Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

Gut, 1980, 21, 662-668

Immunoelectrophoretic studies on human small intestinal brush border proteins: cellular alterations in the levels of brush border enzymes after jejunoileal bypass operation

H SKOVBJERG, E GUDMAND-H0YER, 0 NOREN, AND H SJOSTROM From the Department ofBiochemistry C, The Panum Institute, Medical-gastroenterological Department C, and Surgical-gastroenterological Department D, Herlev Hospital, University of Copenhagen, Denmark

SUMMARY The amounts of lactase (EC 3.2.1.23), sucrase (EC 3.2.1.48), maltase (EC 3.2.1.20), microvillus aminopeptidase (microsomal EC 3.4.11.2), and dipeptidyl peptidase IV (EC 3.4.14. x) in biopsies from proximal and distal were studied by quantitative crossed immuno- electrophoresis and enzymatic assays in obese patients one and six months after jejunoileal bypass operation and compared with peroperative levels. They were related to DNA and protein content. The protein/DNA ratio fell 28-43 % postoperatively. Except for ileal lactase and sucrase all enzymes showed decreased levels when expressed per mg protein and an even more pronounced decrease when related to DNA. Lactase and sucrase levels in ileum were increased or unchanged. A constant correlation between the amount of immunoreactive enzyme protein and enzymatic activity was shown for all enzymes except maltase. The results suggest that the bypass operation is followed http://gut.bmj.com/ by an increased amount of enterocytes devoid of or low in enzymatic activity and protein content. The amounts of lactase and sucrase in ileum are increased in relation to the other enzymes. No immunoreactive enzymes with zero or depressed activity were detected.

Jejunoileal bypass operation frequently used in the to an altered amount of active enzyme molecules in on September 29, 2021 by guest. Protected copyright. treatment of morbid results in malabsorption the cells, or to the presence of more or less active and weight loss. The increase in length and mucosal enzyme molecules. Experimental studies on bypass surface of the functioning shunt seems to be im- operated rats7 have shown an increase in the DNA portant for the weight stabilisation that usually content per segment and a falling occurs 12 to 18 months postoperatively. Increase of activity of brush border enzymes expressed per mg villus height and epithelial cell hyperplasia has been DNA indicating an increased amount of cells described after the operation.'2 Changes in the devoid of or low in enzymatic activity. These cells specific activity of some of the brush border enzymes appear either as crypt cells or immature cells on the (disaccharidases and peptidases) which play an villi. Corresponding studies on human intestine important role in the final digestion have been have not been undertaken. described,' 3 4-6 but the results are conflicting, The present study investigates the cellular changes probably because not only the cellular enzyme in the small intestinal brush border enzymes by the content but also the mucosal protein change post- use of quantitative immunoelectrophoresis and operatively. DNA measurement in combination with measure- The reported changes in brush border enzyme ments of enzymatic activity and protein. Quantita- activity can be caused by an alteration in the relative tive immunoelectrophoresis, which earlier has been number of enterocytes or by a change in enzymatic shown to be a valuable method in the study of activity at a cellular level. Such changes might be due human brush border enzymes,8 gives an expression of the amount of enzyme protein independent of the Received for publication 6 February 1980 enzymatic activity. This enables an estimation of 662 Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

Immunoelectrophoretic studies on human small intestinal brush border proteins 663 possible enzyme molecules with altered enzymatic patients with 12.5 cm jejunum and 375 cm ileum activity. Moreover, changes in the internal relations (type IL) left in continuity. Peroperative biopsies between the enzymes measured can be more precisely from the points of division-that is, proximal determined, as they are quantified on the same im- jejunum and distal ileum-were obtained in all munoelectrophorectic plate. The method also, for patients. One to two and six to eight months post- the first time, allows changes occurring in the operatively peroral biopsies were taken 25 cm below amount of brush border maltase to be estimated, as the ligament of Treitz-that is, jejunal biopsies the enzymatic estimation of maltase activity also from patients with type I and ileal biopsies from includes the activity of at least the sucrase-isomaltase patients with type II operation. Nine patients with complex. type I and seven with type 1I operation had biopsies taken one to two months postoperatively, while Methods nine patients with type l and six with type IL operation had biopsies taken six to eight months postopera- PATIENTS tively. All biopsies were frozen in dry ice immedi- Twenty patients having jejunoileal shunt operation ately after removal. performed as a part of The Danish Obesity Project were investigated. Eleven patients were operated TECHNIQUE according to Payne and De Wind9 with 37.5 cm From the peroperative biopsies mucosal samples jejunum and 12.5 cm ileum (type I) and nine (3-16 mg) were obtained by careful scraping. The

