Pectin-Supplemented Enteral Diet Reduces the Severity of Methotrexate- Induced Enterocolitis in Rats

Y. MAO, B. KASRAVI, S. NOBAEK, L. Q. WANG, D. ADAWI, G. ROOS, U. STENRAM, G. MOLIN, S. BENGMARK & B. JEPPSSON Depts. of Surgery, Pathology, and Food Technology, Lund University, Lund, Sweden

Mao Y, Kasravi B, Nobaek S, Wang LQ, Adawi D, Roos G, Stenram U, Molin G, Bengmark S, Jeppsson B. Pectin-supplemented enteral diet reduces the severity of methotrexate-induced enteromlitis in rats. Scand J Gastroenterol 1996;31:558-567.

Background: Administration of methotrexate (MTX) to rats fed an elemental diet results in a high mortality from severe enteromlitis. Previous studies have shown that pectin is an important precursor of substrates for intestinal structure and function and may facilitate intestinal recovery after enterocolitis. The aim of this study is to evaluate the effect of pectin on MTX-induced enterocolitis in rats. Methodr: Rats received intragastric infusion of either 1% pectin-supplemented or pectin-free elemental diet from the beginning of the study via a gastrostomy. On the 4th day animals received either MTX, 20 mgkg intraperitoneally, or saline injection and were killed on the 7th day for sampling. Results: Pectin supplementation significantly decreased body weight loss, organ water content, and intestinal rnyeloperoxidase levels and increased mucosal protein, DNA, and RNA content in enteromlitis rats. The intestinal permeability was increased by administration of MTX, and pectin supplementation significantly reversed the increased permeability in the distal small bowel and colon. Pectin Supplementation also lowered the magnitude of bacterial translocation, decreased plasma endotoxin levels, and restored bowel microecology. Conclusions: Pectin significantly decreased MTX-induced intestinal injury and improved bowel integrity.

Key words: Bacterial translocation; elemental diet; enteromlitis; methotrexate; microflora; pectin

Bengt Jeppsson, M.D., Ph.D., Dept. of Surgery, Lund University Hospital, S-221 85 Lund, Sweden (fa: + 46 46-1 72004)

Chemotherapeutics exert their antineoplastic effect by being highly toxic to rapidly dividing cells. This cytotoxicity affects not only tumor cells but also rapidly dividing host cells such as those of gastrointestinal mucosa (1). It has been shown that high doses of methotrexate (MTX) cause severe enteromlitis, weight loss, anorexia, and intestinal atrophy (2-4). The use of elemental diets during cancer chemo- therapy has been recommended to avoid the mechanical abrasion on the gastrointestinal tract and reduce the severity of intestinal damage and toxicity (5 ,6) . These diets provide nitrogen as amino acids, and carbohydrate as simple sugars such as glucose or sucrose (1,5-7). Despite these advantages of feeding an elemental diet with chemotherapy, studies (7,8) have indicated that administration of MTX to rats fed an elemental diet has consistently resulted in severe, non- perforated enteromlitis with high mortality.

Most liquid formula diets, such as elemental diets, do not contain fiber and may cause intestinal atrophy. The addition of fiber to an elemental diet can benefit the patients with certain intestinal pathologic conditions (9). Many studies have shown that pectin, a water soluble non-cellulose dietary fiber, is increasingly recognized as an important precursor of substrates for intestinal structure and function (10-12). However, very few studies have focused on the effects of pectin on chemotherapeutics-induced enteromlitis.

The purpose of this study was to evaluate the effects of pectin on MTX-induced enteromlitis in rats. Previous studies on the effects of liquid formula diets on gastrointestinal function have mainly focused on the stomach and small bowel, and the effect of these diets on the colon has been studied less extensively (9). Colonic mucosa depends greatly on the luminal content of short-chain fatty acids (SCFAs), which are mainly produced by fermentation of fiber in the colon. In this study we investigated the colonic and the small- intestinal changes in MTX-induced enterocolitis and its response to pectin supplementation.

Three experiments were performed. The effect of pectin on intestinal injury and nutritional status after MTX administra- tion was studied first. In the second experiment the intestinal mucosal permeability as a measure of the physical barrier function was evaluated. In the third experiment the effect of pectin supplementation on intestinal microecology, bacterial translocation, and endotoxemia was investigated in MTX- induced enteromlitis rats.

