An oligopeptide permeates intestinal tight junctions at glucose-elicited dilatations

An oligopeptide permeates intestinal tight junctions at glucose-elicited dilatations

GASTROENTEROLOGY 1991;100:719-724 An Oligopeptide Permeates Intestinal Tight Junctions at Glucose-Elicited Dilatations Implications for Oligopeptide ...

3MB Sizes 0 Downloads 21 Views

GASTROENTEROLOGY 1991;100:719-724

An Oligopeptide Permeates Intestinal Tight Junctions at Glucose-Elicited Dilatations Implications for Oligopeptide Absorption KANIT ATISOOK and JAMES L. MADARA Departments of Pathology.Brighamand Women'sHospital and HarvardMedicalSchool, and the Harvard DigestiveDiseasesCenter. Boston, Massachusetts

Turnover of the Na+-glucose cotransporter in the apical membrane of intestinal absorptive cells elicits alterations in tight-junction structure including the appearance of intrajunctional dilatations. Paralleling these structural responses, epithelial permeability to ions and nutrient-sized solutes increases. However, it is not known how these observed permeability changes specifically relate to the structural alterations elicited by glucose. Using a hemeconjugated peptide tracer (MP°11; mol wt, = 1900}, the present study shows that the glucose-elicited tight-junction dilatations are specific anatomical sites of junctional permeation. This peptide tracer penetrates tight junctions selectively at sites of dilatations and is detected focally within the paracellular space. This s a m e t r a c e r does not penetrate junctions when glucose is not present. A heineconjugated macromolecule (horseradish peroxidase; mol wt, ~ 40,000) is excluded by both glucoseexposed and glucose-unexposed tissues. The results of this study show a paracellular pathway for small peptides that is regulated during Na+-glucoseactivated absorption. It is speculated that the paracellular pathway may contribute to the meal-related oligopeptide absorption that is known to o c c u r a n d h a s p r e v i o u s l y b e e n wholly attributed to the transcellular route. 'n vivo exposure of mammalian small intestinal linert .epithelium to glucose elicits enhanced clearance of solutes such as creatinine and inulin from the lumen (1,2). As assessed by in vitro studies using either impedance techniques (3) or 'standard directcurrent and flux techniques (2), this effect of glucose seems to be caused by reversible enhancement of tight-junction (TJ) permeability. Because these functional and structural responses are Na ÷ dependent

and phlorizin inhibitable (2), turnover of the Na ÷glucose cotransporter is thought to initiate this series of events. The (reversible} enhancement of junctional permeation by glucose may have substantial physiological importance, because this altered permeability promotes paracellular absorption of nutrients by solvent drag (1). In agreement with this interpretation, structural analyses show that glucose elicits structural alterations in absorptive cell TJs (2,4), including the development of focal intrajunctional dilatations. However, there is at present no specific evidence that junctional dilatations provide a pathway for the enhanced junctional permeation of solutes ranging in size from creatinine to inulin. One approach to localize the site(s} of enhanced permeability elicited by activation of Na+-glucose cotransport is to use morphologically detectable tracers. A common approach is to use tracers with a protein component and peroxidatic activity; subsequent fixation with divalent aldehydes cross-links the tracer with adjacent tissue proteins, thus preserving the anatomical localization of the tracer (5-8). This study examined the role played by glucoseinduced TJ dilatations in the absorption or exclusion of an 11-amino-acid hemepeptide (MP-11; mol wt, 1900) (8) and horseradish peroxidase (HRP; mol wt, 40,000). It shows that junctional dilatations induced by glucose provide the sole detectable pathway for absorption of MP-11 and that HRP fails to penetrate the epithelium even in the presence of glucose. In addition, these observations have implications for paracellular peptide uptake during absorption, as will be discussed. Abbreviations used in this paper: HRP, horseradish peroxidase; TJ, tightjunction. ©1991 by the American Gastroenterological Association