Jejunum Ileum Jejunum Ileum Lactase l 1IAt\** ** NS http://gut.bmj.com/ z c Fig. 1 Changes in amount oj E enzyme protein for each patient cn ,estimated in crossed E immunoelectrophoresis as area U Sucrase Microvillus Aminopeptidase under the precipitate per mg DNA

.at operation, one to two and six to on September 29, 2021 by guest. Protected copyright. eight months after jejunoileal shunt operation. Jejunal values are from patients with type I operation, ileal values from patients with type II operation. NS: not significant. The P values are obtained by testing the z postoperative values against the 0 peroperative values. Interrupted lines: in these patients the one to E CN~ two months postoperative biopsies E Dipeptidyl Peptidase I1: were not taken. U OP 1 6 OP 1 6NS

*** = p< 0.01 200 * * = P< 0.02 *= P< 0.05 100

0 OP 1 6 OP 6 Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

664 Skovbjerg, Gudmand-Hoyer, Noren, and Sjostrirm

Maltase Lactase Sucrase Jejunum Ileum Jejunum Ileum Jejunum Ileum

NS * *..NS ..NS .***NS NS c

0 4

E Fig. 2 Changes in amount of enzyme' protein for each patient estimated in 2 [ 2 l 6 crossed immunoelectrophoresis as area under the precipitate per mg protein at operation, one to two and I: nu six to eight months after jejunoileal OP l 6 Of '1 6 OP 16 OP 16 OP 16 OP 16 shunt operation. Jejunal values are from patients with type I operation, Microvillus Aminopeptidase Dipeptidyl Peptidase XI ileal values from patients with type 1! Jejunum Ileum Jejunum Ileum operation. The bars represent the OP l 6OPNS NS*6PNSNS6Ol median values. NS: not significant. c 8 The P values are obtained by testing the postoperative values against the 0 *** =P 0.01 a. peroperative values. w. cm =p<0.02 E 4 * =p< 0.05 E 2- U ck:WL OP 16 OP 16 0P016 OP 16

mucosal material was then handled as the peroral standard. DNA concentration was measured using http://gut.bmj.com/ biopsies. Each biopsy was homogenised and the the fluorometric method of Kissane and Robins.13 brush border proteins solubilised and quantified in Calf thymus DNA was used as a standard. Extrac- crossed immunoelectrophoresis as earlier described.8 tion of was omitted, as it resulted in irrepro- To summarise briefly: each sample was homogenised ducible loss of material and as the contribution of in 40 ,ul of a 2% Triton X-100 solution and incu- fluorescence from lipids in the homogenate was bated for one hour at 4°C. Papain at a final concen- shown to be less than 8% of the measured value. tration of 2.5 mg/ml was added and 15 pl removed The statistical evaluations were performed with the for enzymatic analysis. The rest of the homogenate Mann-Whitney U test and the Wilcoxon test for on September 29, 2021 by guest. Protected copyright. was incubated for 15 minutes at 37°C and subse- paired data. quently centrifuged at 50.000X g for two hours. The supernatant was analysed in crossed immuno- Results electrophoresis. The precipitates, earlier identified by enzymatic staining,10 were stained by Coomassie As shown earlier8 the solubilisation procedure re- Brillant Blue 250 and their areas expressed in cm2 by leases the studied enzymes almost completely from use of an electronic integrator. the brush border membranes. The mean rest activity Lactase (EC 3.2.1.23), sucrase (EC 3.2.1.48), and of microvillus aminopeptidase in the pellet after maltase activities were assayed in the homogenate solubilisation was for each patient group 2% (range- using their corresponding disaccharides as sub- 0-5%) of the total activity. The protein/DNA ratio strates.1" The activities of microvillus amino- in the peroperative biopsies obtained by mucosal peptidase (microsomal, EC 3.4.11.2) and dipeptidyl scrapings was found to be of the same magnitude peptidase IV (EC 3.4.14. X) were determined using as in peroral biopsies from the ligament of Treitz in L-alanine-p-nitroanilide and glycyl-L-proline-p- patients without gastrointestinal disease. This nitroanilide as substrate respectively.10 As a control indicates that the mucosal scrapings are comparable of the solubilisation procedure the microvillus with the peroral biopsies where their content of aminopeptidase activity was determined in the villi and crypt cells is concerned. pellet. The protein concentration of the homogenate Peroperative sucrase and lactase levels as measured was measured12 in a small sample taken before ad- by immunoelectrophoresis and enzymatic activity dition of papain using bovine serum albumin as a were significantly greater (P<0-05) in jejunal than Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

