
60
50
z-
T
(I)
t
40
11
-
b
30 20
.
N
. E
E E E E
E
OE E
10
X
X N
20
1
11
b
V S PILOCARPINE
EASELIF€
BASELiNE V S C C " R 0 L
lo
0
0
1
=
2
3
5
4
6
7
8
9
1
10
2
3
4
5
6
7
8
9
10
SEGhENT
SEGMKT
Em control
Baseline
Fig. 8. The total amount of stored mucin per mucosal surface area in baseline (n = 5) and control (n = 2) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface.
mm2) was found to have been secreted (Fig. 16). When evaluated in terms of the total amount of product available for secretion, crypts were found to have released 43.1% and villi 27.3% of the total amount of stored mucin. In the jejunum, cr pt stored mucin accounted for 27.5% (10.1 mm3/mm ) and villus stored mucin 72.5% (26.69 mm3/mm2) of the total stored product (Fig. 16). When stimulated by pilocarpine, 86.2% (8.74 mm3/mm2) of crypt mucin and 42.5% (11.4 mm3/mm2) of villus mucin was secreted (Fig. 16). Thus 23.7% of the total amount of product was released by the crypts and 30.8% was released by the villi. In the ileum (segment 10) under unstimulated conditions crypt stored mucin represented 33.4% (12.3 mm3/mm2) and villus stored mucin 66.6% (24.5 mm3/mm2) of the total (Fig. 3). Upon accelerated secretion, 84% (10.3 mm3/mm2)of crypt mucin and 41.1% (10.1 mm3/mm2)of villus mucin was released (Fig. 16). Of the total amount of mucin secreted, 28.1% was released by crypt goblet cells and 27.4% by villus goblet cells.
Fig. 9. The volume of crypt stored mucin per mucosal surface area in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05). CRYPT MUCIN/CRYPT BASAL LAMINA BASELINE V S PILOCAPPINE
*
x
DISCUSSION Previous quantitative studies of the respiratory system have determined the distribution of mucosubstances in both Bonnet (Harkema et al., 1987) and Macaque monkeys (Heidsiek et al., 1987; Plopper et al., 1989). The present study details a quantitative approach to the study of mucin in the mammalian intestinal tract based on the methodology used in these studies. This technique detects changes in the amount and distribution of mucin :tored in the intestinal epithelium, and allows for the determination of the contribution of both functional parts of the epithelium, crypt and villus, to the mucus layer present in the intestine. Earlier studies have shown that there is a n increase proximally to distally of stored mucin in the intestinal epithelium. Quantitatively, this increase occurs predominantly in segments 1-3 in the rat small intestine and thereafter remains relatively constant. It is apparent from the present study that the villus epithelium is
*
*
T t
v
11
b -
3
X
-
2
. nE ' E E
E
0
1
=
2
3
4
5
6
7
8
9
10
SEGMNT
E m Baseline
Pilocarpine
Fig. 10. The volume of crypt stored mucin per crypt basal lamina in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05).
the major contributor to the luminal mucus blanket, with villus stored mucin representing 54% of the total in segment 1 and this figure increasing to 66.6% in the terminal ileum. Goblet cells in both the crypt and villous epithelium are involved with the baseline secretion of the contents of stored intracellular granules with the normal turnover time from synthesis to secretion being 4-6 hours (Bennett et al., 1974). Given that the mucin in both the crypt and villus goblet cells is secreted under nonstimulated conditions, it is apparent from the current study, that the villus is the predominate repository for the mucin that will contribute to the normal, baseline mucus blanket. Not only will baseline secretory rates contribute to this blanket, but the net loss of cells due to the death and exfoliation of villus goblet cells will also be a major contributor.
224
A.C. KEMPEK A N D R.D. SPECIAN VILLUS MUCIN/MUCOSAL SURFACE AREA BASELIN3 VS PILOCARPIE
40
I
1
T 1
0
-
b
20
X
,10 E E
0
1
2
3
4
7
5 6 S E m
8
9
10
Em Baseline
PllOCarpine
VILLUS MUCIN/VILLUS BASAL LAMINA EASO-1P.E V S PILOCARPIE
W
f
T
x
!
co
-t ? 0
-
to
X
.
N
E
E
E
E
0
1
=
2
3
4
5
6
7
8
9
10
SEovOIcT
EZm Basellne
Pl'ocwplne
Fig. 12. The volume of villus stored mucin per mucosal surface area Fig. 11. Light micrograph of jejunum from pilocarpine treated animals demonstrating secretion of both crypt and villus goblet cells. x 265.
