- Email: [email protected]
Phytohemagglutinin derived from red kidney bean (Phaseolus vulgaris): A cause for intestinal malabsorption associated with bacterial overgrowth in the rat
GASTROENTEROLOGY
1983:84:506-15
Phytohemagglutinin Derived From Red Kidney Bean (Phaseolus vulgaris): A Cause for Intestinal Malabsorption Associated With Bacterial Overgrowth in the Rat J, G. BANWELL, D. H. BOLDT, J. MEYERS, F. L. WEBER the technical assistance of B. MILLER and R. HOWARD Division of Gastroenterology; Department Medicine, Lexington, Kentucky
of Medicine,
Plant lectins or carbohydrate binding proteins interact with membrane receptors on cellular surfaces but their antinutritional efiects are poorly defined. Studies were conducted to determine the effects of phytohemaglutinin, a lectin derived from raw red kidney bean (Phaseolus vulgaris), on small intestinal absorptive function and morphology, and on the intestinal microflora. Phytohemagglutinin was isolated in purified form by thyroglobulin-sepharose 4B affinity chromatography. Red kidney bean and (6% and 0.576, respectively, of phytohemagglutinin dietary protein) were fed in a purified casein diet to weanling rats for up to 21 days. Weight loss, associated with malabsorption of lipid, nitrogen, and vitamin B12, developed in comparison with animals pair-fed isonitrogenous casein diets. Antinutritional effects of red kidney bean were reversible on reinstitution of a purified casein diet. An increase in bacterial colonization of the jejunum and ileum occurred in red kidney bean- and phytohemagglutinin-fed animals. When antibiotics were included in Received January 5, 1982. Accepted September 29, 1982. Address requests for reprints to: J. G. Banwell, M.D., Division of Gastroenterology and Clinical Nutrition, Department of Medicine, Case Western Reserve University, Cleveland, Ohio 44106. The research was supported in part by Grant AM 25529 from the National Institutes of Arthritis, Metabolism, and Digestive Disease. This work was presented in part at the American Gastroenterology Meeting, New York, New York, May 1981, and in abstract form in GASTROENTEROLOGY 1981;80:1103. Ralph Giannella, M.D., provided much helpful advice in preparation for the microbiologic studies that were carried out with the assistance of Ms. Peggy Pendley. The assistance of Joyce Cairns and Cheryl Inman in the preparation of this manuscript is gratefully acknowledged. 0 1983 by the American Gastroenterological Association 0016-5085/83/030506-10$03.00
University
Jr., with
of Kentucky, College of
the diet, malabsorption of [3H]triolein and 57Covitamin B12 in red kidney bean-fed animals was partially reversed and, in germ-free animals, purified phytohemagglutinin had no demonstrable antinutritional effect. Mucosal disaccharidase activity was reduced in red kidney bean- and phytohemagglutinin-fed animals, but intestinal mucosal morphology was unchanged. Dietary administration of phytohemagglutinin, alone or as a component of red kidney bean, caused intestinal dysfunction, which was associated with, and dependent upon, small intestinal bacterial overgrowth. Adherence of enteric bacteria to the mucosal surface was enhanced by phytohemagglutinin which may have facilitated small intestinal bacterial overgrowth. Many of the legumes (pulses and beans) that are of major importance as sources of dietary protein for animals and humans contain antinutritional substances [trypsin-inhibitory substances and plant lectins (l-4)]. Impaired growth, weight loss, diarrhea, and ultimately death may occur with feeding of a variety of raw beans to animals (5-7). These deleterious antinutritional effects may be reduced by heat treatment (8). Although the lethal effects of some beans such as the castor bean (Ricinus communis) have been recorded frequently and their biochemical action studied in detail (9), the antinutritional effects of other plant lectins or carbohydrate-binding proteins remain poorly defined. These have potent biological effects (lo),including erythrocyte agglutination (ll),mitogenicity for lymphocytes (121,and inhibition of tumor cell migration (13). Red kidney bean lectins (phytohemagglutinin) comprise a family of five glycoproteins, each con-
March
RED KIDNEY
1983
taining four subunits of similar molecular weight (~32,000 daltons) held together by noncovalent bonds. The individual isolectins contain various proportions of the L (mitogenic) and E (erythro-agglutinating) subunits (14,15). Jaffe (4,6)postulated that the toxicity of phytohemagglutinin (PHA), the major antinutritional factor in the red kidney bean (RKB) (Phaseolus vulgaris), may be related to its interreaction with receptor sites on the surface of intestinal cells, which causes a nonspecific interference with absorption of nutrients. The antinutriand other lectins (18), tional effect of PHA (16,17) however, depends on the presence of intestinal bacteria: in the germ-free state, toxic effects of lectins are ameliorated. These studies were performed in weanling rats to examine the alterations caused by crude RKB or purified PHA on intestinal absorptive function, the small intestinal microflora, and mucosal morphology.
BEAN
LECTINS
AND
MALABSORPTION
507
Diet All diets were constructed from Teklab purified diets (Teklab, Madison, Wise.), free from complex carbohydrates and fiber. The standard control diet was purified diet AIN-76, containing 6% fat and 23.5% protein, including all essential minerals and vitamins. Diets of lower protein content were formulated utilizing a protein-free diet powder to which varying amounts of casein were added. Homogenized crude RKB powder or purified PHA was incorporated in the test diets so that they were isonitrogenous with control diets (Table 1). Control and experimental animals were paired to ensure comparable intake of nutrients. Food consumption was determined daily. All metabolic studies were conducted for 7-day periods after an initial 3-day equilibration period on a dietary regimen. Dietary, fecal, and urine nitrogen were measured with a Coleman Nitrogen Analyser (Coleman Instruments, Perkin Elmer Company, Chicago, Ill.). The RKB diet fed to germ-free animals was irradiated (4.5 Mrads) to ensure sterility before use. Irradiation caused no change in the erythroagglutinating titer of a saline extract of the RKB diet, and irradiated RKB diets caused the same magnitude of weight loss as nonirradiated diets.
Materials and Methods Animals The outbred Sprague-Dawley male rat (Sprague Dawley, Indianapolis, Ind.) was used for all experiments. Batches of 6-12 animals, weighing 40-80 g each, were housed individually in mesh bottom metabolic cages. Rats were maintained at a constant temperature with a 12-h light-dark cycle. Funnels with screen baffles enabled collection of urine separate from feces. Feces were collected on wire screens, urine under mineral oil. Animals were weighed daily. Water was allowed ad libitum. Germ-free animals were introduced into isolators on arrival. Monitoring of germ-free status was carried out at J-day intervals, and by culture of tissues at the time of death. Table
1. Cumulative Containing
Dietary
Regimens
In general, the diets we used were comparable to those employed by animal nutritionists (5-7,16) who have studied the antinutritional effects of RKB or fractions containing purified PHA. Diets with low or normal protein content were used due to the practical difficulty of extracting purified PHA in sufficient quantities for feeding experiments. For instance, when PHA was incorporated in diets as 0.5% dietary protein, a total of 3-4 g were required for each metabolic balance study which represented a yield from five to six affinity columns.
Weight Change and Nitrogen Excretion for Groups of Weanling Rats Fed Isonitrogenous Purified Casein, Raw Red Kidney Bean Extract, or Purified Phytohemagglutinin
Dietary Group”
Dietary
regimen
% N,
N, intake (m&at
day)
Dietary consumption (g/rat. day]
Mean weight change (g/rat 7 day]
Diets
Nitrogen excretion (mgirat day) Fecal
Urine
5.2 + 0.5 36.3 + 2.0"
24.7 2 1.8 13.1 + 2.2'
A
C. T.
Casein protein Casein protein + RKB (6%)
(10%) (4%)
1.37 1.41
53.2 t 4.7 57.8 + 3.7
3.9 r 0.3 4.1 +- 0.2
+1.7 + 1.1 -3.9 i- 1.0"
B.
C. T.
Casein protein Casein protein t RKB (6%)
(20%) (14%,)
3.09 3.18
222.3 + 14.3 237.1 ? 16.2
7.2 ? 0.5 7.5 I 0.5
f11.5 t 0.4 f5.3 + 1.1"
13.5 i- 1.4 70.1 -+ 3.9"
100.7 + 7.2 64.9 + 2.8'
c
c. T.
Casein protein Casein protein t 0.5% PHA
(5%) (4.5%)
0.85 0.99
33.2 t 3.0 37.2 f 1.9
3.9 t- 0.4 3.8 t- 0.9
-2.1 k 1.3 -5.9 + 0.8'
7.35 t 0.4 15.5 t 1.6"
14.1 t 0.9 9.6 2 0.9"
D
C.”
Casein (10%) (irradiated] Germ-free animals
1.20
60.2 2 2.6
5.0 2 0.2
-5.0 + 1.0
13.2 + 1.8
24.6 t 1.6
T.
Casein protein (4%) + RKB (6%) (irradiated) Germ-free animals
1.40
91.3 k 3.3
6.5 t 0.2
PO.7 i 2.0
27.8 t 2.2'~
38.2 2 1.9"
I’n = 6 for each group: T = test; C = control. 1 Animals failed to consume total diet.
” p < 0.001 vs. control
group.
’ p < 0.005 vs.control
group.
I’p < 0.025.I’p c 0.05.’p < 0.1.
508 BANWELL ET AL.
Dietary groups studied
GASTROENTEROLOGY Vol. 84, No. 3
were
Group A C. Casein protein (10%) T. Casein protein (4%) + RKB (6%) This diet was utilized to define effects of homogenized RKB, containing PHA, on small intestinal function. Group B C. Casein protein (20%) T. Casein protein (14%) + RKB (6%) A study to determine effects of RKB in a diet containing normal dietary quantities of protein. Group C C. Casein protein (5%) T. Casein protein (4.5%) + purified PHA (0.5%) A diet to study effects of PHA alone on small intestinal function. Group D C. Casein protein (10%) (irradiated) fed to germ-free animals T. Casein protein (4%) + RKB (6%) (irradiated) fed to germ-free animals A dietary study to determine the effect of RKB in a germfree environment.
