Gut, 1970, 11, 486-492

Morphological changes of the small-intestinalmucosa of guinea pig and hamster followingincubation in vitro and perfusion in vivo withunconjugated bile salts

THOMAS S. LOW-BEER1, ROBERTO E. SCHNEIDER, ANDWILLIAM 0. DOBBINSFrom the Gastroenterology Research Laboratory, Veterans Administration Hospital, and Departmentof Medicine, Division of Gastroenterology, Duke University Medical Center, Durham, North Carolina

SUMMARY Incubation in vitro of the intestine of the hamster and guinea pig with 5 mMsodium cholate and with 2 mM sodium deoxycholate and sodium chenodeoxycholate resultedin significant morphological changes compared with control incubations. Generally, no majordifferences were observed between proximal and distal small intestine or between the speciesused. Only when guinea pig intestine was incubated with 5 mM cholate was less damagefound proximally than distally.

Perfusion in vivo of the intestine of the hamster and guinea pig with Krebs-Ringer phosphateresults in separation of the epithelium from the lamina propria without excessive shedding ofepithelial cells from villous tips. This change was also seen in specimens taken before per-fusion and probably represents unavoidable trauma during handling of the intestine.

In contrast to studies in vitro, regional differences are readily demonstrable with perfusionof bile salts in vivo. Dihydroxy bile salts produce more marked alterations of both proximaland distal small intestine than the trihydroxy bile salt, sodium cholate. Dihydroxy bile saltsresult in significantly greater alterations in proximal than in distal mucosa.When 5 mM cholate at pH 6.8 is perfused in the guinea pig, absorption occurs approxi-

mately 30 times more rapidly from distal than from proximal segments, while in proximalsegments 2 mM chenodeoxycholate is absorbed approximately 15 times more rapidly than5 mM cholate. A correlation is suggested between the morphological alteration produced inthe region of the small intestine by a bile acid and the amount of bile salt passing through thecell. Furthermore, it is proposed that the ileal cells may be damaged to a lesser extent by bileacids normally found in that particular species.

Although unconjugated bile salts may be foundin the small intestine under normal physiologicalconditions, conjugated bile salts predominate(Norman and Sjovall, 1958; Dietschy, Salomon,and Siperstein, 1966; Kim, Spritz, Blum, Terz,and Sherlock, 1966). In contrast, a predominanceof unconjugated bile salts has frequently beendemonstrated in the small intestine of bothhumans and animals where an abnormal bacterialflora exists (Kim et al, 1966; Donaldson, 1965;Tabaqchali and Booth, 1966; Rosenberg, Hardi-Received for publication 8 December 1969.'Present address: Department of Medicine, Bristol Royal Infirm-ary, BS2 8HW.

son, and Bull, 1967). Unconjugated bile saltshave been shown to inhibit several intestinaltransport mechanisms (Pope, Parkinson, andOlson, 1966; Forth, Rummel, and Glasner, 1966),to promote epithelial cell injury (Dawson andIsselbacher, 1960; Fry and Staffeldt, 1964), toaffect the function of a variety of cell fractions(Palade and Siekevitz, 1956; Whitehouse andStaple, 1959; Weissman, 1964), and to inhibitnon-specifically intestinal triglyceride synthesis,cholesterogenesis, and carbon dioxide production(Dietschy, 1967). The experiments reported herewere designed first to study morphologicallythe relative effects of unconjugated bile salts on

on 15 July 2018 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.11.6.486 on 1 June 1970. Dow

nloaded from

Morphological changes of the small-intestinal mucosa oJ guinea pig and hamster

the small intestine following perfusion in vivoand in vitro; second, to determine whether thedistal small intestine (the site of active transportof bile salts) has a different sensitivity to thesematerials than does the proximal small intestine;and third, to investigate whether morphologicalchanges are more related to the quantity of bilesalt being absorbed than to the concentrationwithin the lumen, especially in the proximal smallintestine.