M A T E R W AND METHODS

Animals Eighty-four male Sprague-Dawley rats (M~l leg i rd , Viby,

Denmark), weighing 200-250 g, were housed individually in

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

Pectin in Enterocolitis 559

Table I. Experimental design

Experimental groups Studies performed

Experiment 1, n = 24 Control-ED, n = 6 Control-pectin, n = 6 MTX-ED, n = 6* MTX-pectin, n = 6

Experiment 2, n = 24 Control-ED, n = 6 Control-pectin, n = 6

MTX-pectin, n = 6 MTX-ED, n = 6

Body weight change, histopathologic study, organ water content, and mucosal protein, DNA, and RNA content

Body weight change, and intestinal permeability assessed as "Cr-EDTA clearance

Experiment 3, n = 34 Normal, n = 6 Control-ED, n = 6 Control-pectin, n = 6 MTX-ED, n = 8 MTX-pectin, n = 8

Body weight change, intestinal MPO level, intestinal microflora, bacterial translocation, and plasma endotoxin level

MTX = rnethotrexate; ED = elemental diet; EDTA = ethylene- diaminetetraacetic acid; MPO = myeloperoxidase.

* Excluding two rats in this group that died.

wire-bottomed cages (to limit coprophagy) 5 days before experiments for acclimatization. The rats were subjected to alternate 12-h periods of dark and light and were allowed ad libitum intake of standard rat chow (R3, Lactamin AB, Stockholm) and water. Three days before experiments all the animals were fed a fiber-free liquid elemental diet (Nutri- drink, Nutricia Nordica Al3, Solna, Sweden) ad libitum to eliminate the effect of residual fiber on intestinal nutrition. On the day of experiment the rats were weighed, and a gastros- tomy was placed aseptically under ether anaesthesia and was connected to a swivel apparatus to enable long-term infusion. The rats were randomized to different groups in three experiments (Table I). Experimental design was approved by the Animal Ethics Committee of Lund University.

The rats in the normal group had sham operations (inserting the same type of gastrostomy tube with one end blocked under the skin) and were then allowed to eat rat chow freely throughout the study. The remaining rats received intragastric infusions (2.0 mlh) of normal saline overnight postopera- tively. On the 2nd day the animals received a continuous intragastric infusion (2.5 mlb) of either free elemental diet (ED) (240 kcal/kg and 1.28 g of nitrogenkdday) or supple- mented with 1% citrus pectin (Sigma Chemical Co., St. Louis, Mo., USA). The diets were infused continuously for 6 days. The addition of pectin does not change the osmolarity or the caloric value of the solution (13). On the 4th day the animals were weighed and received an intraperitoneal injection of MTX, 20 mgkg (Methotrexate sodium salt, Lederle Arznei- mittel, Wolfratshausen, Germany), in MTX-ED and MTX- pectin groups or saline in control-ED, control-pectin, and normal groups. Samples were collected on the 7th day. Two rats in the MTX-ED group in experiment 1 died 2 days after MTX injection after blockage of the infusion tube.

Histopathologic study On the 7th day the animals were weighed and then killed

with an overdosage of ether. A midline incision was made, and the intestine from the pylorus to the anus was excised, freed of its mesenteric fat, and rinsed in cold saline. Two 0.5-cm samples of terminal ileum and proximal colon were obtained and placed in phosphate-buffered 4% formaldehyde. Paraffin-embedded samples were cut and studied under light microscopy after being stained with hematoxylin and eosin. At least three sections were studied from each specimen in a blind fashion.

Organ water content The left kidney, spleen, and 5 cm of the proximal ileum and

distal colon, measured under a fixed tension, were quickly removed. Bowel segments were opened, and the contents were wiped. Samples were placed in aluminum desiccation trays and weighed when wet. They were desiccated at 50°C until the dry weight stabilized within 0.05%. The percentage of water content was estimated as follows:

[I-(Dry Weightwet Weight)] x 100

Mucosal mass analysis Ten-centimeter segments of distal small bowel and

proximal colon, measured under a fixed tension, were transected. They were opened longitudinally, cleaned of feces by gentle irrigation with water, and blotted dry. The mucosa was completely scraped with a glass slide for the measurement of nucleotide, RNA, DNA, and protein content. Mucosal nucleotide and RNA were measured in accordance with Munro & Fleek (14). DNA was analyzed by means of the technique of Scott et al. (15). The protein was dissolved in 1 mol/l NaOH and measured by the method of Lowry et al. (16).

Permeability measurement On the 7th day of experiment the rats were anesthetized

under ether. The left femoral vein was cannulated for administration of isotope. Intestinal permeability was mea- sured by the method of von Ritter et al. (17), as modified by Kanwar et al. (18). In brief, three loops (10cm each) of proximal small bowel, distal small bowel, and colon were isolated. Inflow and outflow tubes were fitted in them, and warm Tyrode solution (Sigma Chemical Co.) was perfused through the loop at a rate of 0.4 ml/min. The renal pedicles were tied to prevent excretion of "Cr-labeled ethylenedia- minetetraacetic acid (EDTA) by the kidneys.