Preparation
of Phytohemagglutinin
Red kidney beans (Phaseolus vulgaris) were obtained from a single source (Laurelbrook Farms, Be1 Air, Md.). A saline extract of RKB was made by homogenizing 100-150 g of whole beans in a Waring blender for 3 min at 20,000 rpm. The fine powder obtained was added to 500 ml phosphate-buffered saline (PBS) and stirred overnight at 4°C. This suspension was filtered through cheesecloth, and the filtrate separated by centrifugation at 39,000 g for 20 min. Purified PHA was extracted utilizing thyroglobulin-sepharose 4B affinity chromatography by the method of Felsted and coworkers (14). The single discrete elution peak yielded 0.55-0.85 g total protein (PHA) from lo-15 g of crude bean protein saline extract applied to the column. Polyacrylamide gel electrophoresis, carried out on 7% acrylamide, 0.5% bisacrylamide gel at 4°C according to the method of Reisfeld et al. (20), revealed five separate phytohemagglutinin isolectins with different and distinct mobilities. Columns were reused repetitively after rejuvenation of thyroglobulin sepharose. The PHA yields from several column preparations were combined, analyzed for erythroagglutinating activity and protein content, and mixed with purified casein to provide the requisite PHA diet.
Agglutination
by the reciprocal of the serial dilution immediately preceding the first visible cell button of nonagglutinating cells.
57Co-Vitamin Blz Absorption One-half micro Curie of “Co-vitamin Blz (sp act 1 &i/l pg) (Amersham-Searle, Chicago, Ill.) was administered as a 0.5-ml bolus by intragastric tube to fasting animals with or without the addition of rat intrinsic factor (IF) prepared by the method of Welkos et al. (21). The cobalamin binding capacity of the IF preparation was 18.5 ng cobalamin/ml rat gastric mucosal extract (assayed by a charcoal binding radioassay method in the laboratory of Dr. P. Toskes, Gainesville, Fla.). Animals were counted in a whole-body gamma counter immediately after administration of 57Co-vitamm Blz. They were subsequently recounted 1 wk later and absorption was measured as the percentage of retained radioactivity at that time. Studies to define the effects of RKB or PHA were conducted after a lo-day exposure to the diet. A separate set of studies (see Figure 3B) examined the effects of RKB, before and after cessation of RKB intake, on vitamin Blz absorption measured on day 10, day 20 (6 days after cessation of RKB), and day 26.
[%]OJeic Acid and [14C]TrioJein Absorption Absorption of [3H]oleic acid and [14C]trioJein was studied in separate experiments by the method of Rodgers, Fondacaro, and Kot (22). Lipid emulsions were prepared by first sonicating 4 mM of sodium taurocholate and 3 mM of phospholipid (pig liver phosphatidylcholine, Calbiothem, San Diego, Calif.) in 50 ml of normal saline. The bile salt-phospholipid mixture was added to a beaker containing either radioactive lipid and the mixture was resonicated. Four milliliters of the mixture which contained 10 &i was administered by gavage to fasting animals exposed to RKB or control diets (group A) for 10 days. To determine [3H]oleic acid excretion, 96-h collections of total fecal output were carried out: [‘4C]triolein absorption was determined by killing animals at 5 h, and, after ligature of stomach, small intestine, cecum, and colon, segments were removed and contents washed into separate beakers. Fecal material containing [3H]oleic acid or washings from the intestinal tract segments containing [‘4C]triolein was saponified with 33% KOH solution and boiled with 95% alcohol under reflux. Lipid was then extracted with petroleum ether. An aliquot of the petroleum ether phase was counted by scintillation spectrometry. Radioactive lipids in the gastrointestinal tract and feces were considered unabsorbed.
Assay
Agglutination assays were carried out utilizing red blood cells from a single group 0 donor. A 4% suspension of washed red cells was made and agglutination assays were carried out using microtiter plates and serial dilution with a standardized pipette [Cooke Laboratory Products, Alexandria, Va.). The erythroagglutinin titer was defined
Intestinal
Morphology
Tissue from control and test animals was collected from jejunum, ileum, cecum, and colon immediately after death (within 60 s). Specimens were processed for light microscopy (LM) and transmission electron microscopy (TEM). Sections stained with hematoxylin and eosin were
RED KIDNEY BEAN LECTINS AND MALABSORPTION
March 1983
examined by LM. Specimens for TEM were fixed in 1.75% glutaraldehyde and 0.1% sodium cacodylate and buffered for 1-2 wk before subsequent fixation and examination. Histologic tissue was coded and examined without knowledge of the dietary regimen to which animals had been exposed.
Brush
Border
Enzymes
Activities of lactase, sucrase, and maltase were measured in jejunal and ileal homogenates by the method of Dahlquist (23). One unit of enzyme activity is the amount that hydrolyzes 1 mmol of substrate per milligram mucosal protein.
Immunodiffusion Phytohemagglutinin
Assay
Bacteriological
Studies
All animals were fasted 18 h before death by CO2 narcosis. A lo-cm length of intestine distal to the ligament of Treitz, and a lo-cm segment of ileum, 5 cm proximal to the ileocecal valve, were immediately removed and transferred to prewashed sterile bottles containing PBS. Each segment was opened with sterile scissors and washed with four separate changes of PBS (24). The tissue was then weighed and ground with a Vortis homogenizer in 3.0 ml deoxygenated broth medium at 45,000 rpm for IO s.One milliliter of the first wash, of the fourth wash, and of the intestinal homogenate were mixed with 9 ml of deoxygenated broth and serially diluted to lo-' by ten-fold tube dilutions. One-tenth milliliter of the appropriate dilution was placed on the following media: Aerobic incubations: Blood agar for total aerobes, MacConkeys media for enterobacteriacae, mannitol salt media for staphylococci, Sabauraud’s media for fungi, and Mitis salivarius for streptococci were used. Anaerobic incubation (72-96 h): Total anaerobes were cultured on blood agar and microaerophilic bacilli were cultured in COZ for 72 h on Rogosa SL agar. Bacteria were identified by gross morphology, using a gram stain, and by subculturing and selected biochemical tests. Organisms from anaerobic plates were confirmed as being anaerobic by failure to grow on aerobic subculture. The number of specific organisms was expressed as the logarithm to the base 10 of the mean colony counts on ulates containing l-1000 oreanisms 1251.
of Antibiotics
4% + Red Kidney
Bean
into the Casein
6% Diet
Antibiotics were incorporated into the casein 4% diet containing 6% RKB to study whether “‘Co-vitamin Blz and [‘4C]triolein malabsorption could be reversed. Metronidazole (Searle Co., Chicago, Ill.) and kanamycin (Kantrex, Bristol Laboratories, Syracuse, N.Y.) were carefully dispersed in the feed. The daily dosage of metronidazole (7.5 * 0.3 mg/rat per day) and kanamycin (7.5 2 0.3 mgirat per day) in the diet were similar to levels shown to cause a reduction in both aerobic and anaerobic intestinal bacteria in rats with and without surgically constructed blind loops (21). Control animals were pair fed the same’diet without antibiotics. Statistical
for Fecal
To determine whether PHA was detectable in fecal pellets of PHA-fed animals, pooled g-day fecal collections were extracted in saline and PHA isolated by thyroglobulin sepharose affinity chromatography. The glycine buffer eluate was concentrated on a PM-10 membrane. After dialysis, immunodiffusion plates were run utilizing purified PHA, the fecal PHA extract, and purified antibodies to the E and L isolectins of PHA. Phytohemagglutinin antibody was obtained from Vector Laboratories Inc., Burlingame, Calif.
Intestinal
Incorporation
509
Testing
Results were analyzed with the Student’s t-test for independent variables and expressed as the mean 2 standard error (SEM).
Results In preliminary experiments, both weanling (40-80 g) and adult (~200 g) rats that were fed ad libitum diets containing either RKB or PHA consistently ate less food and grew less or lost weight during the 2-3 wk study period. They remained as active as their control littermates, but their fecal pellets were soft and unformed. Reinstitution of a control casein diet resulted in rapid reversal of weight loss to weight gain within 24 h. Death often supervened in weanling rats by the third or fourth week of dietary treatment with RKB or PHA. Animals fed a 0.5% PHA plus 4.5% casein protein diet ad libitum ate less of their diet (5.4 ? 0.2 g/day] than animals fed 5% casein (11.7 + 0.4 g/day, p < 0.001) during a lo-day study period and lost weight (PHAfed: -1.2 * 0.1 g/day, control-fed: +1.5 +- 0.2 g/day).
To normalize for the effect of variable dietary intake, all subsequent metabolic studies were performed with groups of weanling animals pair-fed isonitrogenous diets (Table 1). Group A diets, containing 6% RKB, caused significantly greater growth depression or weight loss than control diets (p < 0.001). Animals on a normal protein diet containing RKB (group B) grew better than those on a low protein intake, but less than their pair-fed controls (p < 0.001). Animals on a 4.5% casein protein diet + 0.5% PHA (group C) also lost more weight than those pair fed the control 5% casein diet. Germ-free animals (group D) fed 6% RKB lost less weight than the conventional animals in group A (p < 0.025). Nitrogen
urine
Excretion
Studies of nitrogen excretion in feces and during the dietary studies are also shown in
510
GASTROENTEROLOGY Vol. 84. No. 3
BANWELL ET AL.
loss in RKB-fed germ-free animals (group D) was less than that seen in their conventional counterparts (group A).
r
Lipid Absorption With RKB feeding, a greater percentage of radiolabeled lipid was recovered in the 96-h fecal collection than in controls (p < 0.01) (Figure 1). [‘4C]triolein recovery in the small bowel lumen, cecum, and colon 5 h after gastric administration was also greater in RKB-fed animals than controls but failed to reach statistical significance (p < 0.15). Malabsorption of radioactive lipid in rats on the RKB diet was partially reversed by simultaneous antibiotic administration. Figure
1. Rekovery 3H-oleic acid (left side) in feces and l“Ctriolein (right side) from intestinal contents after orogastric administration to weanling rats on test diet (hatched bar], control diet (clear bar), and test animals treated with antibiotics (stippled bar).
