Mechanism of colonic permeation of inulin: Is rat colon more permeable than small intestine

GASTROENTEROLOGY 1995;108:12-20

Mechanism of Colonic Permeation of Inulin: Is Rat Colon More Permeable Than Small Intestine?

THOMAS Y. MA,*'* DANIEL HOLLANDER,* RICHARD A. ERICKSON,*'* HAO TRUONG,* HIEN NGUYEN,* and PAVEL KRUGLIAK* *Division of Gastroenterology, Department of Medicine, Department of Veterans Affairs Medical Center, Long Beach; and +University of California, Irvine, California

Background/Aims: Colonic epithelium is considered to be relatively tight. The colonic "pore" diameter is 6 A; therefore, colonic epithelium has generally been considered to be impermeable to hydrophilic probes with a cross-sectional diameter of > 6 A. This study examined whether rat colon is permeable to inulin, a large hydrophilic macromolecule having a molecular weight of 5000 g/mol and a cross-sectional diameter of 15 A (hydration diameter, 20 A). Methods: The colonic permeation of inulin (10 pmol/L) in vivo was investigated by perfusion of rat colonic segments. Results: There was significant colonic permeation of inulin, but tissue retention of inulin was low. The net colonic flux of inulin was strongly dependent on net water flux, showing a strong solvent drag effect. Addition of 16,16-dimethyl prostaglandin E2 decreased water flux with a corresponding decrease in inulin flux; this process seemed to be mediated by 5'-cyclic adenosine monophosphate because both the phosphodiesterase inhibitor aminophylline and dibutyryl adenosine 5'-cyclic adenosine monophosphate decreased water and inulin flux in a parallel manner. Chenodeoxycholic and taurocholic acids decreased net mucosal-to-serosal water flux but increased inulin flux. The net colonic permeation rate of inulin was higher than the small intestinal permeation rate. Conclusions: Rat colon is permeable to inulin. The higher net colonic permeability may be caused by differences in mucosal surface, permselectivity, solvent drag effect, and differences in net water fluxes of the colon and small intestine.

T he colon plays an important role in the absorption of water and regulation of stool electrolytes and

volume. 1 Another important role of the colon is its function as a barrier against the penetration of toxic substances present in the colonic lumen. Because bacterial concentrations in the colon can reach 10 billion organisms per gram of stool, intact colonic epithelial barrier function is essential to prevent the systemic penetration of these potentially injurious organisms and their toxic by-products. The two potential pathways of epithelial

penetration of luminal substances are the trancellular (through the cell) and paracellular (in-between cells) pathways.2 15 Intestinal permeability studies have been used to indirectly assess the paracellular barrier function of intestinal epithelium. 1'3'8'11 The commonly used permeability markers include polyethylene glycol 400, mannitol, cellobiose, lactulose, and 51Cr-ethylenediamine- tetraacetic acid.~l-I5 These probes are passively absorbed hydrophilic molecules and permeate across the intestinal epithelium predominantly through the paracellular pathways. 3-5 It has been previously shown that tight junctions are the major intercellular structures that restrict permeation through the paracellular pathways. Because carbohydrates are broken down to short-chain fatty acids by the colonic bacteria, the use of carbohydrate markers such as lactulose and mannitol for measurement of colonic permeability has limitations.

The small intestinal epithelium is considered to be relatively "leaky," having low epithelial resistance and high permeability.l-4 Compared with the small intestine, colonic epithelium is considered to be a much "tighter" epithelium, having higher epithelial resistance and low paracellular permeability. 1'2 The colonic "pore" diameter has been previously calculated to be 6 ]11; thus, colonic epithelium has been presumed to be impermeable to medium- or large-sized passively absorbed hydrophilic probes. Inulin is a large noncharged hydrophilic macromolecule with a molecular weight of 5000 g/tool and a molecular modeling-derived cross-sectional diameter of 15 A (hydration diameter, 20 ~).16 The cross-sectional diameter of inulin is much larger than the previously calculated colonic "pore" diameter of 6 ~.1,3 The commonly used permeability probes polyethylene glycol 400 (5.3 A), mannitol (6.7 A), lactulose (9.5 A), and 5ICr-

Abbreviations used in this paper: DBcAMP, dibutyryl adenosine- 5'-cyclic adenosine monophosphate; dmPGE2, 16,16-dimethyl prostaglandin E2,

© 1995 by the American Gastroenterological Association 0016-5085/95/$3,00

January 1995 MECHANISM OF COLONIC PERMEATION OF INULIN 13

e thy lened iamine te t raace t ic acid (10.8 ~l) have lower mo-

lecular weights and smal ler cross-sectional d iameters o 7 (<11 A), but because their size is larger than the pre-

sumed "pore" size of the colon, they were not t hough t

to pe rmea te the colon to any appreciable extent . U n t i l

recently, i t had been p resumed that the small in tes t ine

was impermeab le to larger hydroph i l i c molecules wi th a

cross-sectional d iamete r of > 1 1 ~l. 3 However , recent

s tudies from our labora tory 2 and others 6 have d isproved

this genera l ly held assumpt ion . In the present report , we

examined the poss ib i l i ty tha t colon may also be pe rme-

able to inul in.