Material

The species used in these experiments weregolden hamsters (90 to 170 g) and adult albinoguinea pigs of the Hartley strain (200 to 650 g).Both species have gallbladders but each has adifferent bile salt excretion pattern. Conjugatedchenodeoxycholate is the primary bile salt of thestrain of guinea pigs used (Haslewood, 1964);in the hamster conjugates of cholate predominate,with smaller amounts of chenodeoxycholate anddeoxycholate (Prange, Christensen, and Dam,1962), data further confirmed in our laboratory.The animals were fed Purina hamster laboratorychow and Purina guinea pig chow before study.The unconjugated bile salts used in these experi-ments were obtained either from MaybridgeChemical Co, Cornwall, UK, or from Dr LeonLack, Department of Physiology and Pharmaco-logy, Duke University Medical Center, USA,and 24-14C-cholic acid and 24-14C-chenodeoxy-cholic acid from Mallinckrodt Nuclear, Orlando,Florida, USA. The purity of bile salts used inthese experiments was ascertained by thin-layerchromatography and'assay of the melting point.When necessary, purification and recrystalliza-tion were performed as described by Norman(1955). A final purity of over 98% was achieved.

Methods

INCUBATION in vitroThe unconjugated bile salts were prepared bydissolving the purified material in distilled waterat a temperature of 450 to 50°C. Neutralizationwas accomplished with 1 N NaOH. NaCI, KCI,and Na2HPO4 were added as required to form acalcium- and magnesium-free Krebs-Ringer phos-phate solution. The solution was titrated topH 6-7 with 1 N HCI. In selected experiments,calcium chloride and magnesium sulphate wereadded in physiological amounts to the KRPsolution which contained no bile salts. All solu-tions were clear and isotonic.The animals were fasted for 18 to 24 hours

before sacrifice. Guinea pigs were killed by anoverdose of ether and the hamsters by cervicaldislocation. The abdomen was opened quickly

and the small intestine was removed as rapidlyas possible, taking care to leave the mesenteryintact in order to preserve the serosal surface.The small bowel was placed in cold (20 to 3°C)physiological saline and the proximal and distalportions were identified. Between 10 and 12 cmof each portion was sectioned, rinsed twice withphysiological saline, and everted using a glassrod. The everted small bowel was sectioned insegments of 0.5 to 0.7 cm; one segment eachfrom the proximal and distal small bowel wasfixed immediately in Bouin's solution as control.Four ml aliquots of unconjugated bile salt(sodium cholate, sodium deoxycholate, andsodium chenodeoxycholate 1.0, 2.5, and 5.0 mMrespectively in Krebs-Ringer solution) werepipetted into 25 ml Erlenmeyer flasks before thesacrifice of the animal. An additional flask con-tained only Krebs-Ringer solution and was usedas an incubation control. A segment of evertedsmall bowel was then placed in each flask andincubated in a water bath shaker at 37°C under95 % oxygen for one hour. The shaker water bathhad a speed of 20 to 22 movements per minute.Ten segnments of everted proximal and distalsmall bowel were incubated in each experiment.There was no change in thepH of the incubatingsolutions during these experiments.

PERFUSION in vivoThe preparation for these experimnents has beendetailed previously (Heaton and Lack, 1968;Low-Beer, Tyor, and Lack, 1969). A Sigmamotor pump calibrated to give the same rate offlow of 4.5 ml per minute for two channels wasused, and the infusate maintained at a tempera-ture of 37°C. Unconjugated bile salts were dis-solved in physiological saline and buffered with0.01 N sodium phosphate adjusted to pH 6-8.The final solution was isotonic. The concentra-tions and flow rates were such that the effluentconcentration of bile salt was not appreciablyaltered, hence all parts of the perfused' intestinewere exposed to the same concentration (Heatonand Lack, 1968). The pH was also unchanged atthe end of the perfusion, and the total volumeof effluent recovered from each intestinal seg-ment during the one-hour perfusion was similarin all experiments. When absorption of bile saltwas measured, approximately 5 ,uc of the appro-priate 24-14C isotopic bile salt of high specificactivity was added to each litre of perfusate forsubsequent estimation in the bile. The animalswere anaesthetized, the abdomen was opened, andproximal and distal small bowel were identified.Isolated loops were prepared in both proximaland distal intestine; these measured 15 to 38 cmin the guinea pig and 9 to 10 cm in the hamster.In the hamster the inlet catheter of the proximalloop was inserted 3 cm below the entry of thebile duct into the duodenum, and in the guineapig the inlet catheter of the proximal loop was