Table 1. Fecal nitrogen losses were increased fivefold to sevenfold in the RKB-fed groups (A and B) compared with animals fed isonitrogenous control casein diets (p < 0.001); PHA-fed animals exhibited a similar increase in fecal nitrogen (group C). Urinary nitrogen losses were, in general, slightly lower in RKB-fed animals than in controls. Fecal nitrogen
Disaccharidase
The mucosal activities of maltase, sucrase, and lactase were reduced in animals fed 4% casein protein + 6% RKB diets when compared with pairfed control animals (p < 0.001) (Figure 2). Activity was lower in the jejunum for all these enzymes and in the ileum for maltase and sucrase. Animals fed a diet containing 0.5% purified PHA + 4.5% casein diet had enzyme activity that was higher overall than the group fed RKB; however, jejunal enzyme activity
JEJUNUM
ILEUM
25
Figure 2. Disaccharidase bars]
animals
-
activity in jejunal (left side] and ileal tissue (right side) in control exposed
to RKB or PHA.
Activity
[clear
bars]
and test (hatched
or stippled
March
RED KIDNEY
1983
remained significantly reduced in comparison with the group receiving the control diet (p < 0.001). In the ileum, there was no significant difference in PHA-fed animals vs. controls. Vitamin Blz Absorption 57Co-Vitamin B12 absorption was impaired 714 days after institution of the RKB diet (20.7% * 1.4%) compared with control animals (61.2% +1.5%) (p > 0.001) (Figure 3A, B). Administration of rat gastric intrinsic factor caused no augmentation of 57C~-B12 uptake [RKB: 20.0% +- 2.3%; control: 60% & 1.8% (p < O.OOl)]. Impaired 57Co-vitamin Blz absorption was observed as early as 3 days after commencing the RKB diet (30.2% + 1.1%). Abnormal 57C~-B11 absorption was reversed by reinstitution of a 10% casein diet (65.8% +- 2.2%), and improvement in 57C~-B12 malabsorption towards normal (45.8% k 3.4%) was also observed when antibiotics were included in the diet (Figure 3A). Feeding of heat-treated RKB, which caused inactivation of hemagglutinating activity, resulted in a normal 57Co-vitamin B12 absorption (74.8% -+1.0%). Another group of animals was fed restricted amounts of a 5% casein diet to achieve weight loss comparable to that seen in the animals fed RKB (NZ intake, 0.85%,2.2mg Nz/ rat - day; weight lost -6.6 g/rat - i’ day). This starvation regimen alone did not (57Co-vitamin impair 57Co-vitamin B 12 absorption Blz absorption >60%).
c :
1OOr
;
go-
p
g
6070-
s
50-
cu ai
40-
.f E m .= > b
511
Findings
Microbiologic
Data
1. Influence of RKB diet on intestinal bacterial flora (Table 2). Quantitative culture of bacterial flora from jejunal and ileal segments revealed that RKB feeding caused bacterial proliferation. Counts of total aerobes were increased siginificantly in jejunum in the fourth wash of the mucosa. In ileum, counts were significantly increased in both wash 4 and tissue homogenates. Total anaerobes were also increased in the ileum in wash 4. Escherichia
A-A
6%
?? -.
10% CASEIN
R K B + 4%
PROTEIN
PROTEIN 9-m
i
“\
I
I
I
i
_... _.. .I.,,.:. :.:.:.:.: i ..i..... .:,.::. ..... i I :....:. ................. I ...:..... ................. :....:. ......... i i :.:.:.:.: . ....... .‘. i I .‘ ........‘ .... ..i..... i I ........ “=12 :i::. n=7 i i i
lo-
o
i 57
~ .
I Co-Vitamin
812
S’Co-Vitamin
B12
:,
5’Co-Vitamin
a12
ANTIa:OTICS
A Figure
MALABSORPTION
._.-.-.-..
20-
A
AND
Histologic evaluation by one of us (J.M.) of sections of jejunal and ileal tissue, which had been coded and randomized, revealed no effect of RKB or purified PHA on intestinal morphology. No change in villus morphology, crypt-cell mitotic activity, or round-cell infiltration of the lamina propria was noted. Tissue morphology on scanning EM was also normal in appearance. On TEM, morphological features of the microvilli, globlet cells, mitochrondia, and smooth and rough endoplasmic reticulum also showed no detectable difference from control tissue. In very occasional EM sections, ballooned or fused microvilli were observed in tissue that was otherwise normal. Histologic sections in the liver, pancreas, colon, and stomach showed no pathological changes on light microscopy.
E a-
30-
iiT
LECTINS
Morphological
c
8o
BEAN
3. A. “7Co-Vitamin B,, absorption in rats fed 4% casein + 6% RKB diet (hatched bars] or without added rat intrinsic factor. Antibiotic administration to two groups of animals fed in the bars (- . - . -)I caused “%o-vitamin B,, absorption to increase (stippled bar) compared without antibiotics (hatched bar). B. Absorption of “Co-vitamin B,, before, during the casein diet ( 1 ). Weight change on RKB diet (triangles) and after reinstitution of control
I2 DAY
24
B 10% casein diet (open bars) with or 4% casein + 6% RKB diet [depicted with those animals on an RKB diet RKB diet, and after return to a 10% diet (circles).
512
BANWELL
ET AL.
GASTROENTEROLOGY
cc& counts were greater in RKB-fed animals in jejunum and ileum in both wash 4 (p < 0.001) and tissue homogenates. 2. Influence of purified PHA on intestinal bacterial flora (Table 2). In PHA-treated animals, total aerobes in jejunal fluid, wash 4, and intestinal homogenate were significantly increased (p <
Vol. 84, No. 3
0.001).In the ileum, significant increases were seen with both washes 1 and 4. Particularly large increases in total anaerobes (p < 0.001)were observed in jejunal washings and homogenates, but only in wash 4 from the ileum. Escherichia coli were also increased in all jejunal specimens (p < 0.005) as well as in wash 4 from the ileum.
Table 2. Total, Aerobic, Anaerobic, and Escherichia coli Colony Counts of JejunaJ and Ileal Contents (Wash I), Fourth Mucosal Washing, and Mucosal Homogenate in Rats Fed Red Kidney Bean Diet vs. Control Diet and Purified Phytohemagglutinin Diet vs. Control Diet Bacterial colony count (log,, organism/g tissue wet wt) Wash 1 RKB diet vs. control diet (n = 6 for each group, test and controlj Total aerobes Jejunum 4.25 2 1.74 (NS) Test 2.78 t 1.24 Control Ileum 8.41 2 0.46 (NS) Test 6.79 + 0.70 Control Total anaerobes Jejunum 4.06 2 1.72 (NS) Test 2.81 -c 1.48 Control Ileum 7.79 2 0.76 (NS) Test 7.98 2 0.72 Control Escherichia coli Jejunum 1.04 2 1.56 (NS) Test 0.68 r 1.04 Control Ileum 6.55 ? 1.70 (NS) Test 4.84 k 1.27 Control
Purified PHA Diet vs. Control Diet (n = 7 for each group, test and control) Total aerobes Jejunum 5.96 t 1.20” Test 1.99 r 1.29 Control Ileum 8.40 2 0.27b Test 6.00 t 1.05 Control Total anaerobes Jejunum 7.12 2 0.65” Test 2.90 2 1.18 Control Ileum 8.29 f 0.23 (NS) Test 6.77 2 1.12 Control Escherichia coli Jejunum 6.48 _’0.83” Test 0.75 + 0.69 Control Ileum 5.15 2 1.19 (NS) Test 3.42 f 1.15 Control
Wash 4
Tissue homogenate
1.52 2 1.34” 0.00
1.38 ? 1.08 (NS) 1.80 f 0.83
7.41 t 0.13” 2.34 Ifr 1.05
8.09 2 0.15” 4.33 ‘- 1.29
1.85 2 1.37 (NS) 1.37 + 0.49
4.93 2 1.08 (NS) 3.18 f 0.88
7.08 k 0.33” 2.24 t 1.23
8.17 2 1.18 (NS) 4.56 + 1.34
0.57 ? 3.09” 0.0
3.51 5 1.59” 0.68 * 0.68
6.62 2 0.46” 1.12 c 0.12
6.94 t 0.27” 2.05 t 1.18
4.63 2 0.81” 0.0
7.02 t 0.40” 3.30 t 0.78
3.74 t 1.41’. 1.46 2 0.89
6.44 t 1.03 (NS) 5.85 t 1.17
5.19 2 0.43” 0.35 5 0.33
6.81 r 0.57” 2.89 k 0.73
5.14 + 1.26” 1.60 k 0.96
6.13 + 0.99 (NS) 5.53 * 0.99
3.52 2 1.30” 0.0
6.15 k 0.54” 0.44 2 0.41
4.11 +- 1.56” 0.80 2 0.74
4.17 2 1.59 (NS) 1.82 t 0.80
0 p i 0.001 vs. control group. b p < 0.05 vs. control group. I’p < 0.005 vs. control group. NS = not significant.