Mater ia ls and Me thods

Highly purified [*4C]inulin (mol wt, 5000 g/mol) was

purchased from Du P o n t - N e w England Nuclear Research Products (Boston, MA). [14C]Inulin had radiochemical purity of > 9 9 % as verified by high-pressure liquid chromatography. Recrystallized sodium taurocholate and 16,16-dimethyl prostaglandin E2 (dmPGE2) were purchased from Calbiochem (La Jolla, CA). Indomethacin, sodium chenodeoxycholate, dibutyryl adenosine 5'-cyclic adenosine monophosphate (DBcAMP), and aminophylline were purchased from Sigma Chemical (St. Louis, MO). All other chemicals were of reagent grade.

Animals

Nonfasted male Fisher 344 rats (body wt, 180-290 g) obtained from Charles River Breeding Laboratories (Wil- mington, MA) were used in all experiments.

In Vivo Colonic Perfusion

In vivo colonic perfusions were performed as previously described. ~9'2° After an overnight fast, the rats were anesthe- tized with intraperitoneal pentobarbital sodium (40 mg/kg). After midline incision of the abdomen, 10 cm of midcolonic segments were cannulated in a separate group of rats. The colonic segments were then flushed with normal saline fol- lowed by air flush to remove residual contents, and the perito- neal cavity was closed. The animal was allowed to awaken and

was placed in a Plexiglas restraint cage. A forced-air heating device and feedback temperature controller were used to moni- tor and maintain the animal's body temperature at 37°C. An external recirculating pump was used to recirculate the perfusate at a constant flow rate. The perfusate was recirculated from an enclosed continuously stirred reservoir at 1 mL/min unless otherwise specified. The basal perfusate solution consisted of Krebs ' -phosphate saline buffer (pH 6.5,300 mOsm) containing 10 ~tmol/L inulin with a trace of radioactive [14el- inulin. Where indicated, selected agents were added to the perfusate solution before the start of 3 hours of perfusion. The osmolarity of the perfusate solution was varied by the addition of mannitol. 2'18'2° The perfusate volume was 30 mL and con- tained a total of 300 nmol of inulin at the onset of the perfu-

sion. Outflow from the intestinal segment was drained by gravity into the enclosed reservoir. At the end of the 3-hour perfusion period, the perfused intestinal segments were flushed with air. The absorption rate of inulin was calculated by its

disappearance from the perfusate solution minus the total inulin retained in the perfused intestinal tissue. Duplicate 100-

~L samples were taken from the perfusate solution every 30 minutes throughout the 3-hour colonic perfusion, and radioac-

tivity was counted to a 1% counting error in a scintillation counter (Beckman LS 9000 scintillation counter; Beckman, Fullerton, CA). Absorption values were calculated per 100 cm of length after standardized stretching and drying of the intestinal segments or per 100 mg mucosal protein con- tent. 18-21 Previous studies have shown intestinal length to

correlate best with intestinal surface area. 21 The net water absorption or secretion was calculated from the differences between the initial and final volume of the perfusate after 3 hours of perfusion and was expressed per hour.

Colonic Tissue Retention

The colonic tissue retention of inulin was measured at the end of 3 hours of perfusion using previously described methods. 2'='23 For the purpose of measuring colonic tissue retention of inulin, the perfused segments were flushed twice with air. Segments (4 cm) of perfused colon were removed and opened along the mesenteric side. Colonic sheets were gently blotted with filter paper and allowed to dry. The intestinal sheets were placed individually in scintillation vials containing 0.5 mL of 1 mol/L NaOH. The colonic sheets were digested and liquified by placing the vials in an oven at 90°C for 2 hours. On cooling to room temperature, 0.5 mL of 1 mol/L HC1 and 5 mL of scintillation fluid were added and radioactiv- ity was counted in a scintillation counter.

Ouantitation of Intestinal and Colonic Morphology

On selected animals, 4-cm segments of proximal, mid, and distal small intestines and proximal and distal colon were removed, opened longitudinally on a piece of filter paper, and fixed in 10% buffered formalin. These samples were processed for routine histology, cut longitudinally, and stained with H& E. Coded stained histological samples were examined using a computerized image analyzer system composed of a standard binocular microscope (model BH-2; Olympus Optical Co., Ltd., Tokyo, Japan) with a drawing attachment (model BHW- DA; Olympus Optical Co.) placed on a 20 × 24- inch digitiz- ing board (Numonics 2200; Jandel Scientific, Corte Madera, CA) that was attached to an IBM AT compatible computer using the Sigma Scan program (Jandel Scientific) to measure various morphological parameters. The morphological parameters measured were the total cross-sectional surface length (including the villi and crypts), the longitudinal length of the serosal surface, and the individual villus and/or crypt surface length. At least 20 villi or crypts per specimen were measured from each segment (n = 3 rats). All measurements were per-