0016-S085/91/$3.00

720 ATISOOKAND MADARA

Materials

and Methods

After an overnight fast, male Golden hamsters weighing 180-220 g were, anesthetized with pentobarbital sodium (Nembutal, 5 mE/100 g; Abbott, North Chicago, IL). The small intestine was rapidly removed, and 4-5-cm lengths of intestine were perfused with a recirculating perfusion system previously described in detail (1,2). This method was chosen for tracer experiments because reducing reservoir size substantially diminished the costs of using the MP-11 tracer, which otherwise would have been prohibitive. After a 15-minute equilibration period (2), a 30-minute experimental period began, after which the tissues were harvested for morphological analyses as described below. The electrical measurements were performed on mucosal sheets mounted in Ussing chambers as previously detailed (2). In both experimental systems (Ussing chambers and segments), the buffer solution on the serosal side consisted of 114 mmol/L NaC1, 5 mmol/L KC1, 1.65 mmol/L Na2HPO4, 0.3 mmol/L NaH2PO4, 25 mmol/L NaHCO3, 1.25 mmol/L CaC12, and 1.1 mmol/L Mg2SO~at pH 7.4, maintained at 37°C, and continuously gassed with 95% OJ5% CO2. The mucosal solution was identical except for the addition of flurocarbon emulsion to provide 6 vol% oxygen (Oxypherol-ET, 20% wt/vol; Alpha therapeutic, Los Angeles, CA). After a 15-minute equilibration period, either 20 mmoEL glucose, or 20 mmol/L fructose control (2), was added to both serosal and mucosal solutions.

Tracer Studies Horseradish peroxidase (type II; Sigma Chemical Co., St. Louis, MO), a 40,000-molecular-weight tracer, and MP-11 purified from heme (Sigma, 90% purity) were used as tracers at a concentration of 0.5%. The amino acid sequence of MP-11 (8) is Val-Glu(NH2)-Lys-Cys-Ala-Glu(NH2)-Cys-His-Thr-Val-Glu. Heme is covalently attached to MP-11 at the cysteine residues (7,8). Visualization of HRP was performed using a modification (9) of the method of Graham and Karnovsky (5). MP-11 was localized by a modification of the procedure of Feder (8). Specifically, after fixation (see below) samples were rinsed in 0.05 mol/L Tris buffer, pH 7.56, three times for 5 minutes each. Tissues were placed into 5% agar and, using a tissue chopper, sliced into 135-p.m sections. The slices were preincubated in Tris with 0.5 mg/mL diaminobenzidine and 0.1 mol/L imidazole for 30 minutes at 20°C and then again for 30 minutes in a 37°C shaking-water bath. Tissues were then incubated for I hour at 37°C in a similar Tris-diaminobenzidine-imidazole buffer to which H202 was added (final H202 concentration, 0.05%). Tissues were then processed for routine electron microscopy (see below). In general, HRP yielded more reaction product than MP-11.

Morphological Studies Fixation of all tissues was accomplished by adding glutaraldehyde to the perfusate or reservoir such that the Final concentration was 2% (2,4).; this preserves the altered

GASTROENTEROLOGYVol. 100, No. 3

resistance state elicited by activation of the Na+-glucose cotransporter (2,4). Processing for routine thin- and thicksection microscopy was performed as previously described (2,4,9). Quantitative analysis of the frequency of junctional dilatations and of junctional penetration by the tracer was carried out using thin sections of longitudinally sectioned villi as previously described (4). All sectioned TJs on the top half of villi were assessed. The observer was blinded to the experimental group from which the sections were obtained. Examination was performed on unstained tissues to highlight the tracers. Even with this approach, it was often difficult to be sure that minor increases in density below the TJ, in the intercellular space, were indeed reflective of tracer. Thus, although clear evidence of tracer crossing the TJ was found (see below), the frequency of this finding is less clear than the frequency of finding tracer in the TJ dilatations. Freeze-fracture images of junctions were obtained as previously described (2,4). Results