"Cr-EDTA (Du Pont de Nemours & Co. Inc., Wilmington, Del., USA) was injected (150 pCi/kg) intravenously via the femoral vein. Twenty minutes were allowed for equilibration across the vasculature, and luminal perfusate was collected for three 10-min periods. At the end of the experiment the intestinal loops were removed, rinsed, and weighed. "Cr-EDTA activity in plasma and luminal

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

560 Y. Ma0 et al.

perfusates was measured with a Cobra Series Auto-Gamma spectrometer (Packard Instrument Co, Meriden, Conn., USA). The plasma-to-lumen clearance of "Cr-EDTA was calculated as follows

Cpm, x PR x 100 Clearance =

Cpm,l x '

where clearance of %-EDTA is given in milliliters per minute per 100 g; cpm, is counts per minute per milliliter of perfusate; PR is the perfusion rate in milliliters per minute; cpmPl is counts per minute per milliliter of plasma; and wt is weight in grams of the intestinal segment.

Myeloperoxidase (MPO) Eight-centimeter segments of ileum and colon, measured

under fixed tension, were obtained for measuring MPO activity with the method described by Grisham et al. (19). One unit of activity was defined as the amount of enzyme present that produced 1.0 change in absorbance per minute at 37°C and was expressed as unit/gram tissue.

Intestinal microflora Samples from the ileum and cecum were collected in sterile

tubes containing 5 ml of transport medium (20) and weighed. Tubes were sonicated for 5 min and shaken for 2 min for optimal dispersion. They were serially diluted in 10 ml of the same transport medium five times. Total aerobic bacterial counts were made by placing 1 ml of the sample on brain heart infusion agar (BHI, Difco, Detroit, Mich., USA) and incubated at 37°C for 3 days. Total anaerobic counts were also made in the same manner incubated under anaerobic conditions (Gas Pack, Becton Dickinson Microbiology Systems, Cockeysville, Md., USA) for 3 days. After that the colony-forming units on each plate were counted and corrected to the weight of the original sample. Further microfloral determination was made for gram-negative anae- robes, lactobacilli, and Enterobacteriaceae, using appropriate selective culture media and techniques in accordance with Johansson et al. (20).

Bacterial translocation Under aseptic conditions samples from the caudate lobe of

the liver, spleen, and mesenteric lymph nodes (MLNs) were placed in sterile tubes containing 5 ml transport medium and were weighed. Samples were homogenized, placed in an ultrasonic bath (Millipore, Sweden) for 5 min, and shaken for 2 more min. One milliliter of aortic blood was also collected. Total aerobic and anaerobic bacterial counts were measured as described above and were corrected to the gram of tissue or milliliter of blood.

Endotoxin measurement Blood samples were placed in endotoxin-free EndoTubes

(Chromogenix AB, Molndal, Sweden) and were centrifuged

Fig. 1. Colonic mucosa (hematoxylin and eosin; magrufication, x100). la. The mucosa from a rat 72 h after methotrexate (MTX) administration shows epithelial loss, necrosis, and inflammation. lb. The mucosa from a rat with pectin supplementation 72 h after h4TX administration shows significantly improved mucosal architecture.

at 4 "C for 15 min (2200 g). Plasma endotoxin was measured with the chromogenic limulus assay (Coatest endotoxin, Kabi Diagnostica, Molndal, Sweden) (21). The detection limit was 0.25 endotoxic units/ml plasma (EU/ml).

Statistical analysis Data are presented as the mean f SEM. The incidence of

bacterial translocation was analyzed using Fisher's exact probability test. The rest of the parametric data were statistically analyzed by two-way analysis of variance and Student's t tests when indicated. A p value less than 0.05 was considered significant.

RESULTS

Clinical features In three experiments the rats in the various groups grew

similarly during the study. Forty-eight hours after the administration of MTX, lethargy and onset of diarrhea were noted in enterocolitis rats. MTX-pectin rats had obviously less severity of lethargy and diarrhea than MTX-ED rats. Control

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

Pectin in EnterocoZitis 561

Table II. Kidney, spleen, ileum, and colon water content in different groups

Control Methotrexate

ED Pectin ED Pectin

Kidney 68.7 ? 1.7 70.3 f 1.1 82.4 f 0.6. 74.2 f 1.43 Spleen 77.6 2 0.6 78.8 f 0.8 88.7 f 0.87 81.4 2 1.0 Reurn 80.0 2 0.5 79.0 ? 0.7 96.8 f 0.9t 83.6 f 1.33 Colon 77.6 f 0.8 77.1 2 1.1 91.9? l.lt 79.7 f 1.7

The figures are given as percentage of organ weight. ED = elemental diet; MTX = rnethotrexate. Values expressed as mean ? SEM.