To verify that this methodology could detect secretion, i t was tested under conditions that are known to diminish the amount of stored mucin in the epithelium. Previous studies on regulation of mucin secretion have documented that goblet cell mucin is released in response to cholinergic stimulation (Specian and Neutra, 1980, 1982) and this secretory response occurs primarily in the crypt. The present study confirms that crypt goblet cells secrete in response to systemic cholinergic stimulation, with a secretion of over 93% of the mucin stored in the duodenal crypt epithelium after 30 minutes. The present study also reveals for the first time that to a lesser extent, villus goblet cells are sensitive to cholinergic stimulation, with significant decreases in the amount of mucin stored in the villus epithelium for several segments following pilocarpine-induced secretion. For instance, duodenal villus cells were found to secrete over 50% of their stored product upon stimulation. Secretion from villus goblet cells was not detected by previous qualitative studies (Specian and
in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by an unpaired T-test ( P 5 0.05). Fig. 13. The volume of villus stored mucin per villus basal lamina in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05).
Neutra, 1982). Thus, villus goblet cells are the prime contributors to the baseline mucus blanket, whereas crypt goblet cells are the major contributors to the production of a n emergency layer of mucus in response to acute cholinergic stimulation. It should be emphasized, however, that villus goblet cells do contribute to this emergency mucus blanket to a heretofore unrecognized extent. The differential contribution of mucin to the overlying mucus blanket may have significant physiological implications. Numerous studies have documented differences, both histochemical (Filipe and Fenger, 1979) and immunochemical (Oliver and Specian, personal communication), in the type of mucin stored in goblet cells as they mature and migrate from the base of the
225
A QUANTITATIVE ANALYSIS OF RAT MUCINS
CONTRIBUTION TO TOTAL STOAED MJClN Baseline 8 Pilocarpine
TOTAL MUCIN/MUCOSAL SURFACE AREA
=
BASalNE V S PILOCAFPINE
50 r
+ b
"
Wf
m "llfus
40
1
?
c
30
20
7
X
x
20
1
N
E
. *E
0
2
10
E
E E
-20
0 1
=
2
3
4
7
6
5
S
E
8
9
10
-40
m
Em Baseline
Pilocarpine
Fig. 14. The total amount of stored mucin per mucosal surface area
in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by a n unpaired T-test ( P 5 0.05).
TOTAL MUCIN/TOTAL BASAL LAMINA BASELINE VS. PILOCAFPIK
I
20,
-+ I
" -
t
*
10
T
"
.z
*
X
Fig. 16. Graphic representation of the contribution of crypt and villus mucin to total stored mucin in the duodenum (segment l), jejunum (segment 5), and ileum (segment 10) for both baseline and pilocarpine-treated animals. Positive values represent stored mucin and negative values represent secreted mucin.
titative approaches to the study of goblet cells is not our proposal, we do believe that in some cases quantitative evaluation can provide insight and a more accurate picture of the true ramifications of a n experimental procedure. These data can serve a s a pilot study that will allow future studies to examine the effects of experimental manipulation on stored mucin in the small intestine, and the contribution of this mucin to the physiology of the small bowel. ACKNOWLEDGMENTS
iE
.
The authors would like to thank Dr. Dallas M. Hyde and Dr. Charles G. Plopper for their expert advice, and Mr. Dale Childress for his technical assistance.
E E
0
1
=
2
3
4
5 6 SEGMNT
7
8
9
10
Em Baseline
Pilocarplrp
Fig. 15. The total amount of stored mucin per total basal lamina in
baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences were determined by an unpaired T-test ( P 5 0.05).