RED KIDNEY BEAN LECTINS AND MALABSORPTION
March 1983
Isolation and Purification Phytohemagglutinin
of Fecal
Evidence that PHA was excreted in the feces in animals fed either RKB or purified PHA was derived from two sources: (a) fecal saline extracts from RKB-fed animals (group B) were observed to cause agglutination in group 0 human red cells at a higher titer (1:lZ)than extracts from control feces (1:4),and (b) when a saline extract of pooled feces from 4 PHA-fed animals was applied to a sepharose thyroglobulin affinity column, a protein peak with the characteristics of PHA was eluted. Immunodiffusion analysis of this eluted material utilizing purified PHA-E and PHA-L antibodies revealed that both PHA-E and PHA-L components were present in the eluate and formed lines of identity with the purified mixture of PHA-E and PHA-L antigens. The total calculated fecal excretion of PHA in this study was 6.0 mg PHA, or 1.2% of the total dietary intake of PHA in animals fed 0.5% PHA in the diet.
Discussion Dietary administration of crude RKB or purified PHA caused malabsorption of nitrogen, lipid, and vitamin B12 in conventional but not germ-free weanling rats. Small intestinal bacterial colonization was associated with RKB and PHA feeding, and antibiotics partially reversed lipid and vitamin Blz Phytohemagglutinin, malabsorption. therefore, caused intestinal malabsorption and facilitated bacterial proliferation in the small bowel. Several previous studies (2-8,16,17,24) have classified features of the action of PHA. This work defines the pathophysiology of PHA-induced malabsorption and its dependence on bacterial adherence in the small intestine. The polymicrobial small intestinal bacterial overgrowth was associated with increased adherence of bacteria to the mucosal surface. An increased adherence of microbial organisms to the mucosal surface caused by PHA feeding may have provided the requisite conditions for proliferation of bacteria and development of small bowel bacterial overgrowth. Wilson et al. (26)have described overgrowth of E. coli in rats fed diets containing 10% raw RKB. No bacterial overgrowth occurred until 24 h after feeding began. When a bean diet, containing a reduced lectin content was fed, less prominent changes in coliform counts occurred. Untawale et al. (24)also observed greater attachment of E. coli to the intestinal mucosa of chickens fed raw RKB compared with those fed autoclaved RKB. Recently, Gibbons and Dankers (27) observed that lectin-like activity in aqueous extracts of several commonly ingested fruits affected attachment of Streptococcus mutans to saliva-coated hydroxyapatite beads.
513
Intestinal mucosal morphology after RKB and PHA exposure was essentially unaltered in this study, with the exception that, rarely, occasional microvilli were fused or club shaped. Identifiable histologic changes, as encountered in the small bowel bacterial overgrowth syndrome in experimental animals and humans, may have been absent due to the relatively short period of bacterial overgrowth in our study (14-21 days). Mucosal changes in small bowel bacterial overgrowth had usually been described with chronic bacterial colonization (25,28). Phytohemagglutinins from RKB (29) and other lectins (30)have also been noted to cause histologic lesions in the rat small intestine. Microvillous damage in the study of King et al. (29) was greatest in those animals fed strains of RKB that produced the higher concentrations of PHA lectins and that also caused growth inhibition. Strains of RKB that produced lower lectin concentrations resulted in growth inhibition without mucosal injury. Disaccharidase activity was reduced in RKB- and PHA-treated animals in the present study, which may be attributed to a direct effect on the microvillus membrane, but RKB and PHA concentrations may have been insufficient to cause histologic alterations in mucosal morphology. A variety of lectins (31-34)have been observed to bind to the mucosal surface membranes of both small and large intestine. Evidence for direct PHA adherence to the intestinal wall by immunofluorescent techniques has been obtained in rat studies (34) after crude RKB administration and in work in our own laboratory. Boedecker and Boldt observed that radiolabeled PHA demonstrated saturable binding to ileal brush border membranes and competition with IF-B12 receptor sites (35). Phytohemagglutinin might, therefore, directly alter absorption of nutrients by effects on transport across the intestinal microvillous membrane (36)as well as by changes in intestinal microflora. Phytohemagglutinin is only one of several antinutritional agents present in raw RKB (2):evidence would suggest that it accounts for ?35% of the total growth inhibitory fraction of one variety of Phaseolus vulgaris seed (7). It is not possible to calculate the exact amount of PHA available in RKB from this study since extraction of PHA by affinity chromatography was incomplete and, moreover, not all PHA in RKB may have been bioavailable in the intestine (37).
Specific saccharide inhibitable red blood cell agglutination is often demonstrable in commonly ingested foodstuffs, suggesting that intestinal exposure to dietary lectins in the human diet may be more widespread than supposed (38,39). The toxic effects of hemagglutinins in leguminous foodstuffs are generally eliminated by proper heat treatment (2,6,8),
514
BANWELL ET AL.
GASTROENTEROLOGY Vol.84, No. 3
but they might prevail if destruction is incomplete. Diarrhea1 and toxic effects in humans have been described with ingestion of uncooked kidney beans (40-42).
The present study has also shown that a portion of the purified PHA fed to these animals resisted proteolytic degradation during passage through the intestinal tract and was identifiable in the feces based on its similarity to purified PHA on affinity chromatography and gel immunodiffusion. Jaffe et al. also noted that fecal material from animals fed RKB demonstrated agglutinating activity (6). Resistance to intestinal proteolytic activity may, therefore, be an important feature of PHA (16,17)and other lectins (38,40)that permits their survival throughout the small and large intestine. Phytohemagglutinin and RKB ingestion caused weight loss and growth failure that occurred in association with diminished dietary caloric intake in the preliminary experiments conducted without pair feeding. Reduced caloric intake is known to be a major contributory factor in the weight loss associated with intestinal malabsorption (43).The pair feeding experiments indicated, however, that RKB and PHA were directly responsible for malabsorption. Furthermore, comparable weight loss induced in animals by a starvation diet did not of itself impair nitrogen or vitamin B12 absorption. The relevance of these findings to human disease must await further study. Adherence of bacteria to mucosal surfaces is an important factor in the pathogenicity of several animal (44) and human enteric pathogens (45,46).Lectins may enhance bacterial Lectins other than PHA, adherence (24,28,47,48). and even nonlectin dietary proteins, may have similar antinutritional effects in experimental animals (2-4).There are also several human intestinal diseases in which small intestinal dysfunction may be related to ingested dietary proteins or changes in small intestinal bacterial flora, such as cow’s_milk (49) and soy-protein intolerance (SO), chronic diarrhea1 disease of infancy (51),celiac disease (52),and tropical Sprue (53). Phytohemagglutinin-induced small intestinal bacterial overgrowth and malabsorption may provide a model for certain features of these diseases.
References Patwardhan VN. Pulses and beans in human Clin Nutr 1962;11:12-30. Liener I. Significance for humans of biologically in soybeans and other food legumes. J Am 1979;56:121-9. Rackis JJ. Biological and physiological factors Am Oil Chem Sot 1974;51:161A-74A. Jaffe WG. Hemagglutinins. Liener IE, ed. Toxic plant foodstuffs. New York: Academic Press,
nutrition.
Am J
active factors Oil Chem Sot in soybeans. constituents 1969.
J of
5. Kakade ML, Evans RJ. Growth inhibition of rats fed navy bean fractions. J Agric Food Chem 1965;13:450-2. 6. Jaffe WG, Vega Lette CL. Heat-labile growth-inhibiting factors in beans (Phaseolus vulgaris). J Nutr 1968;94:203-10. 7. Evans RJ, Pusztai A, Watt WB, Bauer DH. Isolation and properties of protein fractions from navy beans (Phaseolus vulgaris) which inhibit growth in rats. Biochim Biophys Acta 1973;303:175-84. 8. Muelenaere HJH. Effect of heat treatment on the haemogglutinating activity of legumes. Nature 1964;201:1029-30. 9. Olsnes S, Pihl A. Abrin, ricin, and their associated agglutinins. In: Cuatrecasas P, Greaves LC, eds. Receptors and recognition: the specificity and action of animal, bacterial and plant toxins. Vol 1. New York: Halsted Press, 1976:131-73. 10. Goldstein IJ, Hayes CE. The lectins: carbohydrate-binding proteins of plants and animals. Adv Carbohydr Chem Biothem 1978;35:127-340. 11. Egorin MJ, Bachur SM, Felsted RL. Leavitt RD, Bachur NR. Phaseolus vulgaris isolectin binding to human erythrocytes. J Biol Chem 1979;254:894-8. 12. Ling NR, Kay JE. Lymphocyte stimulation. New York: American Elsevier Publishing Co. Inc., 1975. RN, Sane1 FT, Felsted RL, Bachur NR. 13. Egorin MJ, Odders Inhibitory effects of phytohemagglutinin isolectins Lq and E, on L1210 cells. Cancer Res 1978;38:1677-87. of the phytohe14. Felsted R, Leavitt RD, Bachur NR. Purification magglutinin family of proteins from red kidney beans (Phaseolus vulgaris) by affinity chromatography. Biochim Biophys Acta 1975;405:72-81. and biochemical 15. Leavitt R, Felsted R, Bachur N. Biological properties of Phaseolus vulgaris isolectins. J Biol Chem 1977;252:2961-6, of 16. Hewitt D, Coates ME, Kakade ML, Liener IE. A comparison fractions prepared from navy (Haricot) beans (Phaseolus vulgaris. L.) for germfree and conventional chicks. Br J Nutr 1973;29:423-35. DJ, Burgess CD. Further observations on the 17. Jayne-Williams toxicity of navy beans (Phaseolus vulgaris) for Japanese quail (Coturnix. Coturnix Japonica). J Appl Bacterial 1974;37:149169. DJ. Influence of dietary jack beans (Canavalia 18. Jayne-Williams ensiformis) and of concanavalin A on the growth of conventional and gnotobiotic Japanese quail (Coturnix. Coturnix Japonica). Nature 1973;243:150-1. A, Clarke EMW, King TP, Stewart JC. Nutritional 19. Pusztai evaluation of kidney beans (Phaeolus vulgaris): chemical composition, lectin content and nutritional value of selected cultures. J Sci Food Agric 1979;30:843-8. DE. Disc electrophoresis of 20. Reisfeld RA, Lewis UJ, Williams basic proteins and peptides on polyacrylamide gels. Nature 1962;195:281-3. 21. Welkos SL, Toskes PP, Baer H. Importance of anaerobic bacteria in the cobalamin malabsorption of the experimental rat blind loop syndrome. Gastroenterology 1981;80:313-20. 22. Rodgers JB, Fondacaro JD, Kot J. The effect of synthetic diether phospholipid on lipid absorption in the rat. J Lab Clin Med 1977;89:147-52. 23. Dahlquist A. Method for assay of intestinal disaccharidase. Anal Biochem 1964;7:18-25. GG, Pietraszek J, McGinnis J. Effect of diet on 24. Untawale adhesion and invasion of microflora in the intestinal mucosa of chicks. Proc Sot Exp Biol Med 1978;159:276-80. RA, Rout WR, Toskes PP. Jejunal brush border 25. Giannella injury and impaired sugar and amino acid uptake in the blind loop syndrome. Gastroenterology 1974:67:965-74. 26. Wilson AB, King TP, Clarket EMW, Pusztai A. Kidney bean (Phaseolus vulgaris) lectin induced lesions in rat small intestine. 2. Microbiological studies. J Comp Path01 1980;90:497602.