487 on 15 July 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.6.486 on 1 June 1970. D

ownloaded from

Thomas S. Low-Beer, Roberto E. Schneider, and William 0. Dobbins

placed at the ligament of Treitz. A segment of thesmall bowel proximal to the inlet catheter of theproximal loop was resected and fixed in Bouin'ssolution as a control. An outlet catheter with alumen diameter double that of the inlet catheterwas placed appropriately. A similar isolated loopwas constructed in the distal bowel, placing theoutlet catheter as near the ileocaecal valve aspossible. A control segment for histology wastaken between the ileocxcal valve and the outletcatheter. The bowel was then replaced in theabdominal cavity.

Bile salt absorption was measured in 10 guineapigs. In these experiments bile was collected fromthe cannulated common bile duct, and only oneloop of intestine was perfused in each animal.The length of proximal intestine perfused aver-aged 26 cm and of distal intestine 32 cm. Themethods of analysis have been described in detailelsewhere (Heaton and Lack, 1968; Low-Beeret al, 1969). In outline, the methods used were asfollows: bile was collected in calibrated centrifugetubes for six 10-minute periods and the volumeswere noted. The bile was then made up to a stan-dard volume with water and aliquots were measur-edfor radioactivity in a liquid scintillationcounter.The bile was sufficiently dilute to avoid complica-tions due to quenching. The amount of bile saltabsorbed was calculated by relating the radio-activity recovered in the bile to the specific activityof the perfusion fluid.At the end of the one-hour perfusion period,

the abdominal cavity was opened, the loopswere carefully resected, placed in cold (20 to 3°C)physiological saline, and measured. Intestinalsegments were then taken from a site 2 cm fromthe inlet catheter and also from the middle ofeach loop and fixed immediately in Bouin'ssolution.The segments fixed in Bouin's solution were de-

hydrated and embedded in paraffin, serially sec-tioned at 5 g thickness, and stained with haema-toxylin and eosin and with periodic acid Schiff andhaematoxylin. Two hundred and fifty or moresections were prepared from each segment, andexamined by light microscopy.

Results

Five hundred and thirty-eight biopsies werestudied by two of the present authors withoutknowledge of the species examined, the bile saltused, or the particular procedure. It was agreedthat morphological changes observed should beclassified on the following scale: 0 should indicateno change, + a nonspecific change (see Fig. 1 aand b). Abnormal changes were graded asfollows: + for a mild shedding of the cells (Fig.2a and b), + + for a moderate shedding of thecells, and + ++ for a severe shedding of the cells(Fig. 3a and b). Agreement between the observers

Fig. la.

w:.

Fig. lb.

Figs. 1, 2 and 3 Light micrographs of Bouin-fixed,paraffin-embedded sections ofproximal intestinalmucosa in the hamster following perfusions in vivo.The histological changes after incubation in vitroand perfusion in vivo were qualitatively identical.Sections were stained with the PAS reaction andhaematoxylin. The low magnifications are x 190and the high magnifications x 750.

Fig. Ia and b Control segment illustrating non-specific change (+) of separation of epithelium fromthe lamina propria.

working independently was obtained in 98%of the biopsies. In the remaining 2%, the dis-agreement was marginal, ie, whether + or ±, or+ or + +, and was resolved by discussion. A fewexperiments were repeated in which controlbiopsies showed extensive lymphocytic infiltra-tion of the mucosa. These changes were presum-ably due to illness of the animal.