* p < 0.01 versus MTX-pectin and p < 0.001 versus control-ED. t p < 0.001 versus MTX-pectin andp < 0.001 versus control-ED. 3 p < 0.05 versus control-pectin.

animals remained alert, and no diarrhea was observed throughout the study. All animals except control-pectin rats had lost weight 3 days after MTX administration. MTX-ED (10.9 2 0.5% decrease) and MTX-pectin (5.9 f 0.4% de- crease) rats lost a significantly greater percentage of body weight than their respective control-ED (3.1 2 0.2% de- crease) and control-pectin (1.2 2 0.4% increase) rats. How- ever, MTX-pectin rats lost a significantly smaller percentage of body weight than the MTX-ED rats.

Histopathologic study On gross examination the small bowel and colon appeared

severely edematous and congestive in MTX-induced entero- colitis rats and appeared normal in both control groups. Microscopic examination showed a significant loss of the mucosal villous tips along with severe inflammation and ulceration in MTX-induced enterocolitis (Fig. la), and a significantly improved mucosa in the enterocolitis rats treated with pectin (Fig. lb). The control-pectin and control-ED groups had a normal and a slightly atrophic mucosal histology, respectively.

Organ water content MTX-induced enterocolitis rats had significantly higher

water content in all of the visceral tissues studied, compared with both control groups (Table 11). However, enterocolitis rats supplemented with dietary pectin had significantly less organ water content than those receiving only elemental diet.

Mucosal protein content MTX administration resulted in a significant ileal and

colonic mucosal protein loss in both dietary groups. The h4TX-pectin animals, however, lost significantly less mucosal protein than did those in MTX-ED group (Table 111). Ileal and colonic mucosal protein in the control-pectin group was also significantly higher than that in control-ED group.

Mucosal DNA, RNA, nucleotide content Both ileal and colonic mucosal DNA, RNA, and nucleotide

in MTX-ED and MTX-pectin groups were significantly less

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

562 Y. Ma0 et al.

Fig. 2. Mucosal permeability assessed as 51Cr-labeled ethylenediaminetetraacetic acid (EDTA) clearance for proximal, distal small bowel, and colon. The measurement was done 72 h after either methotrexate (MTX) or saline administration in the different dietary-supplemented groups. Values expressed as mean 2 SEM. ED = elemental diet. * p < 0.01 versus control-ED. + p < 0.05 versus MTX-ED, p < 0.001 versus control-pectin. + + p < 0.01 versus MTX-ED, p < 0.05 versus control-pectin. $ p < 0.05 versus control-pectin.

than in their respective control groups (Table 111). Pectin supplementation also resulted in a significant increase in ileal and colonic mucosal DNA, RNA, and nucleotide content in both control and MTX groups when compared with non- pectin supplementation, except in ileal mucosal DNA content between control-ED and control-pectin groups.

Intestinal permeability MTX administration resulted in a significant increase in

permeability of the proximal and distal small bowel and colon segments in both dietary groups compared with saline injection (Fig. 2). Furthermore, the alteration of permeability after MTX injection was greater in the distal small bowel and

Fig. 3. Ileal and colonic myeloperoxidase activity in rats fed either a pectin-supplemented elemental diet or a pectin-free elemental diet (ED). Intestines obtained 72 h after either methotrexate (MTX) or saline administration. Values expressed as means 2 SEM of unit/gram (U/g) tissue. * p < 0.001 versus control- ED. $ p < 0.05 versus MTX-ED and p < 0.05 versus control-pectin. $$ p < 0.05 versus MTX-ED and p < 0.001 versus control-pectin.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

Pectin in Enterocolitis 563

3 P

Fig. 4. Different ileal bacteria counts from rats fed either a pectin-supplemented elemental diet or elemental diet (ED) alone. Intestinal content was taken 72 h after either saline or methotrexate (MTX) administration. Values expressed as means t SEM. CFU = colony-forming unit. * p < 0.01 versus MTX- ED. * * p < 0.001 versus MTX-ED. " p < 0.001 versus control-ED. ' p < 0.01 versus normal. ** p < 0.001 versus normal. $ p < 0.05 versus control-ED. I$ p < 0.001 versus control-ED.

Fig. 5. Different cecal bacteria counts from rats fed either a pectin-supplemented elemental diet or an elemental diet (ED) alone. Intestinal content was taken 72 h after either saline or methotrexate (MTX) administration. Values expressed as means 2 SEM. CFU = colony-forming unit. * p < 0.01 versus MTX- ED. * * p < 0.001 versus MTX-ED. ' p < 0.01 versus control-ED. + + p < 0.001 versus control-ED. ' p < 0.01 versus normal. '* p < 0.001 versus normal. 3 p < 0.05 versus control-ED. $$ p < 0.001 versus control-ED.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

