STRUCTURAL AND HORMONAL ALTERATIONS IN THE GASTROINTESTINAL TRACT OF PARENTERALLY FED RATS

Transcription

1 GASTRONTROLOGY 68: , 1975 Copyright 1975 by The Williams & Wilkins Co. Vol. 68, No.5, Part 1 Printed in U.S.A. STRUCTURAL AND HORMONAL ALTRATIONS IN TH GASTROINTSTINAL TRACT OF PARNTRALLY FD RATS LONARD R. JoHNSON, PH.D., DWARD M. CoPLAND, M.D., STANLY J. DuDRICK, M.D., LNARD M. LICHTNBRGR, PH.D., AND GILBRT A. CASTRO, PH.D. Programs in Physiology and Surgery, University of Texas Medial Shool, Houston, Texas, and the Veterans Administration Wadsworth Hospital Center, Los Angeles, California This study examines the effet of prolonged absene of oral food intake on strutural parameters of the gastrointestinal trat in rats maintained nutritionally by intravenous feeding for up to 3 weeks. During this time, their body weights inreased by 22%. Controls fed a nearly isoalori oral diet were sham operated and harnessed in the same manner as their parenterally fed ounterparts. Parenteral feeding resulted in a signifiant derease in the weights (per 1 g body weight) of the oxynti gland area of the stomah, small intestine, and panreas. The weights of the spleen, testes, kidneys, and antral region of the stomah were unaltered. In the small intestine there was a signifiant loss of DNA and a near doubling of the RNA:DNA ratio in the parenterally fed animals. In the absene of an oral diet antral gastrin levels dereased to one-thirtieth of the ontrol level. The following onlusions are suggested by these results. First, the oral intake and/or physial presene of food within the gastrointestinal trat are neessary for strutural maintenane of some tissues of that trat. Seond, the disproportionate derease in weight that ours in ertain tissues is apparently unrelated to the absene of nutrients whih might normally be utilized diretly from the lumen. Third, maintenane of normal tissue stores of the hormone, gastrin, is dependent on stimuli provided by oral ingestion and the presene of food in the gastrointestinal trat. We know little about the importane of the regularly ourring presene of food in the gastrointestinal trat to the over-all maintenane of the strutural and fun- Reeived July 29, Aepted Otober 15, This study was supported by National Institutes of Health Researh Grants AM-1655, Al-11361, and CA Dr. Johnson is the reipient of National Institutes of Health Researh Career Development Award AM The authors are grateful to Paul D. Guthrie, Researh Assoiate, Program in Physiology; Shirley Ann Roy, Researh Tehniian, Program in Physiology; Pattie Watson, Nutritionist, Program in Surgery; and Robert G. Wilfong, Researh Tehniian, Program in Surgery, for expert tehnial assistane. The authors are also grateful to Dr. Travis Solomon for his valuable assistane tional integrity of the gut. As basi as this problem is to fields suh as physiology and nutrition, it remains largely uninvestigated. Previous attempts to study this problem have examined animals in whih all forms of nutrients had been withheld for various lengths of time. Dramati hanges were noted in the funtion and struture of the gastrointestinal trat, but it was impossible to separate the effets aused by the absene of food from the effets of starvation. These studies in starved animals have demonstrated, however, that the gastrointestinal trat is affeted by the absene of oral food intake to a greater degree than is the rest of the body. Steiner et al. 1 found a 32% redution of body weight in rats fasted

2 1178 JOHNSON T AL. Vol. 68, No. 5, Part 1 for 6 days. The small intestinal mass of the same animals, however, dereased 53%. At the same time there were signifiant redutions in RNA, DNA, protein, and water ontent. RNA dereased muh more than DNA, resulting in a signifiant derease in the RNA:DNA ratio. Short periods of food deprivation in rodents are known to ause hanges in muosal arhiteture, as both villus and rypt height diminish. 