As s h o w n in Table 1, addition of 20 mmol/L m u c o s a l glucose elicited both the expected increases in transepithelial voltage and short-circuit current a n d a decrease in transepithelial resistance, as previously reported (2). As s h o w n in Figure 1, the glucoseelicited resistance change obtained from chamberm o u n t e d tissues (Table 1) was paralleled by the d e v e l o p m e n t of focal TJ dilatations. As assessed by freeze fracture, these dilatations represent foci in w h i c h the netlike strand structure of the junction becomes markedly distorted. As previously s h o w n , these changes in TJ structure also o c c u r in Ussing c h a m b e r - m o u n t e d mucosa, where they parallel the decline in resistance (2). We have previously s h o w n (2) that TJ dilatations are elicited by mucosal, not serosal, glucose and that dilatations are not elicited by L-glucose. Analyses of H R P - and M P - 1 1 - p e r f u s e d segments s h o w e d that after fructose exposure TJ dilatations did not o c c u r and that TJ between villous epithelial cells

Table 1. Effects of Glucose and Fructose on the Electrical Profile of Epithelia in Ussing Chambers

Glucose (20 mmolfl.,) Preaddition 30 min postaddition P

Fructose (20 mmol/L) Preadditlon 30 m i n postaddltion P

R

Isc

PD

(/'2 • cm 2)

(p.A/cm2)

(mY)

48 - 2 33 ± 3

101 _ 32 305 ± 75

4.9 ± 0.8 10.2 ± 0.8

<0.01

<0.001

<0.01

49 - 4 48 ± 3

95 ± 56 ±

NS

NS

38 22

4.5 ± 0.6 2.7 ± 0.7 NS

NOTE. Values are mean ± SE of 7-9 mucosal sheets. R, resistance; Isc, short-circuit current; PD, potential difference.

M a r c h 1991

EPITHELIAL P E R M E A T I O N BY PEPTIDE

721

~A+,+'+.,?:~v..:<'~a++~; ' .i:~ +, :,:;N~,,..'+.:;,,+::," ::~ ::, +.+

+~ Figure 1. T h i n - s e c t i o n (A) and freeze-fracture (B) electron micrographs s h o w i n g glucose-elicited dilatations in absorptive cell TJs.

!'~ "+.!:!':'.:"!::gY+: ~'>"g;!c~:,::

~+' '.~ > +- +

haa.~,i'.tL++i.....;1,,+V,:,l~,,:..+-+.~++o.¢,.,.~.e~.,.. :

A. T r a n s v e r s e section at the level of the tight j u n c t i o n s h o w s the close cell-cell a p p o s i t i o n to be distorted by dilatations (orrowheods) (original m a g n i f i c a t i o n x25,000).



!+-.g.~.

-

.

~'~ .

.

.

.

.

.

. ....

.,..

,

,++a

---+-

.:i~:'4~..~.+,~'~r'~:~"~'~

=:":" ~ : +" :+ '" :'"" ;""+""2:':

'~'.}

~Pi'

,.:+',

',

:a.'~ a

,:~+:++~"

'

.+m,

','~

-+

'"

,+',

L++"

,

-~';~'

~. +

'

+-

'. .......55,.~.~ '..'-." '

+ ,

'+,

.~' •

~:/.S.

,

+.,

,+

+~,e,



:

!

7~

,

++"

•,

~1-1~

.. ~..:.:,

+

.,

+

"

~a , .

,'+ • ' +* - .

+

.

.P.~q~.~

i~,'e+..~_:,.~,,~.,.a

~ . + .r-++: +,.

B. Dilatations a p p e a r as areas in w h i c h the n o r m a l netlike p a t t e r n of j u n c t i o n a l s t r a n d s b e c o m e s distorted. T h e f u n c t i o n a l significance of this specific physiologically ind u c e d c h a n g e in j u n c t i o n a l architecture w a s not p r e v i o u s l y k n o w n (original magnification = x50,000).