488 on 15 July 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.6.486 on 1 June 1970. D

ownloaded from

Morphological changes of the small-intestinal mucosa ofguinea pig and hamster

Fig. 2a. Fig. 3a.

Fig. 2b. Fig. 3b.Fig. 2a and b Segment illustrating mild abnormality Fig. 3a and b Segment illustrating severe(+) of separation of epithelium from lamina propria abnormality (++ +) with marked shedding ofwith slight shedding of epithelial cells from villous epithelial cells from villous tips in addition totips. separation of epithelium from lamina propria.

The lesions noted were located exclusively inthe upper third of the villi. When Dresent, theywere similar following incubation in vitro andperfusion in vivo. In both situations, there wasexcessive shedding of epithelial cells from villoustips (Figs. 1-3). The + + and + + + changecould be detected grossly with the dissectingmicroscope. The desquamated cells appearedsimilar in control experiments and in those wherebile salt was used. The only difference was in thedegree of desquamation. No changes wereobserved in the lamina propria, muscularismucosa, or submucosa. The desquamated cellswere rounded in shape and there was increasedeosinophilia of the cytoplasm. The nuclei were

pyknotic and showed prominent clumping ofnuclear chromatin about their periphery, changescharacteristic of cell injury. Other cells appearedswollen and contained disrupted cytoplasmicmembranes, and these were present less often incontrol experiments.

INCUBATION in vitroSeven incubations were performed using thesmall intestine of four hamsters and three guineapigs. A summary of the changes observed isshown in Tables I and II. Nonspecific changes(±) and minimal damage (+) were observed inthe specimens incubated only with Krebs-Ringer

A89 on 15 July 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.6.486 on 1 June 1970. D

ownloaded from

Thomas S. Low-Beer, Roberto E. Schneider, and William 0. Dobbins

solution. Control un-incubated specimens werenormal or showed nonspecific changes only.Morphological changes in villous structure weresimilar in proximal and distal small bowel in thehamster, and in the guinea pig studied withsodium deoxycholate and chenodeoxycholate.Sodium cholate in guinea pigs produced changesnot significantly different from the incubatedcontrols, except when distal small intestine wasincubated at a concentration of 5 mM. In the latterexperiment, moderate damage was produced which

Bile Salt Concentration (mM) Three Guinea Pigs Four Hamsters

Proximal Distal Proximal Distal

Na cholate i + 0 to ±2S + + + +50 + +A+ +A+ +A+

Na deoxycholate 1-0 + + + +2.5 +±+ +-+ +A+A++-+5.0 +A+-+ +A+A++-+-+ +A+A+Na chenodeoxycholate 1-0 ± to + ± + +25 +A+ +A+ +A+ +A+5.0 +A+-+ ++ +A+-+ +A+A+

Table I Histological damage after incubation in vitrousing unconjugated bile salts

Krebs-Ringer Phosphate Solution Five Guinea Pigs Five Hamsters

Proximal Distal Proximal Distal

Alone ± to + ± + ±Ca added ±to + ±to + ± to + to +Mg added ± to + ± to + + +Mg and Ca added + + + +

Table II Histological damage after incubation in vitrousing Krebs-Ringer solution as control

Bile Salt Concen- Guinea Pig Hamstertration(mM) Number Proximal Distal Number Proximal Distal

Control perfusion 2 0 to + 0 to ± 2 0 to 0 to iNa cholate 5 3 0 toAi + + 2 0 +Na deoxycholate 1 2 + + 0 2 +A +Na chenodeoxy-

cholate 1 2 + 0 to A 2 + + +Na deoxycholate 2 2 + + + + + + 2 + +A- +Na chenodeoxy-

cholate 2 3 + + 0 to + 2 +A- i-+

Table III Histological damage after perfusion in vivousing unconjugated bile salts

Bile Salt Concen- Recoveryfrom Bile Fistula during 60-Min Perfusiontration (mM)(mM)

Proximal Small Intestine Distal Small Intestine

Mean Range Exps. Mean Range Exps.