crypt to the villus surface. These data suggest that mucins of different composition may contribute selectively to the protective mucus blanket under baseline and accelerated conditions. Historically, qualitative studies have provided both a rapid and reliable assessment of the effects of pathology and experimental manipulation on intestinal goblet cells. Previously, we have demonstrated that the cholinergic agonist, pilocarpine, caused secretion of crypt goblet cells (Specian and Neutra, 1982).Secretion from villus goblet cells was not reported because only a select population of cells was responsive. We were able to detect this secretory response with the current methodology, and determine that secretion significantly exceeded baseline levels. Although exclusive use of quan-
LITERATURE CITED Allen A., and B. Starkey 1974 The action of proteolytic enzymes and mercaptoethanol on mucoprotein from pig gastric mucus. BioTrans., 2: 630-633. chem. SOC. Bennett, G., C.P. Leblond, and A. Haddad 1974 Migration of glycoprotein from the Golgi apparatus to the surface of various cell types as shown by radioautography after labeled fucose injection into rats. J. Cell Biol., 60: 258-284. Filipe, M.I., and C. Fenger 1979 Histochemical characteristics of mucins in the small intestine. A comparative study of normal mucosa, benign epithelial tumours and carcinoma. Histochem. J., 11: 277-287. Forssmann, W.G., S. Ito, E. White, A. Aoki, M. Dym, and D.W. Fawcett 1976 An improved perfusion method for the testis. Anat. Rec., 188: 307-314. Harkema, J.R., C.G. Plopper, C.M. Hyde, and J.A. St. George 1987 Regional differences in quantities of histochemically detectable mucosubstances in nasal, paranasal and nasopharyngeal epithelium of the Bonnet monkey. J. Histochem. Cytochem., 35: 279286. Heidsiek, J.G., D.M. Hyde, C.G. Plopper, and J.A. St. George 1987 Quantitative histochemistry of mucosubstance in tracheal epithelium of the Macaque monkey. J. Histochem. Cytochem., 35: 435-442. Horowitz, M.I. 1967 Chemistry of the secretion layer. Ann. N.Y. Acad. Sci., 140: 784-796. Humason, G.L. 1979 Animal Tissue Techniques. W.H. Freeman, San Francisco, pp. 208-210. Klipstein, F.A., R.F. Engert, J.D. elements, and R.A. Houglhen 1984 Differences in cross protection in rats immunized with the B-
226
A.C. KGMPEK AND 11.1). SPECIAN
subunits of cholera toxin and Escherichia coli heat-labile toxin. Infect. Immun., 43: 811-816. Lukie, B.E. 1977 Studies of mucus permeability. In: Mucus Secretions and Cystic Fibrosis. Mod. Probl. Pediatr., 19: 46-53. Neutra, M.R., and J.F. Forstner 1987 Gastrointestinal mucus: Synthesis, secretion and function. In: Physiology of the Gastrointestinal Tract. L.R. Johnson, ed. Raven Press, New York, pp. 9751009. Neutra, M.R., L.J. O’Malley, and R.D. Specian 1982 Regulation of intestinal goblet cell secretion. 11. A survey of potential secretagogues. Am. J . Physiol. 242: (Gastrointest. Liver Physiol. 5): G380-387. Plopper, C.G., J.G. Heidsiek, A.J. Weir, J.A. St. George, and D.M. Hyde 1989 Tracheobronchial epithelium in the adult Rhesus
monkey: A quantitative histochemical and ultrastructural study. Am. J. Anat., 184: 31-49. Specian, R.D., and M.R. Neutra 1980 Mechanism of rapid mucus secretion in goblet cells stimulated by acetylcholine. J. Cell Biol., 85: 626-640. Specian, R.D., and M.R. Neutra 1982 Regulation of intestinal goblet cell secretion. I. Role of parasympathetic stimulation. Am. J. Physiol., 242 (Gastrointest. Liver Physiol. 5): G370-379. Specian, R.D., and M.R. Neutra 1984 Cytoskeleton of intestinal goblet cells in rabbit and monkey. The theca. Gastroenterology, 18; 1313-1325. Weibel, E.W. 1979 Stereological Methods. Academic Press, New York, vol. 1. 415 pp.
Rat Small Intestinal Mucins: A Quantitative Analysis ABBIE C. KEMPER AND ROBERT D. SPECIAN Department of Cellular Biology and Anatomy, Louisiana State University Medical Center, Shreveport, Louisiana
ABSTRACT Although much is known about the qualitative distribution of mucin-secreting goblet cells in the small intestine, the quantitative distribution of stored mucins remains undefined. The purpose of this study was to determine the distribution of neutral stored mucin in the rat small intestine by using morphometric techniques and once established, to verify that this methodology could detect secretion in animals exposed to a known mucin secretagogue. Twelve male Wistar rats (five baseline, five pilocarpine-treated, and two vehicle controls) were fixed by vascular perfusion. After a brief fixation the intestine was removed, cut into 10 equal segments, sliced, and fixed overnight. Methacrylate sections from each segment were stained with periodic acid-Schiff and toluidine blue. For morphometry, the volume of epithelium per surface area of epithelial basal lamina was calculated with a Merz grid. The volume density of stored mucin per epithelium was determined by point-counting on a square lattice grid. Volumes were related to either surface area of epithelial basal lamina or mucosal surface area. Due mostly to contributions by villus stored mucin, the total amount of product was found to increase proximally to distally in the small bowel, with the most dramatic increases occurring