Immunoelectrophoretic studies on human small intestinal brush border proteins 665 in ileal biopsies. The levels of maltase, microvillus other enzymes were decreased in parallel during the aminopeptidase, and dipeptidyl peptidase IV seemed first postoperative months and remained decreased, lower in proximal jejunum than in distal ileum but though not significantly, six to eight months post- the differences were not statistically significant. operatively. The variations in enzyme protein per mg DNA When expressed as the area under the precipitate for each patient for maltase, sucrase, lactase, micro- per mg protein (Fig. 2) all enzymes in jejunum villus aminopeptidase, and dipeptidyl peptidase IV tended to decrease. In ileum, lactase and sucrase one to two and six to eight months after operation, increased while the other enzymes decreased almost compared with the peroperative levels, are seen in in parallel. Fig. 1. In jejunum all the measured enzymes de- For all the enzymes studied, except maltase, there creased in parallel during the first postoperative was a close correlation between enzymatic activity months and remained significantly lower than per- and amount of enzyme protein irrespective of de- operatively six to eight months after the operation. creasing or increasing postoperative activity (Fig. 3). In ileum the levels of lactase and sucrase were not Figure 4 exemplifies the results by showing the significantly changed during the study, while the electrophoretic patterns in jejunal biopsies from the

Maltase Lactase . 400k 800k

* 0 0 . 200 - *1 . 4001 %0 . 0 0 0 0 0 1% ° 0 S 00 8o 0 00 a0% lo 04 *oNO~~~~~~ http://gut.bmj.com/ 0L 10 20 1 2 Sucrose Microvillus Aminopeptidase z 400H 0 * 0 600k Fig. 3 The amount of enzyme 0) protein expressed as area under the 0 * E * 0 precipitate versus the enzymatic on September 29, 2021 by guest. Protected copyright. activity. Some (crowded) points It represent more than one patient. Five 2001- *00 1 postoperative values ofdipeptidyl 0 200 00 o peptidase IV were omitted, as the 0 ** enzymatic analysis failed. 0of 0' . Qi 2 4 5 10 Dipeptidyl PeptidaseJE U/mg DNA 0 * Peroperative values

_ =~~~~~~~e o Postoperative values 0 2001

o 100 0 o 0

0' 1 2 U /mg DNA Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

666 Skovbjerg, Gudmand-Hoyer, Noren, and Sjostrom http://gut.bmj.com/

Fig. 4 Crossed immunoelectrophoresis ofbiopsies taken from the proximal part ofjejunum (a) during operation, (b) one month after operation from patient A. The IgG concentration in the gel was 30 [lg/crn2. The amount of antigen is equivalent to 4 and 15 t£g DNA respectively. L: lactase-phlorizin hydrolase. S: sucrase-isomaltase. M: maltase. A: microvillus aminopeptidase. G: dipeptidyl peptidase IV. AP: alkaline phosphatase (this precipitate is visible only after enzymatic staining). on September 29, 2021 by guest. Protected copyright. same patient (type I operation) taken at operation while the other enzymes were all reduced by ap- and one month postoperatively (marked in Fig. 1 by proximately 75% one month postoperatively com- A). This patient*had :;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.-a marked fall in lactase (99%), pared...... with:.the operative...... level. :. ;... :i The Table shows the relation between protein and Table Mean values for ratio mg protein/mg DNA at DNA content in the biopsies. The 28-43% fall in operation, one to two months and six to eight months the ratio postoperatively is of the same magnitude postoperatively for jejunum and ileum* as found in the small intestine from bypass operated rats.7 Operation Postoperatively (months) 1-2 6-8 Discussion 45-8 26 1 29.2 Jejunum (19-9-94-3) (16-6-37.3) (13-5-49 3) In the present study changes in the amount of n= 18 n=9 n=9 P<0-01 P<002 different brush border enzymes have been examined 47.8 27-2 34-5 in relation to changes in DNA and protein content. Ileum (32-3-71-3) (22-9-34.1) (20.7-51.0) in in- n= 17 n=7 n=6 The 28-43%0 fall the protein/DNA ratio P<0t)0 NS dicates that the protein content in the average enterocyte is diminished postoperatively. An in- *Numbers in parentheses are ranges. The P values apply to differences of cells between operation and one to two or six to eight months postopera- creased number dividing crypt increasing ii ely. NS: not significant. the average DNA content in the cells might cont- Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