Na cholate 5 1-2 0-7-2-1 3 35 32-38 3Na chenodeoxycholate 2 18 18 2 28 25-31 2

Table IV Absorption ofunconjugated bile salts duringperfusion in vivo ofthe small intestine in the guinea pig

was greater than in proximal small intestine or inthe controls.

Five experiments were performed to evaluatewhether additions of physiological amounts ofCa and/or Mg to the Krebs-Ringer solutionwould prevent the changes observed in controlincubations. The additions of Ca and/or Mg didnot prevent the changes described (Table II).

PERFUSIONS in vivoControl studies were performed in two animals ofeach species, using physiological saline only andphysiological saline to which was added twicethe amount of IN NaOH usually required in thepreparation of the bile salt solution. All speci-mens obtained showed no change or nonspecificchanges. This is in marked contrast to the findingsin vitro.

Paired animals from each species were perfusedwith 1 mM and 2 mM Na deoxycholate andchenodeoxycholate and with 5 mM Na cholate.The histological findings are summarized inTable Ifl. Sodium cholate (5 mM) producedsignificant alterations in the distal small intestineof both species, although there were no asso-ciated changes in the proximal segment. Sodiumdeoxycholate (1 mM) produced changes in theproximal small bowel of both species, and definitealterations were observed in the hamsters' distalsmall bowel, whereas that of the guinea pig wasunaffected. Sodium deoxycholate (2 mM) resultedin significant injury to the proximal and distalsmall intestine of both species. The degree ofinjury was more pronounced in the proximalsegment of the hamster but was similar in bothsegments of the guinea pig. Sodium cheno-deoxycholate (1 and 2 mM) produced alterationsin the proximal small intestine of both specieswhich generally appeared greater than the distalchanges.When absorption of a perfused bile salt was

measured by recovery from a bile fistula (TableIV), perfusion with 5 mM sodium cholate forone hour led to absorption of 1-2 uM fromproximal segments and 35 ,M from distalsegments. When 2 mM sodium chenodeoxy-cholate was perfused, absorption from proximalsegments was 18 ,M and from distal segments28,uM.

Discussion

In these experiments we have attempted todevelop a preparation for evaluating the acuteeffect of individual bile salts on intestinal mucosalmorphology. We have compared the results ofincubations in vitro with an in-vivo perfusionsystem. During perfusion in vivo we have alsomeasured bile salt absorption in order to try toassess whether penetration through the mucosa

490 on 15 July 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.6.486 on 1 June 1970. D

ownloaded from

Morphological changes of the small-intestinal mucosa ofguinea pig and hamster

may be a factor in causing histological changes.Unconjugated bile salts were used because of therelative ease with which they pass through theintestinal wall by means other than active trans-port; however, they occur in the upper smallintestine to a significant extent only in caseswhere an abnormal bacterial flora exists.

Incubation in vitro tended to obscure moresubtle changes, since small intestinal tissue in-cubated in Krebs-Ringer solution in the absenceof bile salts invariably showed separation ofepithelium from the lamina propria and oftenmild shedding of cells from villous tips. Inclusionof physiological amounts of divalent cations intothe incubating media neither enhanced nor less-ened these effects. Our data extend the morpho-logical findings in vitro of Dawson and Issel-bacher (1960) in the rat. These authors showed'dissolution of the tissue with loss of whole villi'at the end of a one-hour incubation of intestinalrings from the upper two-thirds of the smallintestine in the presence of 1-5 mM deoxycholate,a finding confirmed by Donaldson (1965) at0-6 mM concentration. Dawson and Isselbacherfurther showed that cholate at 15 mM concentra-tion produced 'histological damage to all villi'when compared with incubations in Krebs-Ringer solution containing 1 mM glucose, whichproduced only 'scattered mild mucosal damage'.An appearance identical to the control was foundwith glycocholate and taurocholate at 20 mM(Dawson and Isselbacher, 1960). Our studiesin vitro further show that histological changesgreater than those in control experiments areproduced by sodium chenodeoxycholate anddeoxycholate at 2-5 mM and 5 mM and that thesechanges are similar in degree in gut segmentstaken from proximal and distal bowel. Only with5 mM cholate incubation of guinea pig intestinewas significantly less damage found proximallythan distally. This is interesting in view of thenegligible absorption in vivo of 5 mM cholatefrom guinea pig proximal small intestine com-pared with 2 mM chenodeoxycholate which isbetter absorbed proximally. This suggests thepossibility that histological changes in this regionmay be related in part to the degree to whichbile salts are able to traverse the bowel wall.