in the first three segments. When subjected to pilocarpine, a massive secretory response was evoked, resulting in a near total depletion of crypt stored mucin a t all levels of the small bowel. Secretion of villus stored mucin also occurred throughout the small intestine, however reaching levels of significance a t only a few points. This study describes the distribution of stored mucin in the small intestine under baseline and accelerated secretory conditions. Intestinal mucins are large, negatively-charged glycoproteins that are synthesized, stored, and secreted by goblet cells. Once secreted into the intestinal lumen, mucins form a gel that serves a s a protective barrier for the underlying epithelium. The function of this gel is multifaceted. It protects the epithelium from physical damage by luminal contents, guards against bacterial invasion (Neutra and Forstner, 19871, regulates epithelial hydration (Lukie, 1977), and interacts with secretory immunoglobulin A to produce antibody and antitoxin effects (Klipstein et al., 1984). Degradation of the mucus gel probably begins in the small intestine by lysozymal enzymes and continues in the large intestine by bacterial neuraminadases. In addition, HC1, pepsin, and pancreatic enzymes play a role in the proximal small bowel (Horowitz, 1967; Allen and Starkey, 1974). Due to luminal proteases, the mucus gel must be constantly replaced to adequately protect the epithelium. This renewal is accomplished by the baseline secretion of mucins. Baseline secretion in goblet cells occurs by an intermittent exocytosis of single mucin granules located a t the upper lateral edge of the theca. Under these conditions it has been observed that peripheral granules migrate along the lateral edges of the theca due to the influence of microtubules (Specian and Neutra, 1984). The exact mechanisms that control normal baseline secretion of mucin are not known; GI 1991 WILEY-LISS, INC
however, the effects of cholinergic stimulation on secretion in the small intestine are fairly well established. In organ culture and in vivo, acetylcholine and other cholinergic analogs, such as pilocarpine and carbachol, elicit a compound exocytosis of mucin (Specian and Neutra, 1980). This cholinergic-mediated accelerated release of mucin is limited to the crypts (Specian and Neutra, 1982) but comparable cellular mechanisms allow the rapid release of mucins from surface or villus cells in response to chemical or physical irritation (Neutra et al., 1982). Much is known about the biochemical nature of mucins and the qualitative distribution of goblet cells in the small intestine. Until now, the quantitative distribution of stored mucin has remained undefined. The purpose of this project was to determine the distribution of neutral stored mucin in the small intestine of the rat, using the methodology of Heidsiek et al. (1987). Once the baseline values were determined, a known mucin secretagogue was employed to verify that this methodology could detect secretion.
Received November 27, 1989; accepted J u n e 27, 1990. Address reprint requests to Dr. Robert D. Specian, Department of Cellular Biology and Anatomy, LSU Medical Center, 1501 Kings Highway, Shreveport, LA 71130.
220
A.C. KEMPEK AND R.D. SPECIAN
MATERIALS AND METHODS Baseline: animals and tissue processing Five male Wistar rats weighing between 450 and 550 g each were fasted 24 hours prior to vascular perfusion fixation. Rats were anesthetized with sodium pentobarbital and fixed with 2% paraformaldehyde, 2% glutaraldehyde, and 2.5% polyvinylpyrolidine in 0.1 M phosphate buffer (Forssmann et al., 1976). Following sacrifice, the small intestine was removed, cut into 10 equal lengths, and photographed (Fig. 1).Tissue was then sliced, processed, and embedded in glycol methacrylate (JB-4; Polysciences, Inc., Warrington, PA). Three micrometer cross sections from each segment were stained with periodic acid-Schiff, to demonstrate neutral mucins and toluidine blue (Humason, 1979). A section from each segment was then photographed on 35 mm film a t 10 randomly selected points using the 16 x objective of a Leitz Orthoplan photomicroscope (R. Leitz Inc., Rockleigh, NJ). Pilocarpine: animals and tissue processing Seven male Wistar rats weighing 450 to 550 g each were fasted 24 hours prior to vascular perfusion fixation. Thirty minutes prior to sacrifice, five animals received a single intraperitoneal injection of pilocarpine suspended in phosphate buffered saline (PBS) at 160 mglkg of animal weight (Specian and Neutra, 1982). Two animals, serving a s vehicle controls, received single intraperitoneal injections of PBS. All remaining methods were identical to those conducted under baseline methods.
Morphornetry Line intercept method The surface area of basal lamina per unit volume of epithelium was calculated with a Merz grid (Weibel, 1979). Counts were performed separately on 8” x 1 0 photographic enlargements for both basal lamina intersections and epithelial points. By use of the following equation:
s v = 2P1, where P1 is the number of intersections with epithelial basal lamina per test line length, the surface area of basal lamina per unit volume of epithelium was determined.