Immunoelectrophoretic studies on human small intestinal brush border proteins 667 tribute to the decreased ratio but could not account study therefore presents evidence for the fact that for all of it. The reduced amount of protein in the the increased amount of enterocytes shown in enterocytes is in accordance with the findings of morphological studies are more or less deficient in Dan0 et al.14 and is probably due to the postoperative enzymes. Also, when expressed per mg protein, the protein malabsorptive state.1516 brush border enzymes show a tendency to decrease The almost parallel decrease of all enzymes in except for lactase and sucrase in the postoperative jejunum per mg DNA indicates a decreased amount ileum. Whether the total amount of brush border of enzyme per cell. Morphological studies in rat and enzymes per unit length intestine is changed post- human1 317 have shown an increased amount of operatively cannot be answered in this study. enterocytes on the villi after bypass operation, while However, in experimental studies in rats it has been no alteration in the proportion of enterocytes to the shown that, even if the enzymes decrease per mg total cell number has been described. The falling protein and, even more pronouncedly, per mg DNA enzyme activity per cell indicates the existence of a after jejunoileal shunt operation the amount of these significant number of enterocytes devoid of or low enzymes per unit length intestine is increased.7 in enzyme content. Whether the patients in addition This may also be true for the human. to an increased migration rate of the enterocytes17 In the present study we have also shown that, and increased cell proliferation'8 have an increased except for maltase, there is a close correspondance turnover rate of the enterocytes cannot be settled between the amount of immunoreactive enzyme in this study. protein and enzymatic activity before and after by- In ileum the levels of lactase and sucrase were un- pass operation. Thus we conclude that the decreased changed during the period that was studied, while enzymatic activity is not caused by production of the levels of maltase, microvillus aminopeptidase, immunoreactive enzyme molecules with zero or and dipeptidyl peptidase IV were decreased almost decreased enzymatic activity. Provided that the in parallel when expressed per mg DNA. This solubilisation procedure also releases possible indicates that lactase and sucrase at the cellular level enzyme precursor molecules from intracellular are increased in relation to the other enzymes, and it particles we can also conclude that the suggested, suggests that these disaccharidases are regulated increased population of immature cells on the villi

independently of the other enzymes. An explanation does not contain an increased amount of immuno- http://gut.bmj.com/ might be that there is a relatively increased synthesis reactive enzyme precursors. Where maltase is con- of lactase and sucrase due to enhanced intraluminal cerned the precipitate area expresses directly the stimulation by lactose and sucrose, because the amount of this enzyme, while the enzymatic measure- short functioning jejunum (12.5 cm) in these patients ment represents the activities of both maltase and does not have enough capacity to split these disac- the sucrase-isomaltase complex. Thus the immuno- charides. This result is supported by the concept' 17 electrophoretic measurement of this enzyme was not that the intraluminal nutrition exerts a major in- expected to be correlated with the enzymatic activity. fluence on the splitting and absorptive functions of on September 29, 2021 by guest. Protected copyright. the human small intestine. Alternatively, the catabolism of maltase, microvillus aminopeptidase, The jejunoileal shunt operations were performed at and dipeptidyl peptidase IV in ileum might be surgical department D, Herlev Hospital, Denmark, relatively increased postoperatively. Duodenal juice as a part of the Danish Obesity Project. The authors increases the turnover rate of some brush border are grateful to Ms D Anthonsen, Ms D Rasmussen, enzymes.'9 20 Theoretically, maltase, microvillus Ms M v Holstein, and the nursing staff of medical- aminopeptidase, and dipeptidyl peptidase IV- gastroenterological Department C, Herlev Hospital, which, in contrast with lactase and sucrase, seem to for skilful technical assistance. This work was sup- have a higher peroperative level in the trypsin free" ported by the Danish Medical Research Council distal ileum than in the proximal jejunum-might be (No. 512-10175), P Carl Petersens Fond (No. B more susceptile to the catabolic action of duodenal 1076) and the Novo Foundation. juice. The patient presenting the extreme fall in jejunal References lactase (Fig. 4) is an example of an individual variation. For unknown reasons the jejunal entero- 1Dudrick SJ, Daly JM, Castro G, Akhtar M. Gastro- seem to have a intestinal adaptation following small bowel bypass for cytes in this patient temporarily obesity. Ann Surg 1977; 185: 642-7. selective decreased level of lactase compared with 2Solhaug JH, Tvete S. Adaptive changes in the small the other enzymes. intestine following bypass operation for obesity. A Alterations in the amount of enzyme related to radiological and histological study. Scand J Gastro- DNA reflect the fluctuations at a cellular level. This enterol 1978; 13: 401-8. Gut: first published as 10.1136/gut.21.8.662 on 1 August 1980. Downloaded from