March
1983
27. Gibbons RJ, Dankers I. Lectin-like constituents in foods which react with components of serum, saliva and Streptococcus mutans. Appl Environ Microbial 1981;41:880-8. 28. Toskes PP, Giannella RA, Jervis HR, Rout WR, Takeuchi A. Small intestinal mucosai injury in the experimental blind loop syndrome. Gastroenterology 1975;68:1193-203. 29. King TP, Pusztai A, Clarke EMW. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 1. Light microscope studies. J Comp Path 1980;90:585-95. 30. Lorenzsonn V, Olsen WA. In vivo responses of gut intestinal epithelium to intraluminal dietary lectins. Gastroenterolog! 1982;82:838-48. 31. Freeman HT, Lotan R, Kim YS. Application of lectins for detection of goblet cell glycoconjugate differences in proximal and distal colon of the rat. Lab Invest 1980;42:405-12. 32. Etzler ME, Branstrator ML. Differential localization of cell surface and secretory components in rat intestinal epithelium by use of lectins. J Cell Biol 1974;62:329-43. 33. Freeman HJ, Etzler ME, Garrido AB, Kim YS. Alteration in cell surface membrane components of adapting rat small intestinal epithelium. Studies with lectins after massive proximal jejunoileal resection and jejunoileal transposition. Gastroenterology 1978;75:1066-72, 34. King TP. Pusztai A, Clarke EMW. Immunocytochemical localization of ingested kidney bean (Phaseolus vulgarus) lectins in rat gut. Histochem J 1980;12:201-8. 35. Boedecker EC, Boldt DH. Effect of plant lectins on the binding of human intrinsic factor-vitam:n B12 complex (IF-B,,) to isolated guinea pig brush border membranes (B.B.M.). Gastroenteroiogy 1975;68:866. 36. Kim YS, Brophy EJ, Nicholson JA. Rat intestinal brush border membrane peptidases. J Biol Chem 1976;251:3206-12. 37. Brady PG, Vannier AM, Banwell JG. Identification of the dietary lectin, wheat-germ agglutinin, in human intestinal contents. Gastroenterology 1979;75:236-9. 38. Nachbar MD, Oppenheim JD. Lectins in the United States diet: a survey of lectins in commonly consumed foods and a review of the literature. Am J Clin Nutr 1980;3:2338-45. 39. Nachbar MS, Oppenheim JD, Thomas JO. Lectins in the U.S. diet: isolation and characterization of a lectin from the tomato
RED KIDNEY
BEAN
LECTINS
AND
MALABSORPTION
515
(Lycopersicon esculentum). J Biol Chem 1980;255:2056-61. 40. Freed DLJ, Buckley CH. Mucotractive effect of lectin. Lancet 1978:i:585-6. 41. Public Health Laboratory Service. Unusual outbreak of food poisoning. Br Med J 1976;2:1268. 42. Faschingbauer H, Kofler L. ijber Gifwirkung von Rohen Bohnen and Bohnenkeimlingen. Klin Wochenschr 1929; 42:1069-72. 43. Bray GA, Barry RE, Benfield JR, Castelnuova-Tedesco P. Rodin J. Intestinal bypass for obesity decreases food intake and taste preference. Am J Clin Nutr 1976;29:779-83. 44. Smith H. Microbial surface in relation to pathogenicity. Bacterial Rev 1977;41:475-500. 45. Ofek I, Beachey EH. Bacterial adherence. Adv Int Med 1980;25:503-32. 46. Giannella RA. Pathogenesis of acute bacterial diarrhea1 disorders. Annu Rev Med 1981;32:341-57. 47. Pistole TG. Interaction of bacteria and fungi with lectins and lectin-like substances. Annu Rev Microbial 1981;35:85-112. 48. Gilboa-Garber N. Mizrahi L. Interaction of mannosephilic lectins of Pseudomonas aeroginosa with luminous species of marine enterobacteria. Microbios 1979;26:31-6. 49. Kuitunen P, Visakorpi JK, Savilahti E. Pelkonen P. Malabsorption syndrome with cow’s milk intolerance. Clinical findings and course in 54 cases. Arch Dis Child 1975;50:3516. 50. Ament ME, Rubin CF. Soy protein-another cause of the flat intestinal lesion. Gastroenterology 1972;62:227-34. 51. Challacombe DN, Richardson JM. Rowe B, Anderson CM. Bacterial microflora of the upper gastrointestinal tract in infants with protracted diarrhea. Arch Dis Child 1974; 49:270-7. 52. MacDonald WL, Bradborg LL. Flick AL, Thier JS, Rubin CE. Studies of celiac sprue. IV. The response of the whole length of the small bowel to a gluten-free diet. Gastroenterology 1964:47:573-89. 53. Gorbach SL, Banwell JG, Jacobs B, et al. Tropical sprue and malnutrition in West Bengal. 1. Intestinal microflora and absorption. Am J Clin Nutr 1970:23:1545-58.
1983:84:506-15
Phytohemagglutinin Derived From Red Kidney Bean (Phaseolus vulgaris): A Cause for Intestinal Malabsorption Associated With Bacterial Overgrowth in the Rat J, G. BANWELL, D. H. BOLDT, J. MEYERS, F. L. WEBER the technical assistance of B. MILLER and R. HOWARD Division of Gastroenterology; Department Medicine, Lexington, Kentucky
of Medicine,
Plant lectins or carbohydrate binding proteins interact with membrane receptors on cellular surfaces but their antinutritional efiects are poorly defined. Studies were conducted to determine the effects of phytohemaglutinin, a lectin derived from raw red kidney bean (Phaseolus vulgaris), on small intestinal absorptive function and morphology, and on the intestinal microflora. Phytohemagglutinin was isolated in purified form by thyroglobulin-sepharose 4B affinity chromatography. Red kidney bean and (6% and 0.576, respectively, of phytohemagglutinin dietary protein) were fed in a purified casein diet to weanling rats for up to 21 days. Weight loss, associated with malabsorption of lipid, nitrogen, and vitamin B12, developed in comparison with animals pair-fed isonitrogenous casein diets. Antinutritional effects of red kidney bean were reversible on reinstitution of a purified casein diet. An increase in bacterial colonization of the jejunum and ileum occurred in red kidney bean- and phytohemagglutinin-fed animals. When antibiotics were included in Received January 5, 1982. Accepted September 29, 1982. Address requests for reprints to: J. G. Banwell, M.D., Division of Gastroenterology and Clinical Nutrition, Department of Medicine, Case Western Reserve University, Cleveland, Ohio 44106. The research was supported in part by Grant AM 25529 from the National Institutes of Arthritis, Metabolism, and Digestive Disease. This work was presented in part at the American Gastroenterology Meeting, New York, New York, May 1981, and in abstract form in GASTROENTEROLOGY 1981;80:1103. Ralph Giannella, M.D., provided much helpful advice in preparation for the microbiologic studies that were carried out with the assistance of Ms. Peggy Pendley. The assistance of Joyce Cairns and Cheryl Inman in the preparation of this manuscript is gratefully acknowledged. 0 1983 by the American Gastroenterological Association 0016-5085/83/030506-10$03.00
University
Jr., with
of Kentucky, College of
the diet, malabsorption of [3H]triolein and 57Covitamin B12 in red kidney bean-fed animals was partially reversed and, in germ-free animals, purified phytohemagglutinin had no demonstrable antinutritional effect. Mucosal disaccharidase activity was reduced in red kidney bean- and phytohemagglutinin-fed animals, but intestinal mucosal morphology was unchanged. Dietary administration of phytohemagglutinin, alone or as a component of red kidney bean, caused intestinal dysfunction, which was associated with, and dependent upon, small intestinal bacterial overgrowth. Adherence of enteric bacteria to the mucosal surface was enhanced by phytohemagglutinin which may have facilitated small intestinal bacterial overgrowth. Many of the legumes (pulses and beans) that are of major importance as sources of dietary protein for animals and humans contain antinutritional substances [trypsin-inhibitory substances and plant lectins (l-4)]. Impaired growth, weight loss, diarrhea, and ultimately death may occur with feeding of a variety of raw beans to animals (5-7). These deleterious antinutritional effects may be reduced by heat treatment (8). Although the lethal effects of some beans such as the castor bean (Ricinus communis) have been recorded frequently and their biochemical action studied in detail (9), the antinutritional effects of other plant lectins or carbohydrate-binding proteins remain poorly defined. These have potent biological effects (lo),including erythrocyte agglutination (ll),mitogenicity for lymphocytes (121,and inhibition of tumor cell migration (13). Red kidney bean lectins (phytohemagglutinin) comprise a family of five glycoproteins, each con-
March
RED KIDNEY
1983
taining four subunits of similar molecular weight (~32,000 daltons) held together by noncovalent bonds. The individual isolectins contain various proportions of the L (mitogenic) and E (erythro-agglutinating) subunits (14,15). Jaffe (4,6)postulated that the toxicity of phytohemagglutinin (PHA), the major antinutritional factor in the red kidney bean (RKB) (Phaseolus vulgaris), may be related to its interreaction with receptor sites on the surface of intestinal cells, which causes a nonspecific interference with absorption of nutrients. The antinutriand other lectins (18), tional effect of PHA (16,17) however, depends on the presence of intestinal bacteria: in the germ-free state, toxic effects of lectins are ameliorated. These studies were performed in weanling rats to examine the alterations caused by crude RKB or purified PHA on intestinal absorptive function, the small intestinal microflora, and mucosal morphology.