564 Y. Ma0 et al.

Table IV. Incidence of bacterial translocation in different organs

Control Methotrexate

ED Pectin ED Pectin Normal

Blood 2f6 116 818t 4/8* 016% MLNs 3f6 316 818 618 0164

718 Of@ 8f8 618 0/6((

Liver 3/6 316 Spleen 4f6 316 818

ED = elemental diet; MTX = methotrexate; MLNs = mesenteric

p < 0.05 versus MTX-ED. t p < 0.05 versus control-ED. $ p < 0.001 versus MTX-ED. 8 p < 0.001 versus MTX-ED and p < 0.01 versus MTX-pectin. 11 p < 0.001 versus MTX-ED, p < 0.01 versus MTX-pectin and

lymph nodes.

p < 0.05 versus control-ED by Fisher’s exact probability test.

colon than in the proximal small bowel. Pectin supplementa- tion significantly reversed the intestinal permeability in the distal small bowel and colon but not in the proximal small bowel. The greatest effect of pectin was seen in the colon. There was no statistically significant change in intestinal permeability between control-ED and control-pectin groups.

Myeloperoxidase MTX administration caused an eightfold and a fivefold

increase in MPO activity in ileum and colon, respectively (Fig. 3). Pectin supplementation significantly decreased MPO

levels in both the ileum and colon. No difference was found in control groups with different dietary supplementation.

Intestinal microflora There was a significant increase in ileal and cecal total

bacterial counts and subpopulations in all rats fed the elemental diet, compared with those on normal chow, except lactobacilli (Figs. 4 and 5). MTX administration resulted in a significant increase in bacterial counts in both the ileum and cecum. However, pectin supplementation resulted in a significant reduction of bacterial counts. The only exception was that there was no statistical significance in the lactobacilli content among all experimental groups.

Bacterial translocation Two of six control-ED rats and one of six control-pectin

rats had positive blood cultures, whereas all the eight MTX- ED animals had positive blood cultures (Table IV). MTX- pectin rats had significantly fewer (four of eight) positive blood cultures than the MTX-ED group. No positive blood culture was seen in the normal group.

Translocation of bacteria to the MLNs, liver, and spleen was also not seen in the normal group of rats but occurred in half of the animals in control-ED and control-pectin groups and in all of the MTX-ED rats. MTX-pectin rats had significantly lower magnitude of bacterial translocation to MLNs, liver, spleen, and blood than MTX-ED rats (Fig. 6).

Fig. 6. Effect of a pectin-supplemented elemental diet (ED) on the magnitude of bacterial translocation to mesenteric lymph nodes (MWs), liver, spleen, and aortic blood 3 days after either methotrexate (MTX) or saline administration. None of the rats fed rat chow in the normal group had positive bacteria culture (not shown in the figure). CFU = colony-forming unit. + p < 0.05 versus MTX-ED. + + p < 0.01 versus MTX-ED. * p < 0.05 versus control-ED. ** p < 0.001 versus control-ED. $ p < 0.01 versus MTX-ED, p < 0.01 versus control-pectin.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

Pectin in Enterocolitis 565

No statistical significance was seen in bacterial counts from tested tissues between two control groups.

Plasma endotoxin level Plasma endotoxin levels were significantly increased after

administration of MTX. However, the MTX-pectin group had a significantly lower endotoxin level than the MTX-ED group (0.38 2 0.03 versus 1.66 2 0.84 EU/ml). Plasma endotoxin values in the control-ED and control-pectin groups were in the undetectable range (<0.25 EU/ml).

DISCUSSION

In the present study administration of MTX to rats fed an elemental diet resulted in severe intestinal inflammation and injury, indicated by diarrhea, organ edema, histopathologic changes, and increased MPO activity. MTX also resulted in a reduction of mucosal proliferation shown by lowered mucosal protein, DNA, RNA, and nucleotide content, in both the small and large bowel. Gastrointestinal side effects are the commonest complications of MTX administration which may impose cessation of the therapy (22). MTX administration causes morphologic damage to the immature intestinal crypt cell. When mature cells are sloughed, the villus may not recover quickly, and the mucosa is then exposed to erosive pancreatic and biliary secretions, resulting in autodigestion (23). All of these effects are related to a reduction of mucosal mass and mucosal protein, DNA, and RNA content, and increased intestinal permeability (5,7,24).

This study has shown that pectin supplementation sig- nificantly reduces the degree of enterocolitis and improves the bowel nutritional status by reconstituting almost all of the mentioned variables. It has been shown that pectin has several benefits for the gastrointestinal tract such as enhancing intestinal architecture by stimulating mucosal cell prolifera- tion (9-1 l), protecting the intestinal surface by increasing the thickness of the unstirred water layer (25-27), increasing host enteroglucagon and disaccharidase levels, which are trophic effects on the gastrointestinal tract (12,28,29), and increas- ing lipase activity (30). Furthermore, pectin fermented by bacteria produces SCFAs, which are the preferred energy source of intestinal epithelial cells, especially in the colon (31,32). SCFAs promote sodium and water absorption, protect against alterations of electrolyte transport induced by bile acids, and stimulate colonic cell proliferation (13,33).