14 MA ET AL. GASTROENTEROLOGY Vo1.108, No. 1

4 0

m ~ , 30

~ 20 3 E E.. ~ o I = ' ' ~ 1 0

60 120 180

Perfusion time (rain)

Figure 1. Disappearance of inulin from the luminal perfusate during the 3-hour recycling colonic perfusion period. The basal perfusate solution consisted of Krebs'-phosphate saline buffer (pH 6.5; 300 mOsm) containing 10 pmol/L inulin and tracer [14C]inulin. Values represent mean disappearance of inulin from the luminal perfusate at 30-minute increments (n = 9 rats) during the 3-hour perfusion period. Values were fitted to the linear plot with the linear regression plot.

formed by the same person to prevent interobserver difference in measuring technique.

Statistics

All values are reported as means + SE, and statistical comparisons of control and experimental conditions are made using Student's t test and analysis of variance.

Resul ts

Colonic Permeation and Tissue Retention of Inulin

At physiological luminal osmolarity of 300 mOsm/L, there was a significant colonic permeation of inulin (12 nmol '10 c m - t ' h -1 or 4 .0%'10 c m - l ' h - 1 ) .

The rate of inulin disappearance from the luminal perfusate is shown in Figure 1. There was a linear rate of inulin disappearance from the colonic lumen during the 3-hour perfusion period. Slight deviation or increase in disappearance rate during the first 30 minutes of perfusion was presumably caused by tissue adsorption or extracellular penetration of inulin during the initial stages of absorption as reported for the small intestine in our previous report. 2 Inulin equilibration in the extracellular space occurs rapidly within the first 15 minutes. 2'23 The mean

tissue retention of inulin by the colon at the end of the 3-hour perfusion period was 0.8 nmol/10 cm. In contrast, we previously found small intestinal tissue retention to be higher at 1.1 nmol/10 cm at the end of the 3-hour perfusion period. 8 The colonic tissue retention of inulin

at the end of the 3-hour perfusion period was constant and independent of colonic inulin absorption, indicating that the tissue retention or adsorption of inulin was unre- lated to colonic permeation or absorption rate of inulin. Because carbohydrates are metabolized by the colonic bacteria, the possible bacterial degradation of inulin was determined by incubation of isolated colonic segments in perfusate solution at 37°C for 3 hours. The perfusate solution was sampled (1 mL) at 1-hour intervals, and the changes in inulin composition were detected using high- pressure liquid chromatography. At the end of the 3- hour incubation period, >9 5 % of inulin remained intact, indicating < 5 % colonic degradation of inulin.

Contribution of Solvent Drag to Inulin Permeation

The relative contribution of solvent drag and diffusion to colonic inulin flux was determined. We modi- fied the net colonic water flux by changing the luminal osmolarity. Varying the luminal osmolarity from 300 to 600 mOsm/L resulted in a linear change (r = 0.98) in net mucosal-to-serosal colonic water flux (Figure 2A). There was a linear relationship (r = 0.99) between net water flux and inulin absorption (Figure 2B), indicating a close dependence of inulin absorption on net mucosal- to-serosal water flux. The relative contribution of diffusional and convective movement of inulin (Figure 2C) was calculated using previously reported methods. 2°'24'25

Combining Fick's equation and the Kedem-Katchalsky equation, 25 the relative contributions of diffusive and convective movement on solute flux can be expressed quantitatively:

Js = Pd(C2 -- C1) -F Jv(1 - (sf) [(c1 + C2)/2],

in which J, is the net solute flux resulting from combined contributions of diffusion and solvent drag, Pa is the diffusional permeability coefficient,J~ is the solvent flow rate, (1 - Of) is the coefficient of solvent drag, and Cz and Ce are the concentrations of solute in the luminal and serosal side of the epithelium. Using the equation and the absorption data obtained during water flux studies, the relative contributions of diffusive and convective movement were calculated as previously described. 2°'24

As shown in Figure 2B, increasing net colonic water flux resulted in increasing net inulin flux. The relative contribution of convective movement increased with increasing net water flux (Figure 2C).