668 Skovbjerg, Gudmand-Hoyer, Noren, and Sjostrom

3Iversen BM, Schj0nsby H, Skagen DW, Solhaug JH. in the presence of Triton X-100. Anal Biochem 1975; Intestinal adaptation after jejunoileal bypass operation 63: 414-7. for massive obesity. Eur J Clin Invest 1976; 6: 355-60. 13Kissane JM, Robins E. The fluorometric measurement 4Stein TA, Wise L. Functional adaptation of the in- of deoxyribonucleic acid in animal tissues with special testinal mucosal enzymes after jejunoileal bypass for reference to the central nervous system. J Biol Chem morbid obesity. Am J Clin Nutr 1978; 31: 1143-8. 1958; 233: 184-8. 5Gudmand-H0yer E, Asp NG, Skovbjerg H, Andersen 14Dan0 P, Nielsen OV, Petri M, J0rgensen B. Jejunal B. Lactose malabsorption after bypass operation for morphology and mucosal enzyme activity following obesity. Scand J Gastroenterol 1978; 13: 641-7. intestinal shunt operation for obesity. Scand J Gastro- "Asp NG, Gudmand-H0yer E, Andersen B, Berg NO. enterol 1976; 11: 129-34. Enzyme activities and morphological appearance in '5Editorial. Intestinal adaptation and hepatic decom- functioning and excluded segments of the small in- pensation after jejtunoileal bypass for morbid obesity. testine after shunt operation for obesity. Gut 1979; Nutr Rev 1977; 35: 43-5. 20: 553-8. '6Solhaug JH, Grundt I. Metabolic changes after jejuno- 7McCarthy DM, Kim YS. Changes in sucrase, entero- ileal bypass for obesity. Scand J Gastroenterol 1978; kinase, and peptide hydrolase after intestinal resection. 13: 169-75. The association of cellular hyperplasia and adaptation. '7Gleeson MH, Cullen J, Dowling RH. Intestinal J Clin Invest 1973; 52: 942-51. structure and function after small bowel bypass in the 'Skovbjerg H, Sjostrom H, Noren 0, Gudmand- rat. Clin Sci 1972; 43: 731-42. H0yer E. Immunoelectrophoretic studies on human 18Barry RE, Barisch J, Bray GA, Sperling MA, Morin RJ, small intestinal brush border enzymes. A quantitative Benfield J. Intestinal adaptation after jejunoileal bypass study of brush border enzymes from single small in- in man. Am J Clin Nutr 1977; 30: 32-42. testinal biopsies. Clin Chim Acta 1979; 92: 315-22. 19Alpers DH, Tedesco FJ. The possible role of pancreatic 'Payne JH, DeWind LT. Surgical treatment of obesity. proteases in the turnover of intestinal brush border Am J of Surg 1969; 118: 141-7. proteins. Biochim Biophys Acta 1975; 401: 28-40. "Skovbjerg H, Noren 0, Sjostrom H. Immunoelectro- 20Kwong WKL, Seetharam B, Alpers DH. Effect of phoretic studies on human small intestinal brush exocrine pancreatic insufficiency on small intestine in border proteins. A qualitative study of the protein the mouse. Gastroenterology 1978; 74: 1277-82. composition. Scand J Clin Lab Invest 1978; 38: 723-9. 21Chung YC, Kim YS, Shadchehr A, Garrido A, "Dahlqvist A. Assay of intestinal disaccharidases. Anal MacGregor IL, Sleisenger MH. Protein digestion and

Biochem 1968; 22: 99-107. absorption in human small intestine. Gastroenterology http://gut.bmj.com/ 2Wang CS, Smith RL. Lowry determination of protein 1979; 76: 1415-21. on September 29, 2021 by guest. Protected copyright.