2 3 Several investigators have reported that the ell migration rate from rypt to villus tip is retarded during starvation, 2-4 indiating that the ell renewal mehanism is affeted by starvation. Confliting results on the effets of food deprivation on intestinal disaharidase ativity further ompliate the understanding of the role of food in the gut. 2, s, s A ommon explanation for the disproportionate hanges in small bowel struture during starvation has been that the epithelial ells are probably deprived of essential nutrients that are normally aquired from the lumen. Other suggestions have not been forthoming. The urrent study examines strutural and hormonal hanges whih our in the gastrointestinal trat of animals nourished totally by parenteral feeding. The hanges noted are assumed to be due only to the absene of the oral ingestion of nutrients, the lak of nervous and hormonal reflexes initiated by the physial and hemial stimuli of those nutrients, and the absene of diret exposure of nutrients to the gastrointestinal trat. Methods Animals and tissue preparation. Six male Sprague-Dawley rats with a mean body weight of 154 g were prepared for intravenous alimentation by surgially implanting atheters in the jugular vein. The animals were fasted for 24 hr before parenteral feeding was begun. ah animal reeived from 35 to 45 ml of fluid per day. Average fluid intake was 42 ml amounting to 49 kal. Daily examination of urine revealed a positive nitrogen balane in all animals. The urine was free of gluose and did not ontain abnormal amounts of aetone. The detailed methods for intravenous hyperalimentation of rats have been desribed by Steiger et al. 7 Six ontrol animals were sham operated and plaed in restraint harnesses idential to those used to support the infusion system in the parenterally fed animals. Controls were fasted for 24 hr, plaed in individual ages, and allowed an oral ad libitum diet of Purina rat how and water. The amount of food ingested by eah ontrol rat was alulated eah day by subtrating the weight of the how remaining in the age from the weight of the original amount. The average alori intake of these animals was 54 kal per day. Sine a small amount (less than 5% of the intake) of the food was lost through age bottom eah day, the two groups of rats reeived nearly equal amounts of alories. On eah of days 8, 1, 12, 15, 17, and 19 following surgery, 1 ontrol animal and 1 parenterally fed animal were killed. ah an;mal was weighed and anesthesized with ether. The small intestine was removed and gently perfused with oxygenated Krebs-Ringer biarbonate buffer, blotted, and weighed. The intestine was then suspended vertially and a 2-m setion was removed from the midpoint for hemial analysis. The stomah was removed, opened along the greater urvature, rinsed in ie-old saline, blotted, and weighed. The antrum was isolated by utting along its boundary with the oxynti gland area and weighed. The kidneys, testes, and spleen were removed, rinsed in ie-old saline, blotted, and weighed. From another set of identially treated animals, 1 intravenously alimented and 1 ontrol rat were killed on eah of days 16, 18, 2, and 23 after surgery. The oxynti gland area of the stomah, as well as the antrum, was isolated and weighed. In addition, the panreas was arefully disseted from the mesentary and fat, blotted, and weighed. In order to eliminate the possibility that our results were due to a differene in the nature of the diets rather than the routes of administration, 5 harnessed rats were maintained orally for 8 days on the fluid diet (49 kal per day). The weights of seleted organs were then ompared to the weights of the same organs from animals fed the standard rat how. Analysis of e(l omponents. The 2-m segment of small intestine was mined with sissors and homogenized in old.4 N perhlori aid. After standing in ie for 1 min, the preipitates were olleted by entrifugation at 2 x g for 1 min. The supernatant fluids were disarded, and the preipitates were washed twie with old.2 N perhlori aid and dissolved in 4 ml of.3 N KOH. The RNA was hydrolyzed quantitatively by inubating the tubes at 37 C for 9 min. DNA and protein were