Comparison of Colonic Versus Small Intestinal Permeation

We compared the rate of colonic permeation of inulin to small intestinal permeation data reported in


A

J :

O ¢ )

o E e.-

:K

m

e-

C n

3 0

2 0

1 0

0

- 1 0 , , , , , 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0

Luminal osmolarity (mosM)

B

i i i |

10 0 10 2 0 3 0

1 5

1 0

W a t e r f l u x ( m l / 1 0 0 c m / h r )

o _>

=>* _o,-

o

t -

o

¢~

-g , r -

m

a :1

"o .,,,,,

C

1 . 0

0 .S

0.0 0.0 1 0 . 0

Water flux(181 ml/lO c~hr)

2 0 . 0

Figure 2. (A) Effect of varying perfusate osmolarity on net colonic mucosat-to-serosal water flux. The perfusate osmolarity was varied from 300 to 600 mOsm/L (n = 3 - 9 rats). Plotted values represent mean water flux at various perfusate osmolarity. (B) Relationship between net colonic mucosal-to-serosal water flux and inulin absorption (n = 3 - 9 rats). (C) Relative contribution of diffusive vs. convective movement on net colonic mucosal-to-serosal flux of inulin at varying rates of water flux.

C r - .

g ~

"6E E T -

3 0 -

T

2 0 ¸

10

C W

0 Colon Small intestine

Figure 3. Comparison of small intestine and colon net mucosal-to- serosal flux of inulin and water. El, inulin flux in nmol-lO cm-l.h 1; E3, water flux in mL.lO0 cm-<h -1 (n = 8 - 9 rats).

our previous study. 2 The permeability studies were per-

formed under similar physiological and perfusion conditions. The net colonic flux rates of inulin and water were three-fold higher than the net small intestinal flux rates (Figure 3). The differences in colonic and small intestinal permeability were even greater when inulin permeation was calculated per milligram mucosal protein rather than per centimeter length. The colonic permeation rate of inulin was 354.0 nmol '100 mg prote in- l 'h -1 compared with the small intestinal permeation rate of 55.3 nmol" 100 mg protein-l"h -1. The comparison of reported colonic and small intestinal permeation rates of other permeability markers mannitol, polyethylene glycol 400, and polyethylene glycol 900 are shown in Table 1. The net permeation rates of these probes were also much higher in the colon than in the small intestine.

Morphometric Comparison of Colonic and Small Intestinal Absorptive Surface

The possible differences in absorptive surface area of the small intestine and colon were determined mor- phometrically, similar to previously reported methods. 4 The representative light micrographs of jejunal and co- Ionic mucosa are shown in Figure 4. The prominent protruding "fingerlike" villi were found in the jejunal specimens. The colonic segments showed flat villi with elongated crypts. The ratio of villus surface length to serosal length was much higher in the small intestine than the colon (Figure 5A). There was a progressive decrease in the relative ratio of measured villus surface length to serosal length from the proximal small intestine to distal colon. In contrast, the ratio of crypt surface length to serosal length was much higher in the colon than the small intestine (Figure 5B). The ratio of the

16 MA ET AL. GASTROENTEROLOGY Vol. 108, No. 1

Table 1. Compar ison of Repor ted Intest inal Permeat ion Rate of Polyethy lene Glycol 400 , Manni to l , and Polyethy lene Glycol 9 0 0 Across the Smal l In test ine and Colon

Polyethylene Polyethylene Mannitol glycol 400 glycol 900

Small intestine absorption (#mol- 100 cm- l .h -1) 10.1 _+ 0.5 13.0 _+ 0.5 6.9 _+ 0.3

Colonic absorption (~Lmol • 100 cm- l .h -1) 26.2 _+ 2.0 a 42.2 _+ 2.0 a 31.1 _+ 2.8 a

Small intestine absorption (pmol. 100 mg protein-~.h -1) 12.5 _+ 0.6 16.2 _+ 0.6 9.1 _+ 1.2

Colonic absorption Oxmol. 100 mg protein ~-h ~) 67.4 _+ 7.8 a 140.5 + 6.6 ~ 104 _+ 9.5 ~

Small intestine water absorption (mL. 100 cm -z- h -~) 9.4 4- 0.4 8.9 _+ 0.4 5.1 4- 0.5

Colonic water absorption (mL. 100 cm- l .h 1) 23.0 +_ 0.5 a 34.5 4- 0.8 ~ 35.3 + 5.2 ~

NOTE. The intestinal segments were perfused with 1 mmol/L concen- tration of permeability probes. 19-21

ap < 0.01 when compared with small intestinal absorption.

combined villus and crypt surface length to serosal length was similar between the small intestine and the colon (Figure 5C).

Effect of dmPGE2 and 5'-Cyclic Adenosine Monophosphate on Colonic Permeation of Inulin

In this set of experiments, we examined the possible effect of dmPGE2 on colonic inulin and water flux. Addition of dmPGE2 to the luminal perfusate caused a decrease in net mucosal-to-serosal water flux with a corresponding decrease in net inulin flux (Figure 6). Be- cause prostaglandin E2 stimulates 5'-cyclic adenosine monophosphate (cAMP) production, the effect of the cAMP analogue DBcAMP on inulin permeation was also examined. DBcAMP caused a decrease in net colonic water and inulin flux (Figure 5). Intracellular cAMP is broken down by phosphodiesterases. The phosphodiesterase inhibitor aminophylline blocks this action and leads to an increase in intracellular cAMP. Addition of aminophylline to the perfusate also decreased both net water and inulin flux (Figure 6).