Regional differences were readily demonstrablewith perfusion in vivo of bile salts, and the degreeof damage with similar bile salt concentrationtended to be less severe than that observed afterincubation in vitro. This may be attributed to thepresence of an intact vascular system withperfusion in vivo, and possibly to lessened traumawhen tissues are not everted and shaken mech-anically.The perfusion studies in vivo indicate that

dihydroxy bile salts produce more markedalterations of proximal small intestine than doesthe trihydroxy bile salt, sodium cholate. Cholateconcentrations of 5 mM failed to producemucosal alterations proximally, the changes

again being confined to the distal small intestine.When absorption of sodium cholate was mea-sured in these experiments, the amount of bilesalt absorbed from the distal small intestine wasapproximately thirty times that absorbed fromthe proximal small intestine during the one hourperfusion. In contrast, mild to severe changes inmorphology of both proximal and distal smallintestine were readily apparent with the dihydroxybile salts, deoxycholate and chenodeoxycholate,at only 2 mM concentrations in both animals,and also at 1 mM in the hamster. The changesinduced by dihydroxy bile salts tended to bemore severe in the proximal than in the distalsmall intestine.At least two factors must be considered in

evaluating these observations. First, the pene-tration of bile salts into the cell, and second, thechemical toxicity of the bile salt. In the proximalsmall intestine intracellular penetration occursonly by passive processes (Weiner and Lack,1962). The more polar trihydroxy bile salt wouldbe expected to have a lesser degree of penetrancethan the dihydroxy bile salts. We have shownthat 2 mM sodium chenodeoxycholate is absorbedfrom the proximal small intestine of the guineapig fifteen times more rapidly than 5 mMsodium cholate. The lesser proximal absorptionof cholate may explain in part its lack of effecton the proximal small bowel in the experimentsin vivo.

In the ileum absorption occurs not only bydiffusion, but also by the specialized ileal trans-port system which acts on both conjugated andunconjugated bile salts (Lack and Weiner, 1961;Weiner and Lack, 1962). Here, similar degreesof morphological damage were evident withsodium cholate at 5 mM and with sodiumchenodeoxycholate and deoxycholate at 2 mMconcentration, suggesting a greater chemicaltoxicity of the dihydroxy bile salts. In contrast,conjugated bile salts have been shown to result inminimal morphological changes and to havelittle toxicity as judged by action on metabolismand transport of water-soluble nutrients by ratjejunum in vitro (Pope et al, 1966; Dietschy,1967; Holt, 1966).The fact that unconjugated dihydroxy bile

salts lead to significantly greater alterations inmorphology in proximal than in distal smallintestine can hardly be explained by greatermucosal penetration proximally, since thepotential for cellular entry is greater distallywhere two absorptive processes operate. Indeed,our measurements show greater absorptiondistally, even when corrections are made for theslight differences in length of intestine perfused.Furthermore, radioautographic studies(Schneider, 1967, unpublished data) of gutsamples taken during such perfusion suggest thatconcentrations of bile salts in ileal segments areat least five times those in more proximal seg-ments of small intestine. Rather, it would seem