3 May 1975 GI TRACT OF PARNTRALLY FD RATS 1179 repreipitated with 2 ml of 1% perhlori aid. Mter standing for 1 min in ie, the tubes were entrifuged for 15 min and the supernatant fluid was deanted into a seond tube. The pellets were washed one with 4 ml of.2 N perhlori aid and, after entrifugation, the supernatant fluids were determined at 232 and 26 JL, and the mirograms of RNA per ml were alulated aording to the formula of Munro and Flek. 8 The DNA-ontaining pellet was dissolved in 4 ml of 1% perhlori aid by heating in a boiling water bath for 15 min. Protein was removed by entrifugation for 2 min at 5 x g. Dupliate 2-ml aliquots were assayed for DNA ontent aording to the Burton proedure, as modified by Giles and Myers. 1 Calf thymus DNA was used as a standard. The protein pellet was then redissolved in 1 N NaOH. The protein ontent was determined by the biuret method. Determination of antral gastrin. After weighing, the antrum was mined with sissors and homogenized in 2 ml of old, deionized, distilled water. One milliliter of homogenate was immediately added to 4 ml of boiling distilled water and allowed to boil for 2 min. At the end of the heat extration period, the samples were entrifuged at 2 x g to remove debris. The supernatant fluid was then passed through a Millipore filter (.45 JL) and stored at -2 C. The gastrin onentrations of the filtrates were determined by radioimmunoassay using the method desribed by Yalow and Berson. 11 Results were expressed as piomoles of gastrin per antrum and nanomoles of gastrin per g wet weight of tissue. Statistis. There were no signifiant differenes in organ weight and gastrin levels related to the length of duration of hyperalimentation, so statistial analysis was performed on the two populations, ontrols and parenterally fed, without regard to the day on whih they were killed. Signifiane was assessed using the t-test for unpaired data. The differene between means was onsidered signifiant when the P value was less than.5. Results The 34-g average weight gain shown by the intravenously hyperalimented (IVH) rats was signifiantly less than that of the ontrols (table 1). Nevertheless, this 25% weight gain plus the findings of normal urine nitrogen and aetone and the absene of gluose from the urine indiate that the parenterally fed animals were being maintained in an anaboli and stable state of nutrition. Beause of the differenes in the final weights of the animals in the two groups, the weights of other tissues and organs are expressed per 1 g final body weight. There were no signifiant differenes between the weights of kidneys, spleen, and testes in the two groups of rats (table 1). The small intestine of the ontrol rats was 73% heavier than that of the IVH group, and the weight of the whole stomah of the ontrols was 25% greater. There was, however, no signifiant differene between the mean weights of the antral tissue in the two groups (table 1). Although antral weights were unhanged, there was an approximate 3-fold derease (P <.2) in antral gastrin in the parenterally fed animals (fig. 1). This derease was the same independent of whether gastrin levels were expressed per whole antrum or per g wet weight. There TABL 1. Whole body and relative organ weights of ontrol and parenterally fed ratsa Tissue Control (n 6) IVH' (n 6) p Initial body wt (g) Final body wt (g) Gain body wt (g) Kidneys (mg/1 g body wt) Spleen (mg/1 g body wt) Testes (mg/1 g body wt) Small intestine (mg/1 g body wt) Stomah (mg/1 g body wt) Antrum (mg/1 g body wt) a Results are means and SM. IVH, intravenously hyperalimented. NS, not signifiant. 141 ± ± ± ± ± ± ± ± ± ± 9 NS 188 ± 8 <.2 34 ± 4 < ± 43.7 NS 26 ± 13.8 NS 1526 ± 74 NS 1389 ± 6 < ± 18 < ± 11.6 NS

4 118 JOHNSON T AL. Vol. 68, No. 5, Part 1 were no differenes in antral gastrin of rats killed on the 8th or loth days following initiation of parenteral feeding and those killed later, on the 17th or 19th days. Thus, the derease in gastrin levels ours early and over a short period of time following essation of normal eating. The 2-m segment of gut from the ontrol animals weighed 6% more than the segment from the IVH group of rats and ontained twie as muh DNA (table 2). xpressed as mirograms of DNA per mg of tissue, however, there was no signifiant differene between the DNA ontent of the tissues. Total RNA was signifiantly less in the gut segment of the parenterally fed animal but signifiantly inreased when expressed per weight of tissue. Thus, if one assumes that ellular ontent of DNA is ' ; 8 8 a. > \.') r r \.') 