Effect of Bile Acids on Inulin Permeation

The colon is exposed to bile acids under both physiological and pathological conditions. The effect of

the bile acids chenodeoxycholic acid and taurocholic acid

Figure 4. Representative light micrographs of normal small intestinal and colonic mucosa. Representative light micrograph of (A) rat small intestinal mucosa and (B) colonic mucosa fixed in 10% formalin and stained with H&E (original magnification lOOx). (C) Light micrograph of colonic mucosa perfused with chenodeoxycholic acid. The colonic segment was perfused with perfusate solution containing 5 mmol/L chenodeoxycholic acid for a 3-hour perfusion period.


11t e

m ~ =D ~0

g ; o

q

o

o o

o ~

12 ¸

PROX $1 MIO SI DIST SI PROX LI DIST LI

I n t e s t i n a l regions

PROX SI MID 81 DIST Sl PROX U DIST LI


1 2

c

|

°$ O m

O ~ =~ 4

~ 0

0 PROX SI MID SI DIST SI PROX LI DIST U


Figure 5. Relative differences in crypt and villus surface length of small intestine and colonic mucosat surface as determined by two- dimensional morphometric analysis. Small intestinal and colonic ratio of (A) villus surface length to serosal length, (B) crypt surface length to serosal length, and (C) combined villus and crypt surface length to serosal length was compared (n = 3 rat segments).

3< "O :

(0 . , ,

E u g,o ~ g "6E E. .

! - lU - - 3

30

20

1 0

T

.k

- ' r " 7-

0 Control D M P G D B c A M P A m i n o p h y n i n e

Figure 6. Effect of dmPGE2, DBcAMP, and aminophylline on net co- Ionic mucosal-to-serosal inulin and water flux (n = 3 - 9 rats), m, net inulin flux in 10 nmol.lO0 cm-l.h-~; C], water flux in mL.lO0 cm ~-h ¢. *P < 0.01 compared with controls.

on colonic permeation of inulin was examined. Cheno- deoxycholic acid (5 mmol/L) and taurocholic acid (10 retool/L) caused a significant decrease in net water flux (Figure 7). In contrast to what would be expected based on its effect on solvent drag, there was a marked increase in net inulin flux. To better understand these seemingly contradictory findings, the effect of bile acids on intestinal morphology was examined. The bile acids caused extensive damage of the perfused colonic segments with severe denudation of colonic epithelial cells (Figure 4C). This resulted in large portions of the colonic lamina propria and crypts being directly exposed to luminal contents. Thus, bile acid-induced increase in colonic permeability was probably caused by the extensive dis-

40

X "O :

C U . . , 3 0

E o

"SE

2 0

1 0

C - - 3

0 Control Tauro (SmM) Tauro ( fOruM) C h e n o (SmM)

Figure 7. Effect of chenodeoxycholic acid and taurocholic acid on net colonic mucosal-to-serosal inulin and water flux (n = 3 - 9 rats), i , net inulin flux in 10 nmol.lO0 cm-~-h ~; [2, water flux in mL.lO0 cm ~.h ~. *P < 0.01 compared with controls.


ruption of the colonic epithelial barrier caused by the bile acids.

Discussion

The present data indicate that the rat colon is permeable to inulin. This finding is contrary to the pres- ently held assumption that the colon has a relatively "t ight" epithelium, having a small functional "pore ra- dius" of 3 ~i and being impermeable to molecules with a cross-sectional diameter of > 6 ~.1 In our previous investigation, 2 we showed that the small intestine was permeable to inulin, a commonly used "nonabsorbable" marker. In contrast to the small intestine, the colon has been reported to have a "tighter" epithelium with high transepithelial electrical resistance. 1'27-29 The reported epithelial electrical resistance of the colon is between 2 50 and 500 ~/cm 2, whereas small intestinal resistance is between 20 and 100 ~"~/cm 2.1'23'27-29 Despite these electri-

cal measurements, the present data indicate that the rat colon is permeable to a large electrically neutral probe such as inulin (Figure 1).

As shown in Figure 2, colonic inulin permeation was strongly dependent on solvent drag. Because the paracellular pathways are the major channels of water flux, 1'2'3° this finding is consistent with the paracellular pathways being the major pathway of inulin flux. Whereas the small intestine has a pivotal role in nutrient absorption, the colon plays a key role in the absorption of water and electrolytes, thereby regulating the final stool consistency and volume. The fluid absorptive capacity of the colon has been well established. 1 The colon is more efficient than the small intestine in absorption of water and has the ability to absorb water against a high osmotic gradi- ent. 1 As shown in Figure 3 and Table 1, the net water flux was much greater in the colon than in the small intestine. It is likely that the observed higher colonic permeability of inulin was in part caused by the higher solvent drag effect in the colon.