Fig. 1. Schematic representation of the r a t small intestine, depicting the 10 segments used for morphometric analysis. Coarse cross hatching represents the duodenum, unmarked area the jejunum, and fine cross hatching, the ileum
40 -
T
b
30
x
20
i l
7
N
-. E E E
10
E
n 1
Point counting The volume density of stored mucin per epithelium was determined by point counts using a square lattice grid (Weibel, 1979), and the following equation: Vv(muciniepithe1ium) = Pn(mucin)/Pt(epithelium),
where Pn equals the number of points on stored mucin and Pt the number of points on epithelium. To determine the volume of stored mucin per surface area of basal lamina, volume densities were divided by appropriate basal lamina surface areas per volume of epithelium using the following equaton: VvISv = vs
All calculations were performed separately for crypt and villus areas with the demarcation between these
T
T
t
Y
=
2
3
4
5 6 SEGWzNr
7
8
9
10
Em Crypt
Vlllus
Fig. 2. The total amount of stored mucin per mucosal surface area for baseline animals. All volumes are expressed in cubic millimeters per square millimeter of mucosal surface. n = 5.
two sectors being a line drawn under the surface epithelial basal lamina and passing through the mouths of the crypts. The volume of stored mucin could then be expressed in terms of either crypt or villus basal lamina. By using this line of demarcation a s a representation of the mucosal surface, without amplification of either crypt or villus epithelium, a third set of calculations was performed. Through the use of the line in-
A QUANTITATIVE ANALYSIS OF RAT MUCINS DUODENUM
22 1
ILEUM
Fig. 3. The percent contribution oftotal stored mucin per mucosal surface area for duodenum and ileum in baseline animals.
Fig. 4. a: Light micrograph of duodenum from baseline rats. x 265. b Light micrograph of ileum from baseline rats. x 265.
tercept method, the mucosal surface area per unit volume of epithelium was calculated. By dividing this into either crypt of villus volume density, the amount of crypt or villus stored mucin per mucosal surface area was deteremined. RESULTS Baseline animals
The total amount of stored mucin per mucosal surface area of the small intestine was found to increase
linearly from segments one to three; thereafter, the volume of stored mucin remained relatively constant throughout the remainder of the small bowel (Fig. 2). From these data, it was observed that villus stored mucin accounts for the majority of total stored glycoprotein throughout the length of the small intestine. To determine the relative contribution of both crypt and villus goblet cells to the entire mucosa, the percentage stored in each region was determined for each segment and is presented graphically for segments 1 and 10
222
A.C. KEMPEK A N D K.D. S P E C I A N
(Fig. 3). In duodenum (segment 1, Fig. 4a), crypt stored mucin constituted 46.0% and villus stored mucin 54.0% of the total amount of product present (Fig. 3). As the villus surface area increases upon leaving the duodenum, the relative contribution of the villus increases. In terminal ileum (segment 10, Fig. 4b), crypt stored mucin represents only 33.4% of the total, whereas villus stored mucin represents 66.6% (Fig. 3). The volume of stored mucin per crypt basal lamina revealed a pattern similar to that for crypt stored mucin per mucosal surface area, increasing from 2.30 x mm3/mm2in segment 1 to 3.53 x lop3 mm3/mm2in segment 3 and thereafter remaining relatively constant (Fig. 5). The volume of villus stored mucin per basal lamina follows a comparable pattern, increasing from 4.22 x lo-" mm3/mm2 in segment 1 to 13.71 x lop3 mm3/mm2 in segment 3 and leveling off (Fig. 6). In contrast to the values for total stored mucin per mucosal surface area which were additive for crypt and villus volumes, the volume of stored mucin per total basal lamina is weighted by the relative contribution of surface area of both regions. These values in segments 1, 5, and 10 were 3.07 x mm3/mm2, 8.91 x mm3/mm2, mm3/mm2, respectively (Fig. 7). and 7.83 x Vehicle control animals
Statistical evaluations between baseline (n = 5 ) and vehicle control animals (n = 2) were performed through the use of a n unpaired T-Test, where P 5 0.05 was considered significant. Of 60 tests, only segment 9 of villus stored mucin per villus basal lamina demonstrated a significant difference. No differences were found between the two groups for total stored mucin per mucosal surface area (Fig. 8). Pilocarpine- treated animals
Pilocarpine, as expected, results in the near total depletion of the stored intracellular mucin in crypt goblet cells throughout the small intestine. When quantitatively evaluated, volumes of crypt stored mucin for both the mucosal surface area and crypt basal lamina were found to be decreased significantly in all regions of the small intestine (Figs. 9, 10). Decreases in villus stored mucin were unexpected but did occur consistently throughout the length of the small intestine, although only reaching significant levels in some segments (Fig. 11). Villus stored mucin volume per mucosal surface area was decreased significantly in segments 1-6 and in segment 9, whereas the stored mucin volume per villus basal lamina decreased significantly in segments 2 and 3 (Figs. 12, 13).Total stored mucin per mucosal surface area decreased significantly a t all levels except segment 8 (Fig. 14). Similarly, quantities of total stored mucin per total basal lamina decreased significantly in all segments except 8 and 9, due to the contribution of the crypt epithelium (Fig. 15). From these results we were able to determine the percentage of stored mucin per mucosal surface area that was secreted by both crypts and villi. In the duodenum (segment l ) , 46% (7.8 mm3/mm2) of the total quantity of mucin was found in the crypts (Fig. 3). Of this amount, over 93% (7.3 mm3/mm2)was secreted (Fig. 16). Villus stored mucin in the same segment accounted for 54% (9.1 mms'/mm2) of the total amount of product present (Fig. 3).After cholinergic stimulation, 50.6% (4.6 mm3/
CRYPT MUCIN/CRYPT BASAL LAMINA EASELlhE
20,
.