2 2 u :=: u FIG. 1. Antral gastrin levels in ontrol and parenterally fed rats, expressed as piomoles per antrum and as nanomoles per g wet weight tissue. Bars and vertial lines represent means and SM for 6 observations. In eah ase the ontrol is signifiantly greater, P <.2. IVH, intravenously hyperalimented. Qj ' relatively onstant, the number of ells per segment of tissue has delined greatly in the hyperalimented group. The RNA ontent has not dereased nearly as muh, meaning that eah ell ontains more RNA. This is refleted in the highly signifiant inrease in the RNA:DNA ratio for the IVH group (table 2). Similarly, parenteral feeding resulted in a derease in total protein ontent of the gut but an inrease in protein ontent per mg of tissue or per ell. Colletively, these data indiate that the absene of food from the gut results in hypoplasia of the small intestine but hypertrophy of the remaining ells. In the seond group of animals (table 3), partiular attention was foused on the relationship between the weights of the antral and oxynti gland areas of the stomah. As with the initial study, the stomahs of the IVH group were signifiantly lighter than the stomahs of the ontrols. Again there was no signifiant differene in the antral weights. The oxynti gland area dereased 35% in weight, a perentage derease idential to that found in the whole stomah. The forestomah or rumen was not weighed diretly, but from the above data it is obvious that it also dereased in weight by at least 35%. Therefore, the antrum is the only physiologial or anatomial part of the stomah unaffeted by the absene of oral food intake. The panreati wet weight of the parenterally fed rat was only one-third (P <.1) that of the same organ in the ontrol animal (fig. 2). xpressed per 1 g body weight, the panreas of the ontrol rat was still 2.3 times larger (P <.1) than that TABL 2. Quantities of ell onstituents of 2 em of small intestine from ontrol and parenterally fed ratsa Measured parameter Control (n 6) Wet weight (mg) ± 14.4 DNA total (ILg) ± 12.9 ILg DNNmg tissue 1.38 ±.126 RNA total (ILg) ±.314 ILg RNNmg tissue ± 1.42 RNA:DNA ± 1.14 Protein total (mg) 41.5 ± 2.4 mg protein/mg tissue.177 ±.7 IVH' (n 6) p %of ontrol ± 1.3 < ± 11.6 < ±.134 NS' ±.278 < ± 1.7 < ± 2.7 < ± 1.6 < ±.9 < a Results are means and SM. IVH, intravenously hyperalimented. "NS, not signifiant.

5 May 1975 GI TRACT OF PARNTRALLY FD RATS 1181 TABL 3. Relationship between the weights of the various parts of stomah from ontrol and parenterally fed rats" Tissue Control (n 4) IVH' (n 4) p Bodywt (g) Stomah (mg/1 g body wt) Oxynti gland (mg/1 g body wt) Antrum (mg/1 g body wt) Oxynti gland/antrum 23 ± ± 5.8 <.1 69 ± 2 59 ± 26 < ± ± 22 < ± ± 6.7 NS 6.2 ± ±.45 <.5 a Results are means and SM. IVH, intravenously hyperalimented. NS, not signifiant. "' Q 1 :;: "' 8 u 6 4 I I u u 8 7 ; 6,..., -" "' 5.:::. "' 4 Q "" 3.!!' m 2 FIG. 2. Panreas weights from ontrol and parenterally fed rats, expressed as milligrams wet weight and milligrams wet weight per 1 g body weight. Bars and vertial lines represent means and SM for 4 observations. In eah instane the ontrol is signifiantly heavier, P <.1. IVH, intravenously hyperalimented. of the parenterally fed animal. The hanges in this tissue were espeially dramati and interesting beause this organ is not in ontat with the luminal ontents. As a ontrol experiment a group of rats was fed the liquid diet orally for 8 days. In table 4 the weights of some of their gastrointestinal organs are ompared to those from a group fed the standard rat how. Oral intake of the liquid diet did not inrease body weight. In fat, the final mean body weight of this group was idential to those groups given the same liquid diet by intravenous infusion (tables 1 and 3). Oral administration of this liquid diet, unlike intravenous administration, did not result in dereases of the weights of the.. oxynti gland area and small intestine. The weight of small intestine.was in fat inreased. From this experiment it appears that the dereases in weights of gastrointestinal tissues observed after intavenous administration of the liquid diet are not due to size of the animals or diet omposition. Disussion The data presented here demonstrate that the oral ingestion of food is neessary to maintain the strutural integrity of ertain tissues and aessory organs of the gastrointestinal trat. Maintenane of animals in an anaboli state by total parenteral nutrition has been desribed many times, but it is obvious from the urrent study that the presene of food in the gastrointestinal trat provides a stimulus or additional nutrients to ertain tissues whih annot be provided by intravenous nutrition. The strutural and funtional dependene of the gut on the presene of food in the gastrointestinal trat has also been suggested by two other types of studies. The well known fat that starvation auses a loss of small intestinal mass out of proportion to the weight loss of the rest of the body has already been disussed. Sine these animals are in negative nitrogen balane and losing weight, these data are diffiult to interpret. In earlier experiments we attempted to irumvent this problem by studying intestinal struture in young rats prevented from weaning but maintained on maternal milk. 12 Non weaned 24-day-old rats had a signifiantly lighter small bowel (expressed per 1 g body weight) than did their weaned littermates. In addition the small intestine of nonweaned animals ontained less RNA,