491 on 15 July 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.6.486 on 1 June 1970. D

ownloaded from

492 Thomas S. Low-Beer, Roberto E. Schneider, and William 0. Dobbins

reasonable to speculate on a possible adaptationof ileal epithelial cells, which, during normalenterohepatic circulation, areexposed continuallyto relatively large quantities of bile salts. Thisadaptation may be species specific. In the guineapig, deoxycholate is present in very low concen-trations, if at all, whereas it is a normal con-stituent of hamster bile. An increase in sodiumdeoxycholate concentration from 1 mM to 2 mMled to a much greater increase in damage to theguinea pig ileum than to the hamster ileum. Theobservation that 2 mM sodium chenodeoxy-cholate produces considerably less change inthe ileum than in the proximal small intestine ofthe guinea pig is in accord with this hypothesis,since chenodeoxycholate is the principal bilesalt native to the guinea pig. The bile salts usedin this study are all native to the hamster. Inkeeping with the above thesis of species specificresistance, morphological alterations with allthese bile salts, at all concentrations tested except2 mM chenodeoxycholate, were more apparentin the proximal than in the distal small intestineof the hamster. It is possible that resistance to theaction of unconjugated bile salts may also beacquired. Fry and Staffeldt (1964) showed thatin mice fed 2% deoxycholate, the upper intestineshowed changes within two days of institutionof the diet. Their illustration closely resemblesthe 1+ abnormality shown in our report. Ateight days the damage extended down the wholeintestine. Thirty-nine days after starting the diet,however, the appearance of the villi was almostnormal except for evidence of increased prolifera-tion in the crypts (Fry and Staffeldt, 1964).The markedly greater absorption of sodium

chenodeoxycholate than sodium cholate fromthe proximal small intestine at pH 6X8 has notbeen demonstrated previously by direct measure-ment of bile salt in the bile. Recent preliminarydata of Hislop, Hoffman, and Schoenfield (1966),using a triple-lumen tube to perfuse isotopicbile acids and sample them in the human jejunum,are in broad agreement with this finding. Observa-tions in the guinea pig (Weiner and Lack, 1962)have shown negligible absorption of sodiumcholate buffered to pH 7.0 when introduced intoan isolated proximal segment of small intestinein vivo. Tidball (1964) found 43% recovery ofcholic acid in three hours from rat jejunum, andmentions that studies in the guinea pig weresimilar. The pH of the bile salt introduced intothe segment is not stated. Relatively slightchanges in thepH of the mucosal solution in therange of pH 6.0-8.0 result in marked changes inthe relative amounts of cholate ion and unionisedcholic acid and, therefore, in the degree of passivenon-ionic diffusion (Dietschy et al, 1966).There are no published data other than those

reported here on the absorption of chenodeoxy-cholate from the proximal small bowel of experi-mental animals. It would seem likely that thegreater proximal absorption of sodium cheno-

deoxycholate at a comparable pH in theseexperiments may be related to a greater fatsolubility of the undissociated molecule.

We are particularly grateful to Dr M. P. Tyor and toDr Leon Lack for continuous advice and assistance.This work was supported by Research grants AM11730 and AM 12107 from the United States PublicHealth Service, and that of Dr Schneider by train-ing grant AM 5093 from the United States PublicHealth Service.

Please address requests for reprints to William0. Dobbins, III, MD, Veterans AdministrationHospital, 50 Irving Street, N.W., Washington, DC20422.References

Dawson, A. M., and Isselbacher, K. J. (1960). Studies on lipidmetabolism in the small intestine with observations on therole of bile salts. J. clin. Invest., 39, 730-740.

Dietschy, J. M. (1967). Effects of bile salts on intermediary meta-bolism of the intestinal mucosa. Fed. Proc., 26, 1589-1598.

Dietschy, J. M., Salomon, H. S., and Siperstein, M. D. (1966).Bile acid metabolism, I. Studies on the mechanism ofintestinal transport. J. clin. Invest., 15, 832-846.