Prostaglandins have been shown to affect intestinal water and electrolyte transport and contribute to inflammatory diarrhea) *-34 Therefore, we studied the influence of dmPGE2 on colonic water and inulin flux (Fig- ure 6). We found that dmPGE2 decreased net colonic mucosal-to-serosal water and inulin flux. Because the effect of dmPGE2 on water and electrolyte transport has been shown to be mediated through stimulation of cAMP production, 32-34 the possible role of cAMP on colonic inulin and water flux was examined. Both the cAMP analogue DBcAMP and the phosphodiesterase inhibitor aminophylline decreased net flux of inulin and water in a parallel manner as dmPGE2, suggesting that cAMP

may be the intracellular mediator responsible for the observed effect. In addition to their effect on water flux, dmPGE2 may also directly affect tight junctional structures. 35

Normally, <0.5 g of bile acids reaches the colon in 24 hours. In pathological situations, such as in active Crohn's disease or after extensive ileal resection, large amounts of bile acids may be presented to the colon. The bile acid-induced increase in colonic permeability (Figure 7) was partly caused by its injurious effect on colonic epithelium (Figure 4C). Bile acids decreased net colonic water flux but also caused extensive mucosal injury, producing large openings in the colonic mucosa (Figure 4C). Recent in vitro studies in our laboratory indicate that bile acids can also cause direct injury or disruption of tight junctional complexes (unpublished observations, July 1992).

Under the present experimental conditions, the rat colon was more permeable than the small intestine to inulin (Figure 3). This finding contradicts the existing view that the colon is a "tighter" epithelium than the small intestine and therefore less permeable. 1'23'27-29'3s In

support of our present findings, other studies have shown higher colonic permeability for a wide range of hydrophilic molecules, 18'19'36'37 including Na +, CI-, mannitol,

polyethylene glycol 400, and polyethylene glycol 900 (Table 1). Under similar perfusion conditions used in the present experiment, Powell and Malawer 36 reported greater net colonic fluxes o fNa + and CI- compared with the small intestine in rats. Davis et al. 37 also found net colonic absorption of Na + and C1- to be higher than in the small intestine. They attributed the higher net co- Ionic flux to high "back flux" or "leakiness" in the small intestinal shunt pathway compared with the "tighter" colonic epithelia. There is also a "greater relative solute transport with respect to water transport in the colon than in the small intestine") 6 The net osmolarity of transported solutes has been calculated by Powell and Malawer to be 325 4- 3 mOsm/mL in the jejunum, 405 4- 10 mOsm/mL in the ileum, and 515 4- 14 mOsm/ mL in the colon. Parsons 38 also reported similar solute transport of 357 ~mol/mL for the jejunum, 382 btmol/ mL for the ileum, and 514 ~mol/mL for the colon. In contrast to our present findings, Chadwick et al) 9 found a progressive decrease in permeability of polyethylene glycol 400 from the proximal small intestine to the colon in humans. The reason for this discrepancy is unclear but may be caused by differences in animal species.

It has been shown that there is a difference in permselectivity between the colon and the small intestine) The small intestine has more cationic selectivity, whereas the


colon has more anionic selectivity) '36'4° Perhaps there exists a greater colonic permselectivity to noncharged molecules compared with the small intestine. The pres- ence of such permselectivity may explain the difference in permeability between the colon and small intestine to inulin.

Another possible explanation for the high rat colonic permeability to inulin may be related to the differences in serosal-to-mucosal "back flux" between the colon and small intestine. 1'37'41 The net solute flux or mucosal absorption is the net difference between mucosal-to-serosal and serosal-to-mucosal flux of a solute. The small intestine is "leaky" and has relatively high permeability to water-soluble solutes in both mucosal-to-serosal and serosal-to-mucosal directions. 9'25"34 In contrast, the colon has

"moderate" mucosal-to-serosal permeability but has minimal "back leak" or serosal-to-mucosal back flux of solutes) 7 Because "back" flux through the paracellular pathway of the colonic epithelium is relatively small, the net colonic absorption of solutes closely reflects unidirec- tional mucosal-to-serosal flux. 37 On the other hand, there is a large "back flux" or "back leak" through the leaky shunt pathway of the small intestine, causing the net small intestinal absorption to be dependent on both mucosal-to-serosal and serosal-to-mucosal flux.