6 1182 JOHNSON T AL. Vol. 68, No. 5, Part 1 TABL 4. Relationship of weights of various tissues from rats fed oral solid and liquid diets Tissue Solid (n 5) Liquid (n 5) p Bodywt (g) Oxynti gland (mg/1 g body wt) Antrum (mg/1 g body wt) Small intestine (mg/1 g body wt) 24 ± ± 8.8 < ± ± 11.3 NS 55.4 ± ± 5.8 NS 319 ± ± 134 NS Results are means and SM. NS, not signifiant. DNA, and protein. 12 We onluded that some aspets of gut development are dependent on the presenoe of solid food. There are several fators whih ould possibly ontribute to the dependene of the gastrointestinal trat on orally ingested nutrients. The at of eating and the hemial and physial presene of food within the trat stimulate gastri, panreati, and intestinal seretion. 13 Motility of the entire gastrointestinal trat is stimulated and regulated by the presene of food. In the absene of food these funtions would be greatly redued, 14 and hypotrophy of the organs, ells, and musles involved ould our. This, however, does not seem to be the explanation for our results, for in the small intestine hypoplasia has ourred and the remaining ells are hypertrophied. Furthermore, it is diffiult to explain on this basis why antral mass was unaffeted by parenteral nutrition. The antrum is heavily musled and is responsible for most of the mixing of gastri ontents and the emptying of those ontents into the duodenum. Yet this tissue showed no evidene of disuse atrophy. The degree of dependene of tissues lining the gastrointestinal trat on nutrients absorbed from the lumen an only be surmised. Presumably the flow of nutrients to the serosal side of the trat is unaltered during total parenteral nutrition. xept for ontributions from desquamated ells and minimal seretions, the large amounts of monosaharides, amino aids, and glyerides normally entering the small intestinal muosa are absent. Dependene of the intestinal epithelial ells on these absorbed substanes ould explain some of the observed results. The dramati dereases in relative weights of the panreas and stomah, however, annot be explained by this hypothesis. The panreas is not in ontat with the lumen and obviously derives no nutrients from the luminal ontents. The gastri muosa normally absorbs only short hain fatty aids (C 2 -C 4 ) and then only if the ph is low enough, usually below 4, so that they are unionized. 15 The quantity of water-soluble nutrients, gluose, and amino aids normally absorbed by the stomah is nutritionally negligible. 15 Sine eating and the presene of food within the digestive trat are the normal stimulants of gastrointestinal hormone release, 16 this release is absent in the parenterally fed animals. Not only is hormone release impaired, but the atual synthesis or maintenane of antral gastrin supplies appears to depend upon the stimuli aompanying the oral ingestion of food and the presene of the food within the gut. In rats fasted for 3 days, antral gastrin levels dereased to one-tenth normal, and serum gastrin levels dropped from 33 to 7 pg per ml. 16 Dependene on solid food for the development of antral gastrin levels has been demonstrated in young rats. 12 Before weaning, antral gastrin ontent is quite low but undergoes a 15-fold inrease to adult levels within a day or two following the first ingestion of solid food. Young rats prevented from weaning had antral gastrin levels equal to only one-third of those present in littermates weaned at the normal time. 12 We have only studied gastrin, but it is not unreasonable to assume that maintenane of normal levels of holeystokinin and seretin may also depend upon the presene of foodstuffs within the gut. Gastrin has been shown to be an important trophi hormone for the oxynti gland area of the stomah, the small intestine, and the panreas. 19 Protein synthesis and growth of the antral region of the stomah are unaffeted by exogenous