BEAN
LECTINS
AND
MALABSORPTION
507
Diet All diets were constructed from Teklab purified diets (Teklab, Madison, Wise.), free from complex carbohydrates and fiber. The standard control diet was purified diet AIN-76, containing 6% fat and 23.5% protein, including all essential minerals and vitamins. Diets of lower protein content were formulated utilizing a protein-free diet powder to which varying amounts of casein were added. Homogenized crude RKB powder or purified PHA was incorporated in the test diets so that they were isonitrogenous with control diets (Table 1). Control and experimental animals were paired to ensure comparable intake of nutrients. Food consumption was determined daily. All metabolic studies were conducted for 7-day periods after an initial 3-day equilibration period on a dietary regimen. Dietary, fecal, and urine nitrogen were measured with a Coleman Nitrogen Analyser (Coleman Instruments, Perkin Elmer Company, Chicago, Ill.). The RKB diet fed to germ-free animals was irradiated (4.5 Mrads) to ensure sterility before use. Irradiation caused no change in the erythroagglutinating titer of a saline extract of the RKB diet, and irradiated RKB diets caused the same magnitude of weight loss as nonirradiated diets.
Materials and Methods Animals The outbred Sprague-Dawley male rat (Sprague Dawley, Indianapolis, Ind.) was used for all experiments. Batches of 6-12 animals, weighing 40-80 g each, were housed individually in mesh bottom metabolic cages. Rats were maintained at a constant temperature with a 12-h light-dark cycle. Funnels with screen baffles enabled collection of urine separate from feces. Feces were collected on wire screens, urine under mineral oil. Animals were weighed daily. Water was allowed ad libitum. Germ-free animals were introduced into isolators on arrival. Monitoring of germ-free status was carried out at J-day intervals, and by culture of tissues at the time of death. Table
1. Cumulative Containing
Dietary
Regimens
In general, the diets we used were comparable to those employed by animal nutritionists (5-7,16) who have studied the antinutritional effects of RKB or fractions containing purified PHA. Diets with low or normal protein content were used due to the practical difficulty of extracting purified PHA in sufficient quantities for feeding experiments. For instance, when PHA was incorporated in diets as 0.5% dietary protein, a total of 3-4 g were required for each metabolic balance study which represented a yield from five to six affinity columns.
Weight Change and Nitrogen Excretion for Groups of Weanling Rats Fed Isonitrogenous Purified Casein, Raw Red Kidney Bean Extract, or Purified Phytohemagglutinin
Dietary Group”
Dietary
regimen
% N,
N, intake (m&at
day)
Dietary consumption (g/rat. day]
Mean weight change (g/rat 7 day]
Diets
Nitrogen excretion (mgirat day) Fecal
Urine
5.2 + 0.5 36.3 + 2.0"
24.7 2 1.8 13.1 + 2.2'
A
C. T.
Casein protein Casein protein + RKB (6%)
(10%) (4%)
1.37 1.41
53.2 t 4.7 57.8 + 3.7
3.9 r 0.3 4.1 +- 0.2
+1.7 + 1.1 -3.9 i- 1.0"
B.
C. T.
Casein protein Casein protein t RKB (6%)
(20%) (14%,)
3.09 3.18
222.3 + 14.3 237.1 ? 16.2
7.2 ? 0.5 7.5 I 0.5
f11.5 t 0.4 f5.3 + 1.1"
13.5 i- 1.4 70.1 -+ 3.9"
100.7 + 7.2 64.9 + 2.8'
c
c. T.
Casein protein Casein protein t 0.5% PHA
(5%) (4.5%)
0.85 0.99
33.2 t 3.0 37.2 f 1.9
3.9 t- 0.4 3.8 t- 0.9
-2.1 k 1.3 -5.9 + 0.8'
7.35 t 0.4 15.5 t 1.6"
14.1 t 0.9 9.6 2 0.9"
D
C.”
Casein (10%) (irradiated] Germ-free animals
1.20
60.2 2 2.6
5.0 2 0.2
-5.0 + 1.0
13.2 + 1.8
24.6 t 1.6
T.
Casein protein (4%) + RKB (6%) (irradiated) Germ-free animals
1.40
91.3 k 3.3
6.5 t 0.2
PO.7 i 2.0
27.8 t 2.2'~
38.2 2 1.9"
I’n = 6 for each group: T = test; C = control. 1 Animals failed to consume total diet.
” p < 0.001 vs. control
group.
’ p < 0.005 vs.control
group.
I’p < 0.025.I’p c 0.05.’p < 0.1.
508 BANWELL ET AL.
Dietary groups studied
GASTROENTEROLOGY Vol. 84, No. 3
were
Group A C. Casein protein (10%) T. Casein protein (4%) + RKB (6%) This diet was utilized to define effects of homogenized RKB, containing PHA, on small intestinal function. Group B C. Casein protein (20%) T. Casein protein (14%) + RKB (6%) A study to determine effects of RKB in a diet containing normal dietary quantities of protein. Group C C. Casein protein (5%) T. Casein protein (4.5%) + purified PHA (0.5%) A diet to study effects of PHA alone on small intestinal function. Group D C. Casein protein (10%) (irradiated) fed to germ-free animals T. Casein protein (4%) + RKB (6%) (irradiated) fed to germ-free animals A dietary study to determine the effect of RKB in a germfree environment.
Preparation
of Phytohemagglutinin
Red kidney beans (Phaseolus vulgaris) were obtained from a single source (Laurelbrook Farms, Be1 Air, Md.). A saline extract of RKB was made by homogenizing 100-150 g of whole beans in a Waring blender for 3 min at 20,000 rpm. The fine powder obtained was added to 500 ml phosphate-buffered saline (PBS) and stirred overnight at 4°C. This suspension was filtered through cheesecloth, and the filtrate separated by centrifugation at 39,000 g for 20 min. Purified PHA was extracted utilizing thyroglobulin-sepharose 4B affinity chromatography by the method of Felsted and coworkers (14). The single discrete elution peak yielded 0.55-0.85 g total protein (PHA) from lo-15 g of crude bean protein saline extract applied to the column. Polyacrylamide gel electrophoresis, carried out on 7% acrylamide, 0.5% bisacrylamide gel at 4°C according to the method of Reisfeld et al. (20), revealed five separate phytohemagglutinin isolectins with different and distinct mobilities. Columns were reused repetitively after rejuvenation of thyroglobulin sepharose. The PHA yields from several column preparations were combined, analyzed for erythroagglutinating activity and protein content, and mixed with purified casein to provide the requisite PHA diet.
Agglutination
by the reciprocal of the serial dilution immediately preceding the first visible cell button of nonagglutinating cells.
57Co-Vitamin Blz Absorption One-half micro Curie of “Co-vitamin Blz (sp act 1 &i/l pg) (Amersham-Searle, Chicago, Ill.) was administered as a 0.5-ml bolus by intragastric tube to fasting animals with or without the addition of rat intrinsic factor (IF) prepared by the method of Welkos et al. (21). The cobalamin binding capacity of the IF preparation was 18.5 ng cobalamin/ml rat gastric mucosal extract (assayed by a charcoal binding radioassay method in the laboratory of Dr. P. Toskes, Gainesville, Fla.). Animals were counted in a whole-body gamma counter immediately after administration of 57Co-vitamm Blz. They were subsequently recounted 1 wk later and absorption was measured as the percentage of retained radioactivity at that time. Studies to define the effects of RKB or PHA were conducted after a lo-day exposure to the diet. A separate set of studies (see Figure 3B) examined the effects of RKB, before and after cessation of RKB intake, on vitamin Blz absorption measured on day 10, day 20 (6 days after cessation of RKB), and day 26.
[%]OJeic Acid and [14C]TrioJein Absorption Absorption of [3H]oleic acid and [14C]trioJein was studied in separate experiments by the method of Rodgers, Fondacaro, and Kot (22). Lipid emulsions were prepared by first sonicating 4 mM of sodium taurocholate and 3 mM of phospholipid (pig liver phosphatidylcholine, Calbiothem, San Diego, Calif.) in 50 ml of normal saline. The bile salt-phospholipid mixture was added to a beaker containing either radioactive lipid and the mixture was resonicated. Four milliliters of the mixture which contained 10 &i was administered by gavage to fasting animals exposed to RKB or control diets (group A) for 10 days. To determine [3H]oleic acid excretion, 96-h collections of total fecal output were carried out: [‘4C]triolein absorption was determined by killing animals at 5 h, and, after ligature of stomach, small intestine, cecum, and colon, segments were removed and contents washed into separate beakers. Fecal material containing [3H]oleic acid or washings from the intestinal tract segments containing [‘4C]triolein was saponified with 33% KOH solution and boiled with 95% alcohol under reflux. Lipid was then extracted with petroleum ether. An aliquot of the petroleum ether phase was counted by scintillation spectrometry. Radioactive lipids in the gastrointestinal tract and feces were considered unabsorbed.