n
4
E E
0 1
2
3
4
5
6
7
8
9
10
SEGMNT
Fig. 5. The volume of crypt stored mucin per crypt basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5. VILLUS MUCIN/VILLUS BASAL LAMINA EASELIfE
2
20
W
CD
-t D
b r
10
X
.
N
E
E
E
E a 1
2
3
4
5 6 sEav€NT
7
8
9
10
Fig. 6. The volume of villus stored mucin per villus basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5. TOTAL MUCIN/TOTAL BASAL LAMINA BASELINE
20,
T X n
. E E
E
E 0
1
2
3
4
5 6 sE(;MNT
7
8
9
10
Fig. 7 . The total amount of stored mucin per total basal lamina for baseline animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. n = 5.
223
A QUANTITATIVE ANALYSIS OF RAT MUCINS
CRYPT MUCIN/MUCOSAL SURFACE AREA
TOTAL MUCIN/MUCOSAL SURFACE AREA
5 -t
60
50
z-
T
(I)
t
40
11
-
b
30 20
.
N
. E
E E E E
E
OE E
10
X
X N
20
1
11
b
V S PILOCARPINE
EASELIF€
BASELiNE V S C C " R 0 L
lo
0
0
1
=
2
3
5
4
6
7
8
9
1
10
2
3
4
5
6
7
8
9
10
SEGhENT
SEGMKT
Em control
Baseline
Fig. 8. The total amount of stored mucin per mucosal surface area in baseline (n = 5) and control (n = 2) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface.
mm2) was found to have been secreted (Fig. 16). When evaluated in terms of the total amount of product available for secretion, crypts were found to have released 43.1% and villi 27.3% of the total amount of stored mucin. In the jejunum, cr pt stored mucin accounted for 27.5% (10.1 mm3/mm ) and villus stored mucin 72.5% (26.69 mm3/mm2) of the total stored product (Fig. 16). When stimulated by pilocarpine, 86.2% (8.74 mm3/mm2) of crypt mucin and 42.5% (11.4 mm3/mm2) of villus mucin was secreted (Fig. 16). Thus 23.7% of the total amount of product was released by the crypts and 30.8% was released by the villi. In the ileum (segment 10) under unstimulated conditions crypt stored mucin represented 33.4% (12.3 mm3/mm2) and villus stored mucin 66.6% (24.5 mm3/mm2) of the total (Fig. 3). Upon accelerated secretion, 84% (10.3 mm3/mm2)of crypt mucin and 41.1% (10.1 mm3/mm2)of villus mucin was released (Fig. 16). Of the total amount of mucin secreted, 28.1% was released by crypt goblet cells and 27.4% by villus goblet cells.
Fig. 9. The volume of crypt stored mucin per mucosal surface area in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05). CRYPT MUCIN/CRYPT BASAL LAMINA BASELINE V S PILOCAPPINE
*
x
DISCUSSION Previous quantitative studies of the respiratory system have determined the distribution of mucosubstances in both Bonnet (Harkema et al., 1987) and Macaque monkeys (Heidsiek et al., 1987; Plopper et al., 1989). The present study details a quantitative approach to the study of mucin in the mammalian intestinal tract based on the methodology used in these studies. This technique detects changes in the amount and distribution of mucin :tored in the intestinal epithelium, and allows for the determination of the contribution of both functional parts of the epithelium, crypt and villus, to the mucus layer present in the intestine. Earlier studies have shown that there is a n increase proximally to distally of stored mucin in the intestinal epithelium. Quantitatively, this increase occurs predominantly in segments 1-3 in the rat small intestine and thereafter remains relatively constant. It is apparent from the present study that the villus epithelium is
*
*
T t
v
11
b -
3
X
-
2
. nE ' E E
E
0
1
=
2
3
4
5
6
7
8
9
10
SEGMNT
E m Baseline
Pilocarpine
Fig. 10. The volume of crypt stored mucin per crypt basal lamina in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05).
the major contributor to the luminal mucus blanket, with villus stored mucin representing 54% of the total in segment 1 and this figure increasing to 66.6% in the terminal ileum. Goblet cells in both the crypt and villous epithelium are involved with the baseline secretion of the contents of stored intracellular granules with the normal turnover time from synthesis to secretion being 4-6 hours (Bennett et al., 1974). Given that the mucin in both the crypt and villus goblet cells is secreted under nonstimulated conditions, it is apparent from the current study, that the villus is the predominate repository for the mucin that will contribute to the normal, baseline mucus blanket. Not only will baseline secretory rates contribute to this blanket, but the net loss of cells due to the death and exfoliation of villus goblet cells will also be a major contributor.