Another possible explanation for the "relatively high" net colonic permeability may be related to the greater crypt surface area of the colon compared with the small intestine (Figures 3 and 4). Whereas the small intestinal surface is covered mostly by villi, the colonic mucosal surface consists mostly of crypts (Figure 4). In a freeze- fractured morphometric study, Marcial et al. 4 calculated the contributions of crypts and villi to net ileal passive transjunctional ion flow. They found that villi consti- tuted 87% and the crypts 15% of the total surface of ileal mucosa. Using Claude's equation 5 relating tight junction resistance to tight junction structure and taking into account the differences in serosal-to-mucosal ampli- fication of villi and crypts, they calculated that the majority (73%) of ileal paracellular conductance was attributable to the crypts and concluded that "crypt com- partment may be responsible for the majority of net ileal paracellular conductance." If the majority of paracellular conductance takes place through the crypt region as calculated by Marcial et al., 4 then perhaps the high crypt surface area present in the colonic mucosa could account for the high paracellular permeability of this organ.

We conclude that there is a sizable colonic permeation by inulin in the rat. Colonic permeation of inulin occurs by a combination of diffusion and solvent drag and is strongly modulated by solvent drag. Under physiological

conditions, the net mucosal-to-serosal flux of inulin is higher in the colon than the small intestine of rats. The higher net colonic permeability to inulin may be caused by differences in the back flux, permselectivity, solvent drag affect, and distribution of villus and crypt surfaces of the colon and small intestine, dmPGE2 decreases colonic permeation of inulin, and this effect seems to be mediated by cAMP. Bile acids decrease colonic water flux but cause "breakage" or large openings in the mucosa, leading to a net increase in inulin permeation.

References

1. Powell DW. Intestinal water and electrolyte transport. In: Johnson LR, ed. Physiology of the gastrointestinal tract. 2nd ed. New York: Raven, 1987 :1267-1305 .

2. Ma TY, Hollander D, Erickson RA, Truong H, Krugliak P. Is the small intestinal epithelium truly " t ight" to inulin permeation? Am J Physiol 1991;260:G669-G676.

3. Madara JL. Loosening tight junction lessons from the intestine. J Clin Invest 1989 ;83 :1089-1094 .

4. Marcial MA, Carlson SL, Madara JL. Partitioning of paracellular conductance along the ileal crypt-villus axis: a hypothesis based on structural analysis with detailed consideration of tight junction-structure function relationships. J Membr B io11984;80:59- 70.

5. Claude P. Morphological factors influencing transepithelial permeability: a model for the resistance of the zonula occludens. J Membr Biol 1978 ;39 :219-232 .

6. Pappenheimer JR, Reiss KZ. Contribution of solvent drag through intercellular junctions to absorption of nutrient by the small intestine of the rat. J Membr Biol 1987 ;100 :137-148 .

7. Hollander D, Ricketts D, Boyd CAR. The importance of probe molecular geometry in determining intestinal permeability. Can J Gastroenterol 1988 ;2 :35a-38a .

8. Hollander D. Crohn's d isease- -a permeability disorder of the tight junction? Gut 1988; 29 :1621-1624 .

9. Menzies IS. Transmucosal passage of inert molecules in health and disease. In: Sdadhauge E, Heintze K, eds. Intestinal absorption and secretion. Higham, MN: MTP, 1984 :527-544 .

10. Peters T J, Bjarnason I. Uses and abuses of intestinal permeability measurements. Can J Gastroenterol 1988 ;2 :127-132 .

11. Anonymous. Intestinal permeability (ed). Lancet 1985 ;1 :256 - 258.

12. Bjarnason I, Williams P, Smethurst P, Peters T J, Levi AJ. Effect of nonsteroidal antiinflammatory drugs prostaglandins on the permeability of the human small intestine. Gut 1986 ;27 :1292 - 1297.

13. Hamilton I, Cobden I, Rothwell J, Axon ATR. Intestinal permeability in celiac disease: the response to gluten withdrawal and sin- gle-dose gluten challenge. Gut 1982; 23 :202-210 .

14. Jenkins RT, Rooney P J, Jones DB, Bienenstock J, Goodacre RL. Increased intestinal permeability in patients with rheumatoid ar- thritis: a side-effect of oral nonsteroidal antiinflammatory drug therapy? Br J Rheumatot 1987 ;26 :103 -107 .

15. Ma TY, Hollander D, Katz KD, Krugliak P. Crohn's disease and NSAID enteropathy. Gastroenterology 1990; 99 :1190-1192 .

16. Middleton E. The molecular configuration of inulin: implication for ultrafiltration theory and glomerular permeability. J Membr Biol 1977 ;34 :93 -101 .

17. Hollander D. Intestinal absorption of vitamins A, E, D and K. J Lab Clin Med 1981 ;97 :449 -462 .

18. Hollander D, Koyama S, Dadufalza V, Tran DQ, Krugliak P, Ma


TY, Ling KY. Polyethylene glycol 900 permeability of rat intestinal and colonic segments in vivo and brush border membrane vesi- cles in vitro. J Lab Clin Med 1989;113:505-515.

19. Krugliak P, Hollander D, Ma TY. Basic mechanisms of mannitol permeability of perfused rat intestine (abstr). Gastroenterology 1992; 102:A219.