+++',"V"

~+,

p+.,+:

+ ~...'v .x .~+.-'..:.a~:;:~' '-':.: ~-,~" ..'+ ;..'"-.'.1~.'~<;:*.:'..~'+,:+.< ~; ,.:.'"~'~'" ,~ "?,.^i.:+',~+,~i:+r~ .... s ,",.a+,',-,'.~..'~,~:~.-:...:~'

,~

...

..~++ :+..:.. + + ,. . .,. .+. .+ + +,*,.;

+' +'~,~'+~ :.

.+,.

'"::: :+'::2.,.',::::,7' :

++.,++,.,,

..... •..... , + ,++++; ,.; ,+

; "~:++,'+~''L%~' +l~,;':a'~'.,++".+'+~.~".;' +' %~X'~'~:

~'

++~.

+, f!~

~::~"i"~,~:-/'+~+! -'

Q[

• .. ~

,-..+-~.' .~ ~

't

J"

+" +"

+

+:

; '+"

'

-~

,

+

.~.

-

:: i

;..

~'~,.~ ~ .~..-.

Figure 2. Unstained electron micrographs of absorptive cell TJs s h o w i n g localization of the tracers HRP (A,C) or MP-11 (B, D) in tissues unexposed (A, B) or e x p o s e d (C, D) to 20 mmol/L glucose. Tight junctions do not leak either tracer in glucoseu n e x p o s e d t i s s u e s or in g l u c o s e e x p o s e d tissues in areas w h e r e ~ilatations are not present (original magnification --- x 45,000).

722

ATISOOK AND MADARA

GASTROENTEROLOGY Vol. 100, No. 3

excluded both tracers (Figure 2A and B). After glucose exposure, both tracers were excluded from TJ in areas not containing dilatations (Figure 2C and D). In contrast, as shown in Figure 3A, several of the glucoseelicited TJ dilatations leaked MP-11. However, TJ did not leak HRP even at sites of dilatations (Figure 3B). Occasionally sites were found at which MP-11 had penetrated the most apical portion of the TJ but had not yet filled the dilatation (Figure 3C). These latter findings show that tracer in TJ can be visualized when present even if dilatations do not exist, thus validating our interpretation of negative results in glucoseunexposed TJ. MP-11 staining in the subjunctional paracellular space (Figure 3C) was less readily detected than the dense staining that occurred in the dilatations and was restricted to glucose-exposed tissues (Figure 4). Table 2 summarizes the quantitation of these morphological results. Discussion

The present results show that glucose-elicited dilatations in the TJ of small intestinal absorptive cells are specific anatomical sites at which junctional penetration of a heme-linked oligopeptide (MP-11) occurs. This tracer has the advantage of being crosslinked to other proteins during fixation, thus maintaining its physiological position postfixation. It is possible that MP-11 might be degraded into a slightly smaller peptide during its passage into the TJ. However, loss of more than 3 of the 11 amino acids results in loss of its enzymatic activity (7). The fact that the reaction product is found in TJ only after exposure to glucose also argues against degradation to a much smaller peptide. Our findings also suggest that paracel-

i:! .~ "i.

:" ~

'~'

,:i"~

Figure 4. Unstained electron micrograph showing example of focal paracellular MP-11 leak {arrows). Arrowhead marks the paracellular space at an interdigitation--a common feature in absorptive cells {original magnification = × 70,000).

~ "~' '~: .... !i"

Figure 3. Electron micrographs of TJs in glucose-exposed tissues. MP-11 (it) but not the macromolecule HRP (B) can leak into TJs at sites of dilatations. At sites where MP-11 leaks into TJs, obvious deeper paracellular penetration is often not present, although unequivocal paracellular MP-11 leaks were focally detected in glucose-exposed tissues {see Figure 4). Rarely was MP-11 seen penetrating the TJ between the lumen and dilatation.(C, arrowhead} but not yet entering the dilatation {original magnification = x45,000}.