224
A.C. KEMPEK A N D R.D. SPECIAN VILLUS MUCIN/MUCOSAL SURFACE AREA BASELIN3 VS PILOCARPIE
40
I
1
T 1
0
-
b
20
X
,10 E E
0
1
2
3
4
7
5 6 S E m
8
9
10
Em Baseline
PllOCarpine
VILLUS MUCIN/VILLUS BASAL LAMINA EASO-1P.E V S PILOCARPIE
W
f
T
x
!
co
-t ? 0
-
to
X
.
N
E
E
E
E
0
1
=
2
3
4
5
6
7
8
9
10
SEovOIcT
EZm Basellne
Pl'ocwplne
Fig. 12. The volume of villus stored mucin per mucosal surface area Fig. 11. Light micrograph of jejunum from pilocarpine treated animals demonstrating secretion of both crypt and villus goblet cells. x 265.
To verify that this methodology could detect secretion, i t was tested under conditions that are known to diminish the amount of stored mucin in the epithelium. Previous studies on regulation of mucin secretion have documented that goblet cell mucin is released in response to cholinergic stimulation (Specian and Neutra, 1980, 1982) and this secretory response occurs primarily in the crypt. The present study confirms that crypt goblet cells secrete in response to systemic cholinergic stimulation, with a secretion of over 93% of the mucin stored in the duodenal crypt epithelium after 30 minutes. The present study also reveals for the first time that to a lesser extent, villus goblet cells are sensitive to cholinergic stimulation, with significant decreases in the amount of mucin stored in the villus epithelium for several segments following pilocarpine-induced secretion. For instance, duodenal villus cells were found to secrete over 50% of their stored product upon stimulation. Secretion from villus goblet cells was not detected by previous qualitative studies (Specian and
in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by an unpaired T-test ( P 5 0.05). Fig. 13. The volume of villus stored mucin per villus basal lamina in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences between the two groups were determined by an unpaired T-test ( P s 0.05).
Neutra, 1982). Thus, villus goblet cells are the prime contributors to the baseline mucus blanket, whereas crypt goblet cells are the major contributors to the production of a n emergency layer of mucus in response to acute cholinergic stimulation. It should be emphasized, however, that villus goblet cells do contribute to this emergency mucus blanket to a heretofore unrecognized extent. The differential contribution of mucin to the overlying mucus blanket may have significant physiological implications. Numerous studies have documented differences, both histochemical (Filipe and Fenger, 1979) and immunochemical (Oliver and Specian, personal communication), in the type of mucin stored in goblet cells as they mature and migrate from the base of the
225
A QUANTITATIVE ANALYSIS OF RAT MUCINS
CONTRIBUTION TO TOTAL STOAED MJClN Baseline 8 Pilocarpine
TOTAL MUCIN/MUCOSAL SURFACE AREA
=
BASalNE V S PILOCAFPINE
50 r
+ b
"
Wf
m "llfus
40
1
?
c
30
20
7
X
x
20
1
N
E
. *E
0
2
10
E
E E
-20
0 1
=
2
3
4
7
6
5
S
E
8
9
10
-40
m
Em Baseline
Pilocarpine
Fig. 14. The total amount of stored mucin per mucosal surface area
in baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of mucosal surface. *, Significant differences between the two groups were determined by a n unpaired T-test ( P 5 0.05).
TOTAL MUCIN/TOTAL BASAL LAMINA BASELINE VS. PILOCAFPIK
I
20,
-+ I
" -
t
*
10
T
"
.z
*
X
Fig. 16. Graphic representation of the contribution of crypt and villus mucin to total stored mucin in the duodenum (segment l), jejunum (segment 5), and ileum (segment 10) for both baseline and pilocarpine-treated animals. Positive values represent stored mucin and negative values represent secreted mucin.
titative approaches to the study of goblet cells is not our proposal, we do believe that in some cases quantitative evaluation can provide insight and a more accurate picture of the true ramifications of a n experimental procedure. These data can serve a s a pilot study that will allow future studies to examine the effects of experimental manipulation on stored mucin in the small intestine, and the contribution of this mucin to the physiology of the small bowel. ACKNOWLEDGMENTS
iE
.