20. Krugliak P, Hollander D, Ma TY, Tran D, Dadufalza VD, Katz KD, Le K. Mechanisms of polyethylene glycol 400 permeability in perfused rat intestine. Gastroenterology 1989;97:1164-1170.

21. Meshkinpour H, Smith M, Hollander D. Influence of aging on the surface area of the small intestine in the rat. Exp Gerontol 1981; 16:399-404.

22. Said HM, Smith R, Redha R. Studies on the intestinal surface acid microclimate: developmental aspects. Pediatr Res 1987; 22:497 -499.

23. Schultz SG, Fuisz RE, Curran PF. Amino acid and sugar transport in rabbit ileum. J Gen Physiol 1966;49:849-886.

24. Kedem O, Katchaisky A. Thermodynamic analysis of the permeability of biological membranes to non-electrolytes. Biochem Bio- phys Acta 1958;27:229-246.

25. Schaefer JA, Andreoli TE. Principles of water and nonelectrolyte transport across membranes. In: Andreoli TE, Hofman JF, Fanes- til DD, Schultz SG, eds. Physiology of membrane disorders. New York: Plenum, 1986:177-190.

26. Said HM, Blair JA, Lucas ML, Hilburn ME. Intestinal surface acid microclimate in vitro and in vivo in the rat. J Lab Clin Med 1986; 107:420-424.

27. Curran PF, Schultz GF. Na, CI, and water transport by colon. J Gen Physiol 1960;43:555-571.

28. Frizzell RA, Schultz SG. Ionic conductance of extracellular shunt pathway in rabbit ileum. J Gen Physiol 1972;59:318-346.

29. Schultz SG, Frizzell RA, Nellans HN. Active sodium transport and electrophysiology of rabbit colon. J Membr Biol 1977;33:351- 384.

30. Pappenheimer JR. Physiological regulation of transepithelial im- pedance in the intestinal mucosa of rats and hamsters. J Membr Biol 1987; 100:137-148.

31. Bjarnason i, Smethurst P, Fenn GC, Lee CE, Levi AJ. Misoprostol reduces indomethacin induced changes in human small intestinal permeability. Gastroenterology 1988;94:A37.

32. Kimberg DV, Field M, Johnson J, Henderson A, Gershon E. Stimu-

lation of intestinal mucosal adenyl cyclase by cholerenterotoxius and prostaglandins. J Ciin Invest 1971;50:1218-1230.

33. Krugliak P, Hollander D, Le K, Ma TY, Dadufalza VD, Katz KD. Regulation of polyethylene glycol 400 intestinal permeability by endogenous and exogenous prostanoids. Influence of non-steroi- dal anti-inflammatory drugs. Gut 1990;31:417-421.

34. Sernka T J, Rood RP, Mah MY, Tseng CH. Antiabsorptive effects of 16,16-dimethyl prostaglandin E= in isolated rat colon. Prosta- glandins 1982;23:411-426.

35. Duffey ME, Hainan B, Ho S, Bentze CJ. Regulation of epithelial tight junction permeability by cyclic AMP. Nature 1981; 294:451- 453.

36. Powell DW, Malawer SJ. Relationship between water and solute transport from isosmotic solutions by rat intestine in vivo. Am J Physiol 1968;215:49-54.

37. Davis GR, Santa Ana CA, Morawski SG, Fortran JS. Permeability characteristics of human jejunum, ileum, proximal colon and distal colon: results of potential difference measurements and unidi- rectional fluxes. Gastroenterology 1982;83:844-850.

38. Krugliak P, Hollander D, Schlaepfer CC, Nguyen H, Ma TY. Mecha- nism of mannitol permeability of small and large intestine in the rat. Dig Dis Sci 1993;38:796-801.

39. Chadwick VS, Phillips SF, Hofmann AF. Measurements of intestinal permeability using low molecular weight polyethylene glycols (PEG 400). Gastroenterology 1977; 73:247-251.

40. Powell DW. Intestinal conductance and permselectivity changes with theophylline and choleragen. Am J Physio11974; 227:1436- 1443.

41. Madara JL, Trier JS. Functional morphology of the mucosa of the small intestine. In: Johnson LR, ed. Physiology of the gastrointestinal tract. 2nd ed. New York: Raven, 1987:1209-1249.

Received August 4, 1993. Accepted June 6, 1994. Address requests for reprints to: Thomas Y. Ma, M.D., Ph.D., Divi-

sion of Gastroenterology, DVA Medical Center, 5901 East Seventh Street, Long Beach, California 90812. Fax" (310) 494-5675.

Supported in part by a Merit Review grant to T.Y.M. from the Veterans Administration and Goldsmith Research Foundation.

The authors thank Violetta Dadufalza for excellent technical assis- tance.

Mechanism of colonic permeation of inulin: Is rat colon more permeable than small intestine

Documents

Transcript of Mechanism of colonic permeation of inulin: Is rat colon more permeable than small intestine