Donaldson, R. M., Jr. (1965). Studies on the pathogenesis ofsteatorrhea in the blind loop syndrome. J. clin. Invest.,44, 1815-1825.

Forth, W., Rummel, W., and Glasner, H. (1966). Zur resorptions-hemmenden Wirkung von Gallensauren. Naunyn-Schmied-berg's Arch. exp. Path. Pharmak., 354, 364-380.

Fry, R. J., and Staffeldt, E. (1964). Effect of a diet containingsodium deoxycholate on the intestinal mucosa of themouse. Nature (Lond.), 203, 1396-1398.

Hasiewood, G. A. D. (1964). The biological significance ofchemical differences in bile salts. Biol. Rev., 39, 537-574.

Heaton, K. W., and Lack, L. (1968). Ileal bile salt transport:Mutual inhibition in an in vivo system. Amer. J. Physiol.,214, 585-590.

Hislop, I. G., Hoffman, A. F., and Schoenfield, L. G. (1966).Determinants of the rate and site of bile acid absorption inman. J. clin. Invest., 46 (abstr.), 1070-1071.

Holt, P. R. (1966). Competitive inhibition of intestinal bile saltabsorption in the rat. Amer. J. Physiol., 210, 635-639.

Kim, Y. S., Spritz, N., Blum, M., Terz, J., and Sherlock, P. (1966).The role of altered bile acid metabolism in the steatorrheaof experimental blind loops. J. clin. Invest., 45, 956-962.

Lack L., and Weiner, I. M. (1961). In vitro absorption of bilesalts by small intestine of rats and guinea pigs. Amer. J.Physiol., 200, 313-317.

Low-Beer, T. S., Tyor, M. P., and Lack, L. (1969). The effects ofsulfation of taurolithocholic and glycolithocholic acids ontheir intestinal transport. Gastroenterology, 56, 721-726.

Norman, A. (1956). Preparation of conjugated bile acids usingmixed carboxylic acid anhydrides. Bile acids and steroids.Arkiv. Kemi., 8, 331-334.

Norman, A., and Sjovall, J. (1958). On the transformation andenterohepatic circulation of cholic acid in the rat. J. biol.Chem., 233, 872-885.

Palade, G. E., and Siekevitz, P. (1956). Liver microsomes: Anintegrated morphological and biochemical study. J. Bio-phys. Biochem. Cytol. 2, 171-200.

Pope, J. L., Parkinson, T. M., and Olson, S. A. (1966). Action ofbile salts on the metabolism and transport of water-solublenutrients by perfused rat jejun um in vitro. Biochim. biophysActa (Amst.), 130, 218-232.

Prange, I., Christensen, F., and Dam, H. (1962). Alimentaryproduction of gallstones in hamsters. II. Relation betweendiet and composition of the bladder bile. Z. Erndhrungsw.,3, 59-78.

Rosenberg, I. H., Hardison, W. G., and Bull, D. M. (1967).Abnormal bile salt patterns and intestinal bacterial over-growth associated with malabsorption. New Eng. J. Med.,276, 1391-1397.

Tabaqchali, S., and Booth, C. C. (1966). Jejunal bacteriology andbile salt metabolism in patients with intestinal malabsorp-tion. Lancet, 2, 12-15.

Tidball, C. S. (1964). Intestinal and hepatic transport of cholateand organic dyes. Amer. J. Physiol., 206, 239-242.

Weiner, I. M., and Lack, L. (1962). Absorption of bile salts fromthe small intestine in vivo. Amer. J. Physiol., 202, 155-157.

Weissman, G. (1964). Lysosomes. Blood, 24, 594-606.Whitehouse, M. W., and E. Staple (1959). Regulation ofcholesterol

oxidation by the liver in vitro. Proc. Soc. exper. Biol., 101,439-441.

on 15 July 2018 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.11.6.486 on 1 June 1970. Dow

nloaded from