The mechanism of the healing in enterocolitis is not fully explained. However, we know that mucosal repair involves at least two processes. One process is the rapid migration of enterocytes from the crypts into the injured area of the mucosa. The second process is the replacement of the mucosa by cell replication (34). Pectin has been shown to increase DNA synthesis and increase the crypt cell production rate in gastrointestinal mucosa. Addition of pectin to the elemental diet may facilitate the mucosal healing from MTX-induced enterocolitis. The effect of pectin on reestablishment of the

mucosal mass was much greater in the colon than that in the small bowel in our study. This may be related to the fact that pectin is mainly fermented to SCFAs in the cecum and colon (35,36), and colonic cells depend largely on the luminal fermentation products. Thus, pectin, by providing luminal nutrients to the colon, plays a great beneficial role for colonic mucosa.

Administration of MTX caused severe mucosal damage, which greatly increased the clearance of "Cr-EDTA from blood to lumen in the proximal and distal small bowel and colon. Addition of pectin to the elemental diet could significantly reverse the extent of intestinal damage and help to reestablish the bowel physical barrier function, indicated by the reduced intestinal permeability in the distal small bowel and colon. However, in our study, the permeability of the proximal small bowel was not affected significantly by the administration of pectin. This may also indicate that the colonic and ileal mucosa get more benefit from fiber fermentation than the jejunal mucosa.

In agreement with other studies (1,37), our study showed that administration of a fiber-free elemental diet resulted in a significant increase in the number of all bacterial groups in both the small bowel and cecum, except lactobacilli. Administration of MTX caused even higher intestinal bacterial counts than the elemental diet. A suggestion is that the increase in the bacterial load after MTX administration reflects a decrease in the absorption of nutrition in the intestine. Thus, when the intestinal status deteriorates, more nutrients are left to the bacteria, and their numbers increase. Furthermore, it is the number of the more aggressive types of bacteria that prosper, such as Enterobacteriaceae. The advantage of dietary pectin supplementation to the elemental diet in the present study was the maintenance of normal gut microflora by significantly reversing intestinal microfloral disruption, protecting against more aggressive bacteria. The reestablishment of normal microflora may be a mediator in reducing intestinal inflammation (38) and enhancing the function of gastrointestinal surface protection system.

Administration of MTX to the rats fed the elemental diet resulted in a significant increase in rate and amount of bacterial translocation from the gut to the extraintestinal sites. Pectin supplementation significantly lowered the bacterial counts isolated from tissues and blood in enterocolitis rats. Recent studies (39,40) have suggested that bacterial trans- location, which is a sign of intestinal barrier failure, happens in severe clinical and experimental conditions. It is known that the gastrointestinal surface protection system, which is structurally complex and consists of the mucosal epithelial layer, surfactant, mucus, probiotic bacteria, and gut immune cells, normally is an efficient barrier that prevents invasion of microorganisms. Under certain conditions, such as thermal injury (41), endotoxin administration (M), and cytotoxic drug injection (7), intestinal bacteria may translocate to extra- intestinal sites. The clinical significance of bacterial translo- cation is controversial, but, together with increased passage of

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

566 Y. Ma0 et al.

endotoxin, it may trigger multiple organ failure in certain

the physical mucosal barrier and the disruption of intestinal

produced by parenteral nutrition following methotrexate. Circ Shock 1993;39:263-8.

dutmine administration on mortditv followine exoerbental

groups Of patients (39)' In OUT study the Of 3. Burke DJ, Averdy JC, Aoys E, Moss GS, Effect of route of

microflora caused by MTX injection may be reasons behind enteromlitis. J Parenteral Enteral Nu& 1990;14 suppi 1%. the bacterial banslocation observed. And the addition of 4. Nverdy JC, AOYS E, Moss GS. Parenteral nutrition results in

bacterial translocation from the gut and death following pectin resulted in decreased bacterial translocation not only chemotherapy. J parentera] Enteral Nut1 1990;14 suppl p8, by reducing mucosal injury and improving bowel mucosal 5. Bounous G. The use of elemental diets during cancer therapy. structure but also by maintaining the intestinal microecology Anticancer 1983;3:2965-7. 6. Hudson MJ, Borviellos SP, Hill MJ. Elemental diets and the

to the effect Of bacterial flora of the gastrointestinal tract. In: Russell RI, editor. fiber on decreasing bacterial translocation might also include Elemental diets. West Palm Beach (FL): CRC Press, 1981: improvements in immune responsiveness, induction of 126-43.