The authors would like to thank Dr. Dallas M. Hyde and Dr. Charles G. Plopper for their expert advice, and Mr. Dale Childress for his technical assistance.
E E
0
1
=
2
3
4
5 6 SEGMNT
7
8
9
10
Em Baseline
Pilocarplrp
Fig. 15. The total amount of stored mucin per total basal lamina in
baseline (n = 5) and pilocarpine-treated (n = 5) animals. All values are expressed in cubic millimeters per square millimeter of basal lamina. *, Significant differences were determined by an unpaired T-test ( P 5 0.05).
crypt to the villus surface. These data suggest that mucins of different composition may contribute selectively to the protective mucus blanket under baseline and accelerated conditions. Historically, qualitative studies have provided both a rapid and reliable assessment of the effects of pathology and experimental manipulation on intestinal goblet cells. Previously, we have demonstrated that the cholinergic agonist, pilocarpine, caused secretion of crypt goblet cells (Specian and Neutra, 1982).Secretion from villus goblet cells was not reported because only a select population of cells was responsive. We were able to detect this secretory response with the current methodology, and determine that secretion significantly exceeded baseline levels. Although exclusive use of quan-
LITERATURE CITED Allen A., and B. Starkey 1974 The action of proteolytic enzymes and mercaptoethanol on mucoprotein from pig gastric mucus. BioTrans., 2: 630-633. chem. SOC. Bennett, G., C.P. Leblond, and A. Haddad 1974 Migration of glycoprotein from the Golgi apparatus to the surface of various cell types as shown by radioautography after labeled fucose injection into rats. J. Cell Biol., 60: 258-284. Filipe, M.I., and C. Fenger 1979 Histochemical characteristics of mucins in the small intestine. A comparative study of normal mucosa, benign epithelial tumours and carcinoma. Histochem. J., 11: 277-287. Forssmann, W.G., S. Ito, E. White, A. Aoki, M. Dym, and D.W. Fawcett 1976 An improved perfusion method for the testis. Anat. Rec., 188: 307-314. Harkema, J.R., C.G. Plopper, C.M. Hyde, and J.A. St. George 1987 Regional differences in quantities of histochemically detectable mucosubstances in nasal, paranasal and nasopharyngeal epithelium of the Bonnet monkey. J. Histochem. Cytochem., 35: 279286. Heidsiek, J.G., D.M. Hyde, C.G. Plopper, and J.A. St. George 1987 Quantitative histochemistry of mucosubstance in tracheal epithelium of the Macaque monkey. J. Histochem. Cytochem., 35: 435-442. Horowitz, M.I. 1967 Chemistry of the secretion layer. Ann. N.Y. Acad. Sci., 140: 784-796. Humason, G.L. 1979 Animal Tissue Techniques. W.H. Freeman, San Francisco, pp. 208-210. Klipstein, F.A., R.F. Engert, J.D. elements, and R.A. Houglhen 1984 Differences in cross protection in rats immunized with the B-
226
A.C. KGMPEK AND 11.1). SPECIAN
subunits of cholera toxin and Escherichia coli heat-labile toxin. Infect. Immun., 43: 811-816. Lukie, B.E. 1977 Studies of mucus permeability. In: Mucus Secretions and Cystic Fibrosis. Mod. Probl. Pediatr., 19: 46-53. Neutra, M.R., and J.F. Forstner 1987 Gastrointestinal mucus: Synthesis, secretion and function. In: Physiology of the Gastrointestinal Tract. L.R. Johnson, ed. Raven Press, New York, pp. 9751009. Neutra, M.R., L.J. O’Malley, and R.D. Specian 1982 Regulation of intestinal goblet cell secretion. 11. A survey of potential secretagogues. Am. J . Physiol. 242: (Gastrointest. Liver Physiol. 5): G380-387. Plopper, C.G., J.G. Heidsiek, A.J. Weir, J.A. St. George, and D.M. Hyde 1989 Tracheobronchial epithelium in the adult Rhesus
monkey: A quantitative histochemical and ultrastructural study. Am. J. Anat., 184: 31-49. Specian, R.D., and M.R. Neutra 1980 Mechanism of rapid mucus secretion in goblet cells stimulated by acetylcholine. J. Cell Biol., 85: 626-640. Specian, R.D., and M.R. Neutra 1982 Regulation of intestinal goblet cell secretion. I. Role of parasympathetic stimulation. Am. J. Physiol., 242 (Gastrointest. Liver Physiol. 5): G370-379. Specian, R.D., and M.R. Neutra 1984 Cytoskeleton of intestinal goblet cells in rabbit and monkey. The theca. Gastroenterology, 18; 1313-1325. Weibel, E.W. 1979 Stereological Methods. Academic Press, New York, vol. 1. 415 pp.