March 1991

EPITHELIALPERMEATIONBY PEPTIDE 723

Table 2. Quantitation of MP-11 Leak at Sites of Glucose-Elicited Junctional Dilatations

Glucose-exposed Glucose*unexposed

TJ with dilatations (~)

Dilatations with MA-11 leak (%)

19° 2

11 0

NOTE. For purposes of quantitation, seven blocks containingvilli were examinedfrom four glucose-exposedanimals and eight from three non-glucose-exposed animals. A total of 191 (glucoseexposed) and 164 (glucose-unexposed)TJs were examined,respectively, for the data given. Data represent an unselected sample of the total animals (n = 23) and blocks (n = 69) examined.Resultsof examinationof these other tissues agreed with these findings,but were not performed in a blinded or quantitative fashion as was done here. °Thin sections of TJ contain <0.1 ~m of TJ. Becauseeach absorptive cell has -- 35 p.m of TJ, the above findings indicate that each absorptive cell has numerousTJ dilatationsaround its perimeter. lular uptake of peptides is regulated and should accompany glucose-induced paracellular solute uptake by solvent drag. The movement of inert solutes across the intestine is largely determined by solvent drag across TJs (1), and activation of glucose absorption stimulates such transjunctional water flow (1,2). It is known that mammalian small intestine absorbs small peptides (review, 10). However, the current view holds that oligopeptide absorption occurs transcellularly by a mechanism that includes carriermediated uptake of peptides across the apical membrane of absorptive cells, intracellular hydrolysis of peptides into flee amino acids, and release of free amino acids into the basolateral compartment, presumably by a diffusion or carrier pathway (review, 11). Evidence for peptide absorption by intestinal epithelial cellg originated, in part, from observations of intestinal amino acid absorption in patients having defective mucosal amino acid transport systems. For example, cystinuria and Hartnup disease are inherited conditions in which transport of different amino acids by the proximal renal tubule cells and the intestinal epithelium is markedly impaired (12). Cystinuria patients appear unable to concentrate lysine in intestinal mucosal cells because of a lack of the specific transporter for this molecule (13). Yet nutrition is essentially normal in these patients (12) despite such grossly defective cellular transport of the amino acids involved. The ability of the intestine to absorb oligopeptides via the above described transcellular route is thought to account for the unaffected nutritional status of these patients (11,12,14). It is likely that some of the lysine absorbed by such patients is taken up nonselectively as single amino acids by solvent drag (1) across TJs with enhanced permeability (1,2,4). In addition, the Observations

reported here suggest that uptake of intact oligopeptides might also occur via the paracellular route. This possibility is supported by other observations. For example, a recent study of calves (10) indicated that up to 70% of absorbed amino acids appearing in the plasma of portal blood are in the form of peptides, largely in the 3 00-1500-molecular-weight range, rather than in the form of amino acids as would have been predicted by the current view of transcellular oligopeptide absorption. Studies of rat small intestinal segments have similarly suggested that substantial transepithelial movement of intact oligopeptides occurs during active absorption, although this accounts for <30% of the total amino acid absorbed in these studies (15). Similarly, studies of diglyceride, triglyceride, and tetraglycine absorption by the rat small intestine show appearance of these peptides in portal blood (16). Thus, it is likely that paracellular and transcellular uptake of oligopeptides occurs, although the relative contributions of those two pathways is unclear. In future studies of oligopeptide absorption it might be useful to reexamine the characteristics of the putative transcellular carrier-mediated, pathway. For example, one key piece of evidence supporting a carrier-mediated pathway is the observed saturability of this pathway. However, saturability data are often obtained using solutions with 110-140 mmo]/L NaCl and varying in dipeptide concentration from 0-100 mmol/L (17). At the higher dipeptide concentrations, an effective osmotic pressure that would attenuate convective water flow in the mucosal-to-serosal direction would be present. Because such convective flow is the major determinant of inert solute uptake across TJ [via solvent drag (1)], it is possible these experiments may have primarily exerted effects on the paracellular rather than the transcellular pathway. In considering selective carrier-mediated transcellular uptake of oligopeptides, it is also somewhat difficult to envision how such carrier(s) could mediate uptake. For example, for tetrapeptides alone there are 160,000 potential variations (II). To efficiently absorb the variety of oligopeptides present during the digestion of a meal, such carriers would have to be extremely numerous or have a surprising ability to recognize this broad array of molecules in a "selective" fashion.