7. Fox AD, Kripke SA, DePaula J, Berman JM, Settle BG, trophic and toxin binding (39)7 which are Of great Rombeau JL. Effect of a glutamine-supplemented enteral diet on interest for further investigations. methotrexate-induced entermlitis. J Parenteral Enteral Nutr

the gut lumen to the 8. McAnena OJ, Harvey LP, Bonau RA, Daly JM. Alteration of systemic circulation is a well-recognized, yet poorly under- methowexate toxicity in by manipulation of dietary stood, process (40). Increased passage of endotoxin from the components. Gastroenterology 1987;92:354-60. lumen to he systemic circulation has been observed under 9. Rombeau JL, Rolandelli RH, Settle RG. Effect of liquid formula

diets supplemented with citrus pectin on intestinal structure and many stressful clinical and experimental conditions (4245). function. z Gasmoenterol 1989;27 suppl 2:27-30. In our study we detected a high level of circulating endotoxin 10. Demigne C, Levrat MA, Remesy C. Effects of feeding after administration of m. nis may explain the high fermentable carbohydrates on the cecal concentrations of

minerals and their fluxes between the cecum and blood plasma mortality rate observed in this model. Application of pectin in the rat. J Nutr 1989;119:163-30. significantly decreased the systemic endotoxin levels in 11. Tasman-Jones C, Owen RL, Jones AL. Semipurified dietary enterocolitis rats, which may also be mediated by the fiber and small bowel morphohy in rats. Dig Dis Sci 1982;

275 19-24. improvement of gut surface protection system. 12. Koruda MJ, Rolandelli RH, Settle RG, Rombeau JL. Small

Much attention has been focused on the efficacy of bowel disaccharidase activity in the rat is affected by intestinal antibiotics or free scavengers in preventing and resection and pectin feeding. Am J Clin Nutr 1988;47:448-53.

13. Rolandelli RH, Saul SH, Settle RG, Jacobs DO, Trerotola SO, treating the effects Of Iumina1 pemeabilib' to Rombeau JL. Comparison of parenteral nutrition and enteral microorganisms and their toxic products, but few studies have feeding with pectin in experimental colitis in the rat. Am J Clin investigated the effect of specific nutrients on prevention of 14. Munro HN, Fleek A. Recent developments in the measurements these phenomena. We have shown that the provision Of of nucleic acid in biological materials. Analyst 1966;91:78-88. pectin, a preferred fuel for the intestinal mucosa, in MTX- 15. Scott JE, Fraccastoro AP, Taft EB. Studies in histochemistry. I. induced enterowlitis model resulted in significantly improved Determination of nucleic acid in microgram amounts of tissue. J

Histochem Cytochem 1956;4:1-10. nutritional and morphometric status and decreased intestinal 16. hwry OH, Rosebrough NJ, F~~ a, Randall RG, protein inflammatory extent and bowel permeability. The effects of measurement with the Folin phenol reagent. J Biol Chem pectin on preventing bacterial translocation and endotoxemia 1951;193:265-75.

17. Von Ritter C, Grisham MB, Hollwarth M, Inauen W, Granger were associated not only with improved mucosal structure but DN. Neutrophil-deived oxidants mediate formylmethionyl- also with reestablished intestinal microecology. leucyl-phenylalanine-induced increases in mucosal permeability

in rats. Gastroenterology 1989;97:778-80. 18. Kanwar S, Wallace JL, Befus D, Kubes P. Nitric oxide synthesis

inhibition increases epithelial permeability via mast cells. Am J Physiol 1994;266:G222-9.

19. Grisharn MB, Benoit JN, Granger DN. Assessment of leucocyte

Methods Enzvmol 1990;186:729-42.

Other potential mechanisms

1988;12:33-31. The movement Of

1988;47:715-21.

ACKNOWLEDGEMENTS

The authors thank Dr' Inga Hagerstrand, Dept* Of involvement during ischemia and reperfusion of intestine, Lund University Hospital, Lund, for histopathologic exam- -

inations. This study was supported by grants from the Medical Research Council, Stockholm, grant B96-174- 11616-OlA, Lisa och Johan Gronbergs Stiftelse, Stockholm; and Medical Faculty, Craafordska Stiftelsen, Albert PHhls- sons Stiftelse, and Royal Physiographic Society, Lund University, Lund, Sweden.

REFERENCES

1. Shou J, Liebennan MD, Hofman K, Leon P, Redmond HP, Davies H, et al. Dietary manipulation of methotrexate-induced enteromlitis. J Parenteral Enteral Nutr 1991;15:307-12.

2. Zaloga GP, Roberts P, Black KW, Prielipp R. Gut bacterial translocation/dissemination explains the increased mortality

20. Johansson Mi, Molin G , Jeppsson B, Nobaek S, Ahrne S, Bengmark S. Administration of different Lactobacillus strains in fermented oatmeal soup: In vivo colonization of human intestinal mucosa and effect on the indigenous flora. Appl Environ Microbiol 199339: 15-20.

21. Sturk A, van Deventer SJH, Wortel CH, Levels JHM, Ten Cate JW, Buller HR, et al. Detection and clinical relevance of human endotoxemia. Z Med Lab Diagn 1992;31:147-58.