7 May 1975 GI TRACT OF PARNTRALLY FD RATS 1183 gastrin This pattern is idential to that observed in the parenterally fed animals. Following surgial removal of the antrum and the major endogenous supply of gastrin, the RNA and DNA ontent of the small intestine dereased and the RNA:DNA ratio inreased. 22 Again these findings parallel those of the urrent study. The defiieny in gastrointestinal hormones has never been impliated, until reently, as a ausative fator in the disproportionate hanges whih our to various gastrointestinal and related tissues when food is kept out of the gut. Although muh more work is needed, the findings of the urrent study make this hypothesis one to be onsidered in future experimentation in this area. Several definitive onlusions an be drawn from the results presented here. First, the stimuli arising from the oral ingestion and/or physial and hemial presene of food within the gastrointestinal trat are neessary for the maintenane of the strutural integrity of ertain tissues within or related to that trat. Seond, the disproportionate hanges noted are speifi for ertain tissues and, therefore, appear to be independent of general hanges suh as disuse hypotrophy and the absene of luminal nutrients. Third, maintenane of normal endogenous stores of gastrin is dependent on oral ingestion and the physial and hemial presene of food in the gastrointestinal trat. RFRNCS 1. Steiner M, Bourges HR, Freeman LS, et a!: ffet of starvation on the tissue omposition of the small intestine of the rat. Am J Physiol215:75-77, MNeil LK, Hamilton JR: The effet of fasting on disaharidase ativity in the rat small intestine. Pediatris 47:65, Altmann GG: Influene of starvation and refeeding on muosal size and epithelial renewal in the rat small intestine. Am J Anat 133: , Brown OH, Levine ML, Lipkin M: Inhibition of intestinal epithelial ell renewal and migration indued by starvation. Am J Physiol 25: , Deren JJ, Broitman SA, Zamhek N: ffet of diet upon intestinal disaharidases and disaharide absorption. J Clin Invest 46: , Levin RJ, Newey H, Smyth DH: The effets of adrenaletomy and fasting on intestinal funtion in the rat. J Physiol 177:58-73, Steiger, Vars HM, Dudrik SJ: A tehnique for longterm intravenous feeding in unrestrained rats. Arh Surg 14:33-332, Munro HN, Flek A: Reent developments in the measurement of nulei aids in biologial materials. Analyst 91:78-89, Burton K: A study of the onditions and mehanism of the diphenylamine reation for the olorimetri estimation of deoxyribonulei aid. Biohem J 62: , Giles KW, Myers A: An improved diphenylamine method for the estimation of deoxyribonulei aid. Nature 26:93, Yalow RS, Berson SA: Radioimmunoassay of gastrin. Gastroenterology 58:1-14, Lihtenberger L, Johnson LR: A possible role of gastrin in the ontogeni development of the small intestine. Am J Physiol227:39-395, Grossman MI: Neural and hormonal stimulation of gastri seretion of aid. Handbook of Physiology, Set. 6: Alimentary Canal, vol. 2. dited by CF Code. Washington DC, Amerian Physiologial Soiety 1967, p Towne JB, Hamilton RF, Stephenson DV: Mehanism of hyperalimentation in the suppression of upper gastrointestinal seretion. Am J Surg 126: , Davenport HW: Physiology of the Digestive Trat. Chiago, Year Book Medial Publishers, 1966, p Lihtenberger LM, Castro GA, Copeland M, et a!: The effet of food on rat antral gastrin onentration. Gastroenterology 66:A22, Johnson LR, Aures D, Yuen L: Pentagastrin indued stimulation of the in vitro inorporation of "C-leuine into protein of the gastrointestinal trat. Am J Physiol 217: , Johnson LR, Chandler AM: RNA and DNA of gastri and duodenal muosa in antretomized and gastrin treated rats. Am J Physiol 224:937-94, Mayston PD, Barrowman JA: The influene of hroni administration of pentagastrin on the rat panreas. Q J xp Physiol 56: , Crean GP, Marshall MW, Rumsey RD: Parietal ell hyperplasia indued by the administration of pentagastrin (ICI 5, 123) to rats. Gastroenterology 57: , Johnson LR, Aures D, Hakanson R: ffet of gastrin on the in vivo inorporation of 14 C-leuine into protein of the digestive trat. Pro So xp Bioi Med 132: , Johnson LR, Guthrie PD: Muosal DNA Synthesis: A short term index of the trophi ation of gastrin. Gastroenterology 67: , 1974