References 1. Pappenheimer JR, Reiss KZ. Contribution of solvent drag throughintercellularjunctionsto absorptionof nutrientsin the small intestineof the rat. J MembrBiol 1987;100:123-136. 2. Atisook K, Carson S, Madara JL. Effects of phlorizin and sodium on glucose-elicited alterations of cell junctionstin intestinalepithelia.AmJ Physio11990;258:C77-C85. 3. PappenheimerJR. Physiologicalregulation of transepithelial

724

4.

5.

6.

7.

8. 9.

10. 11.

ATISOOK AND MADARA

impedance in the intestinal mucosa of rat and hamsters. J M e m b r Bio11987;100:137-148. Madara JL, Pappenheimer JR. Structuralbasis for physiological regulation of paracellular pathways in intestinal epithelia. J M e m b r Bio11987;100:149-164. Graham RD, Karnovsky MJ. The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructuralchemistry by a new technique. J Histochem Cytochem 1966;14:291-300. Seligman AM, Karnovsky MJ, Wasserkrug HI,, Hanker JS. Non-droplet ultrastructuraldemonstration of cytochrome oxidase activitywith a polymerizing osmiophilic reagent diaminobenzidine. J Cell Biol 1968;38:1-14. K.raehenbuhl JP, Garlardy RE, Jamieson J. Preparation and characterization of an irnmunoelectron microscope tracerconsisting of a heme-octapeptide coupled to Fab. J Exp M e d 1974;139:208-223. Feder N. A heme-peptide as an ultrastructural tracer. J Histochem Cytochem 1970;18:911-913. Madara JL, Trie~ JS. Structure and permeability of goblet cell tightjunctions in rat small intestine.JM e m b r Bio11982;66:145157 Webb KE. Amino acid and peptide absorption from the gastrointestinaltract.Fed Proc 1986;45:2268-2271. Alpers D. Carbohydrate and Protein Absorption. In: Johnson LR, ed. Physiology of the gastrointestinaltract. 2nd ed. N e w York: Raven, 1986.

GASTROENTEROLOGY

Vol. 100, No. 3

12. Asatoor AM, Crouchman MR, Harrison AR, Light FW, Loughridge LE, Milne MD, Richards AJ. Intestinal absorption of oligopeptides in cystinuria.Clin Sci 1971;41:23-33. 13. Their SO, Segal S, Fox M, Blair A, Rosenberg LE. Cystinuria: defective intestinal transport of dibasic amino acids and cystine. J Clin Invest 1965;44:442--448. 14. HellierMD, PerrettD, Holdsworth CD. Dipeptide absorption in cystinuria.Br M e d J 1970;4:782-783. 15. Gardner MG. Absorption of intact peptides--studies on transport of protein digests and dipeptides across rat small intestine in vitro. QJ Exp Physiol 1982;67:629-637. 16. Matthews DM, Craft IL, Geddes DM, Wise IJ, Hyde CW. Absorption of glycine and glycine peptides from the small intestine of the rat. Clin Sci 1968;35:415-424. 17. Adibi SA, Soleimanpour MR. Functional characterization of dipeptide transport system in human jejunum. J Clin Invest 1974;53:1368-1374.

Received April 24, 1990. Accepted August 27, 1990. Address requests for reprints to: James L. Madara, M.D., Department of Pathology, Brigham and Women's Hospital, 20 Shattuck Street, Room 1423, Boston, Massachusetts 02115. Supported by National Institutes of Health grant DK-35932 and by a grant from the Siriraj Foundation, Thailand {K.A.}and a Prince Mahidol Scholarship, Thailand {K.A.). The authors thank Susan Carlson for expert technical assistance.