22. Rombeau JL. A review of the effects of glutamine-enriched diets on experimentally induced enteromlitis. J Parenteral Enteral Nutr 1990;14 Suppl 4:lOO-5.

23. Torosian MH, Buzby GP, Prest ME, Stein TP, Zinnser K, Mullen JL. Reduction of methotrexate toxicity with improved nutritional status. Surg Forum 1982;33:109-12.

24. Bjarnason I, Smethurst P, Levi AJ, Peters TJ. Intestinal permeability to "Cr-EDTA in rats with experimentally induced enteropathy. Gut 1985;26:579-85.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.

Pectin in Enterocolitis 567

25. Huang JD. Role of unstirred water layer in the exsorption of quinidine. J Pharm Pharmacol 1990;42:435-7.

26. Sugiyama K, Bamba T, Nakajo S, Hosoda S. A study on the mechanism of indomethacin induced intestinal ulcer-from a view point of unstirred water layer. Nippon Shokakibyo Gakkai Zasshi 1991;263&43.

27. Kohen R, Shadmi V, Kakunda A, Rubinstein A. Prevention of oxidative damage in the rat jejunal mucosa by pectin. Br J Nutr 1993;69:789-800.

28. Chun W, Bamba T, Hosoda S. Effect of pectin, a soluble dietary fiber, on functional and morphological parameters of the small intestine in rats. Digestion 1989;42:22-9.

29. Bamba T, Chun W, Nakajo S, Hosoda S. Effect of pectin on formyl methionyl-leucyl-phenylalanine-induced intestinal mu- cosa of rat. J Nutr Sci Vitamin01 Tokyo 1992;38:197-201.

30. Forman LP, Schneeman BO. Effect of dietary pectin and fat on digestive enzyme activities in rat [abstract]. Fed Proc 1980; 39:785A.

31. Roediger WEW. Role of anaerobic bacteria in the metabolic wellfare of the colonic mucosa in man. Gut 1980;21:793-8.

32. Rombeau JL, Kripke SA. Metabolic and intestinal effect of short-chain fatty acids. J Parenteral Enteral Nutr 1990;14:

33. Sakata T, Von Engelhardt W. Stirnulatory effect of short chain fatty acids on the epithelial cell proliferation in rat large intestine. Comp Biochem Physiol [A] 1983;47:459-62.

34. Luck MS, Bass P. Effect of epidermal growth factor on experimental colitis in rat. J Pharmacol Exp Ther 1993;264: 984-90.

181 5-55.

35. Silk DBA. Fiber and enteral nutrition. Gut 1989;30:24&64. 36. Armstrong EF, Eastwood MA, Edwaeds CA, Brydon WG. The

effect of weaning diet on the subsequent colonic metabolism of dietary fiber in the adult rat. Br J Nutr 1992;68:741-51.

37. AJverdy JC, Aoys E, Moss GS. Effect of commercially available chemically defined liquid diets on the intestinal microflora and bacterial translocation from the gut. J Parenteral Enteral Nutr 1990;14:1-6.

38. Firsner JB, Shorter RG. Recent developments in nonspecific inflammatory bowel disease. N Engl J Med 1982;306:83745.

39. Frankel W, Zhang W, Singh A, Bain A, Satchithanandarn S, Klurfeld D, et al. J. Fiber: effect on bacterial translocation and intestinal mucin content. World J Surg 1995;19:144-9.

40. Deitch EA, Berg R, Specian R. Endotoxin promotes the translocation of bacteria from the gut. Arch Surg 1987;122:

41. Maejima K, Deitch EA, Berg R. Promotion by bum stress of the translocation of bacteria from the gastrointestinal tracts of mice. Arch Surg 1984;119:166-72.

42. Gardiner KR, Anderson NH, McCaigue MD, Ersin PJ, Halliday MI, Rowlands BJ. Absorbents as antiendotoxin agents in experimental colitis. Gut 1993;34:51-5.

43. Tai C. Clinical and experimental endotoxaemia related to the intestinal ischemia. Detection of endotoxin by means of radioimmunoassay. Jpn J Med Sci Biol 1974;27: 11 1-4.

44. Gathirarn P, Wells MT, Brock-Utne JG, Wessels BC, Gaffin SL. Prevention of endotoxaemia by non-absorbable antibiotics in heat stress. J Clin Pathol 1987;40:1364-8.

45. Prytz H, Holst-Christensen J, Komer B, Lieher H. Portal venous and systemic endotoxaemia in patients without liver disease and systemic endotoxaemia in patients with cirrhosis. Sand J Gastroenterol 1976;11:857-63.

185-90.

Received 20 September 1995 Accepted 16 November 1995

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

Col

lege

Lon

don

For

pers

onal

use

onl

y.