Assay
Agglutination assays were carried out utilizing red blood cells from a single group 0 donor. A 4% suspension of washed red cells was made and agglutination assays were carried out using microtiter plates and serial dilution with a standardized pipette [Cooke Laboratory Products, Alexandria, Va.). The erythroagglutinin titer was defined
Intestinal
Morphology
Tissue from control and test animals was collected from jejunum, ileum, cecum, and colon immediately after death (within 60 s). Specimens were processed for light microscopy (LM) and transmission electron microscopy (TEM). Sections stained with hematoxylin and eosin were
RED KIDNEY BEAN LECTINS AND MALABSORPTION
March 1983
examined by LM. Specimens for TEM were fixed in 1.75% glutaraldehyde and 0.1% sodium cacodylate and buffered for 1-2 wk before subsequent fixation and examination. Histologic tissue was coded and examined without knowledge of the dietary regimen to which animals had been exposed.
Brush
Border
Enzymes
Activities of lactase, sucrase, and maltase were measured in jejunal and ileal homogenates by the method of Dahlquist (23). One unit of enzyme activity is the amount that hydrolyzes 1 mmol of substrate per milligram mucosal protein.
Immunodiffusion Phytohemagglutinin
Assay
Bacteriological
Studies
All animals were fasted 18 h before death by CO2 narcosis. A lo-cm length of intestine distal to the ligament of Treitz, and a lo-cm segment of ileum, 5 cm proximal to the ileocecal valve, were immediately removed and transferred to prewashed sterile bottles containing PBS. Each segment was opened with sterile scissors and washed with four separate changes of PBS (24). The tissue was then weighed and ground with a Vortis homogenizer in 3.0 ml deoxygenated broth medium at 45,000 rpm for IO s.One milliliter of the first wash, of the fourth wash, and of the intestinal homogenate were mixed with 9 ml of deoxygenated broth and serially diluted to lo-' by ten-fold tube dilutions. One-tenth milliliter of the appropriate dilution was placed on the following media: Aerobic incubations: Blood agar for total aerobes, MacConkeys media for enterobacteriacae, mannitol salt media for staphylococci, Sabauraud’s media for fungi, and Mitis salivarius for streptococci were used. Anaerobic incubation (72-96 h): Total anaerobes were cultured on blood agar and microaerophilic bacilli were cultured in COZ for 72 h on Rogosa SL agar. Bacteria were identified by gross morphology, using a gram stain, and by subculturing and selected biochemical tests. Organisms from anaerobic plates were confirmed as being anaerobic by failure to grow on aerobic subculture. The number of specific organisms was expressed as the logarithm to the base 10 of the mean colony counts on ulates containing l-1000 oreanisms 1251.
of Antibiotics
4% + Red Kidney
Bean
into the Casein
6% Diet
Antibiotics were incorporated into the casein 4% diet containing 6% RKB to study whether “‘Co-vitamin Blz and [‘4C]triolein malabsorption could be reversed. Metronidazole (Searle Co., Chicago, Ill.) and kanamycin (Kantrex, Bristol Laboratories, Syracuse, N.Y.) were carefully dispersed in the feed. The daily dosage of metronidazole (7.5 * 0.3 mg/rat per day) and kanamycin (7.5 2 0.3 mgirat per day) in the diet were similar to levels shown to cause a reduction in both aerobic and anaerobic intestinal bacteria in rats with and without surgically constructed blind loops (21). Control animals were pair fed the same’diet without antibiotics. Statistical
for Fecal
To determine whether PHA was detectable in fecal pellets of PHA-fed animals, pooled g-day fecal collections were extracted in saline and PHA isolated by thyroglobulin sepharose affinity chromatography. The glycine buffer eluate was concentrated on a PM-10 membrane. After dialysis, immunodiffusion plates were run utilizing purified PHA, the fecal PHA extract, and purified antibodies to the E and L isolectins of PHA. Phytohemagglutinin antibody was obtained from Vector Laboratories Inc., Burlingame, Calif.
Intestinal
Incorporation
509
Testing
Results were analyzed with the Student’s t-test for independent variables and expressed as the mean 2 standard error (SEM).
Results In preliminary experiments, both weanling (40-80 g) and adult (~200 g) rats that were fed ad libitum diets containing either RKB or PHA consistently ate less food and grew less or lost weight during the 2-3 wk study period. They remained as active as their control littermates, but their fecal pellets were soft and unformed. Reinstitution of a control casein diet resulted in rapid reversal of weight loss to weight gain within 24 h. Death often supervened in weanling rats by the third or fourth week of dietary treatment with RKB or PHA. Animals fed a 0.5% PHA plus 4.5% casein protein diet ad libitum ate less of their diet (5.4 ? 0.2 g/day] than animals fed 5% casein (11.7 + 0.4 g/day, p < 0.001) during a lo-day study period and lost weight (PHAfed: -1.2 * 0.1 g/day, control-fed: +1.5 +- 0.2 g/day).
To normalize for the effect of variable dietary intake, all subsequent metabolic studies were performed with groups of weanling animals pair-fed isonitrogenous diets (Table 1). Group A diets, containing 6% RKB, caused significantly greater growth depression or weight loss than control diets (p < 0.001). Animals on a normal protein diet containing RKB (group B) grew better than those on a low protein intake, but less than their pair-fed controls (p < 0.001). Animals on a 4.5% casein protein diet + 0.5% PHA (group C) also lost more weight than those pair fed the control 5% casein diet. Germ-free animals (group D) fed 6% RKB lost less weight than the conventional animals in group A (p < 0.025). Nitrogen
urine
Excretion
Studies of nitrogen excretion in feces and during the dietary studies are also shown in
510
GASTROENTEROLOGY Vol. 84. No. 3
BANWELL ET AL.
loss in RKB-fed germ-free animals (group D) was less than that seen in their conventional counterparts (group A).
r
Lipid Absorption With RKB feeding, a greater percentage of radiolabeled lipid was recovered in the 96-h fecal collection than in controls (p < 0.01) (Figure 1). [‘4C]triolein recovery in the small bowel lumen, cecum, and colon 5 h after gastric administration was also greater in RKB-fed animals than controls but failed to reach statistical significance (p < 0.15). Malabsorption of radioactive lipid in rats on the RKB diet was partially reversed by simultaneous antibiotic administration. Figure
1. Rekovery 3H-oleic acid (left side) in feces and l“Ctriolein (right side) from intestinal contents after orogastric administration to weanling rats on test diet (hatched bar], control diet (clear bar), and test animals treated with antibiotics (stippled bar).
Table 1. Fecal nitrogen losses were increased fivefold to sevenfold in the RKB-fed groups (A and B) compared with animals fed isonitrogenous control casein diets (p < 0.001); PHA-fed animals exhibited a similar increase in fecal nitrogen (group C). Urinary nitrogen losses were, in general, slightly lower in RKB-fed animals than in controls. Fecal nitrogen
Disaccharidase
The mucosal activities of maltase, sucrase, and lactase were reduced in animals fed 4% casein protein + 6% RKB diets when compared with pairfed control animals (p < 0.001) (Figure 2). Activity was lower in the jejunum for all these enzymes and in the ileum for maltase and sucrase. Animals fed a diet containing 0.5% purified PHA + 4.5% casein diet had enzyme activity that was higher overall than the group fed RKB; however, jejunal enzyme activity
JEJUNUM
ILEUM
25
Figure 2. Disaccharidase bars]
animals
-
activity in jejunal (left side] and ileal tissue (right side) in control exposed
to RKB or PHA.
Activity
[clear
bars]
and test (hatched
or stippled
March
RED KIDNEY
1983
remained significantly reduced in comparison with the group receiving the control diet (p < 0.001). In the ileum, there was no significant difference in PHA-fed animals vs. controls. Vitamin Blz Absorption 57Co-Vitamin B12 absorption was impaired 714 days after institution of the RKB diet (20.7% * 1.4%) compared with control animals (61.2% +1.5%) (p > 0.001) (Figure 3A, B). Administration of rat gastric intrinsic factor caused no augmentation of 57C~-B12 uptake [RKB: 20.0% +- 2.3%; control: 60% & 1.8% (p < O.OOl)]. Impaired 57Co-vitamin Blz absorption was observed as early as 3 days after commencing the RKB diet (30.2% + 1.1%). Abnormal 57C~-B11 absorption was reversed by reinstitution of a 10% casein diet (65.8% +- 2.2%), and improvement in 57C~-B12 malabsorption towards normal (45.8% k 3.4%) was also observed when antibiotics were included in the diet (Figure 3A). Feeding of heat-treated RKB, which caused inactivation of hemagglutinating activity, resulted in a normal 57Co-vitamin B12 absorption (74.8% -+1.0%). Another group of animals was fed restricted amounts of a 5% casein diet to achieve weight loss comparable to that seen in the animals fed RKB (NZ intake, 0.85%,2.2mg Nz/ rat - day; weight lost -6.6 g/rat - i’ day). This starvation regimen alone did not (57Co-vitamin impair 57Co-vitamin B 12 absorption Blz absorption >60%).
c :
1OOr
;
go-
p
g
6070-
s
50-
cu ai
40-
.f E m .= > b
511
Findings
Microbiologic
Data
1. Influence of RKB diet on intestinal bacterial flora (Table 2). Quantitative culture of bacterial flora from jejunal and ileal segments revealed that RKB feeding caused bacterial proliferation. Counts of total aerobes were increased siginificantly in jejunum in the fourth wash of the mucosa. In ileum, counts were significantly increased in both wash 4 and tissue homogenates. Total anaerobes were also increased in the ileum in wash 4. Escherichia
A-A
6%
?? -.
10% CASEIN
R K B + 4%
PROTEIN
PROTEIN 9-m
i
“\
I
I
I
i
_... _.. .I.,,.:. :.:.:.:.: i ..i..... .:,.::. ..... i I :....:. ................. I ...:..... ................. :....:. ......... i i :.:.:.:.: . ....... .‘. i I .‘ ........‘ .... ..i..... i I ........ “=12 :i::. n=7 i i i
lo-
o
i 57
~ .
I Co-Vitamin
812
S’Co-Vitamin
B12
:,
5’Co-Vitamin
a12
ANTIa:OTICS
A Figure
MALABSORPTION
._.-.-.-..
20-
A
AND
Histologic evaluation by one of us (J.M.) of sections of jejunal and ileal tissue, which had been coded and randomized, revealed no effect of RKB or purified PHA on intestinal morphology. No change in villus morphology, crypt-cell mitotic activity, or round-cell infiltration of the lamina propria was noted. Tissue morphology on scanning EM was also normal in appearance. On TEM, morphological features of the microvilli, globlet cells, mitochrondia, and smooth and rough endoplasmic reticulum also showed no detectable difference from control tissue. In very occasional EM sections, ballooned or fused microvilli were observed in tissue that was otherwise normal. Histologic sections in the liver, pancreas, colon, and stomach showed no pathological changes on light microscopy.
E a-
30-
iiT
LECTINS
Morphological
c
8o
BEAN
3. A. “7Co-Vitamin B,, absorption in rats fed 4% casein + 6% RKB diet (hatched bars] or without added rat intrinsic factor. Antibiotic administration to two groups of animals fed in the bars (- . - . -)I caused “%o-vitamin B,, absorption to increase (stippled bar) compared without antibiotics (hatched bar). B. Absorption of “Co-vitamin B,, before, during the casein diet ( 1 ). Weight change on RKB diet (triangles) and after reinstitution of control
I2 DAY
24
B 10% casein diet (open bars) with or 4% casein + 6% RKB diet [depicted with those animals on an RKB diet RKB diet, and after return to a 10% diet (circles).
512
BANWELL
ET AL.
GASTROENTEROLOGY
cc& counts were greater in RKB-fed animals in jejunum and ileum in both wash 4 (p < 0.001) and tissue homogenates. 2. Influence of purified PHA on intestinal bacterial flora (Table 2). In PHA-treated animals, total aerobes in jejunal fluid, wash 4, and intestinal homogenate were significantly increased (p <
Vol. 84, No. 3
0.001).In the ileum, significant increases were seen with both washes 1 and 4. Particularly large increases in total anaerobes (p < 0.001)were observed in jejunal washings and homogenates, but only in wash 4 from the ileum. Escherichia coli were also increased in all jejunal specimens (p < 0.005) as well as in wash 4 from the ileum.
Table 2. Total, Aerobic, Anaerobic, and Escherichia coli Colony Counts of JejunaJ and Ileal Contents (Wash I), Fourth Mucosal Washing, and Mucosal Homogenate in Rats Fed Red Kidney Bean Diet vs. Control Diet and Purified Phytohemagglutinin Diet vs. Control Diet Bacterial colony count (log,, organism/g tissue wet wt) Wash 1 RKB diet vs. control diet (n = 6 for each group, test and controlj Total aerobes Jejunum 4.25 2 1.74 (NS) Test 2.78 t 1.24 Control Ileum 8.41 2 0.46 (NS) Test 6.79 + 0.70 Control Total anaerobes Jejunum 4.06 2 1.72 (NS) Test 2.81 -c 1.48 Control Ileum 7.79 2 0.76 (NS) Test 7.98 2 0.72 Control Escherichia coli Jejunum 1.04 2 1.56 (NS) Test 0.68 r 1.04 Control Ileum 6.55 ? 1.70 (NS) Test 4.84 k 1.27 Control
Purified PHA Diet vs. Control Diet (n = 7 for each group, test and control) Total aerobes Jejunum 5.96 t 1.20” Test 1.99 r 1.29 Control Ileum 8.40 2 0.27b Test 6.00 t 1.05 Control Total anaerobes Jejunum 7.12 2 0.65” Test 2.90 2 1.18 Control Ileum 8.29 f 0.23 (NS) Test 6.77 2 1.12 Control Escherichia coli Jejunum 6.48 _’0.83” Test 0.75 + 0.69 Control Ileum 5.15 2 1.19 (NS) Test 3.42 f 1.15 Control
Wash 4
Tissue homogenate
1.52 2 1.34” 0.00
1.38 ? 1.08 (NS) 1.80 f 0.83
7.41 t 0.13” 2.34 Ifr 1.05
8.09 2 0.15” 4.33 ‘- 1.29
1.85 2 1.37 (NS) 1.37 + 0.49
4.93 2 1.08 (NS) 3.18 f 0.88
7.08 k 0.33” 2.24 t 1.23
8.17 2 1.18 (NS) 4.56 + 1.34
0.57 ? 3.09” 0.0
3.51 5 1.59” 0.68 * 0.68
6.62 2 0.46” 1.12 c 0.12
6.94 t 0.27” 2.05 t 1.18
4.63 2 0.81” 0.0
7.02 t 0.40” 3.30 t 0.78
3.74 t 1.41’. 1.46 2 0.89
6.44 t 1.03 (NS) 5.85 t 1.17
5.19 2 0.43” 0.35 5 0.33
6.81 r 0.57” 2.89 k 0.73
5.14 + 1.26” 1.60 k 0.96
6.13 + 0.99 (NS) 5.53 * 0.99
3.52 2 1.30” 0.0
6.15 k 0.54” 0.44 2 0.41
4.11 +- 1.56” 0.80 2 0.74
4.17 2 1.59 (NS) 1.82 t 0.80
0 p i 0.001 vs. control group. b p < 0.05 vs. control group. I’p < 0.005 vs. control group. NS = not significant.
RED KIDNEY BEAN LECTINS AND MALABSORPTION
March 1983
Isolation and Purification Phytohemagglutinin
of Fecal
Evidence that PHA was excreted in the feces in animals fed either RKB or purified PHA was derived from two sources: (a) fecal saline extracts from RKB-fed animals (group B) were observed to cause agglutination in group 0 human red cells at a higher titer (1:lZ)than extracts from control feces (1:4),and (b) when a saline extract of pooled feces from 4 PHA-fed animals was applied to a sepharose thyroglobulin affinity column, a protein peak with the characteristics of PHA was eluted. Immunodiffusion analysis of this eluted material utilizing purified PHA-E and PHA-L antibodies revealed that both PHA-E and PHA-L components were present in the eluate and formed lines of identity with the purified mixture of PHA-E and PHA-L antigens. The total calculated fecal excretion of PHA in this study was 6.0 mg PHA, or 1.2% of the total dietary intake of PHA in animals fed 0.5% PHA in the diet.
Discussion Dietary administration of crude RKB or purified PHA caused malabsorption of nitrogen, lipid, and vitamin B12 in conventional but not germ-free weanling rats. Small intestinal bacterial colonization was associated with RKB and PHA feeding, and antibiotics partially reversed lipid and vitamin Blz Phytohemagglutinin, malabsorption. therefore, caused intestinal malabsorption and facilitated bacterial proliferation in the small bowel. Several previous studies (2-8,16,17,24) have classified features of the action of PHA. This work defines the pathophysiology of PHA-induced malabsorption and its dependence on bacterial adherence in the small intestine. The polymicrobial small intestinal bacterial overgrowth was associated with increased adherence of bacteria to the mucosal surface. An increased adherence of microbial organisms to the mucosal surface caused by PHA feeding may have provided the requisite conditions for proliferation of bacteria and development of small bowel bacterial overgrowth. Wilson et al. (26)have described overgrowth of E. coli in rats fed diets containing 10% raw RKB. No bacterial overgrowth occurred until 24 h after feeding began. When a bean diet, containing a reduced lectin content was fed, less prominent changes in coliform counts occurred. Untawale et al. (24)also observed greater attachment of E. coli to the intestinal mucosa of chickens fed raw RKB compared with those fed autoclaved RKB. Recently, Gibbons and Dankers (27) observed that lectin-like activity in aqueous extracts of several commonly ingested fruits affected attachment of Streptococcus mutans to saliva-coated hydroxyapatite beads.
513
Intestinal mucosal morphology after RKB and PHA exposure was essentially unaltered in this study, with the exception that, rarely, occasional microvilli were fused or club shaped. Identifiable histologic changes, as encountered in the small bowel bacterial overgrowth syndrome in experimental animals and humans, may have been absent due to the relatively short period of bacterial overgrowth in our study (14-21 days). Mucosal changes in small bowel bacterial overgrowth had usually been described with chronic bacterial colonization (25,28). Phytohemagglutinins from RKB (29) and other lectins (30)have also been noted to cause histologic lesions in the rat small intestine. Microvillous damage in the study of King et al. (29) was greatest in those animals fed strains of RKB that produced the higher concentrations of PHA lectins and that also caused growth inhibition. Strains of RKB that produced lower lectin concentrations resulted in growth inhibition without mucosal injury. Disaccharidase activity was reduced in RKB- and PHA-treated animals in the present study, which may be attributed to a direct effect on the microvillus membrane, but RKB and PHA concentrations may have been insufficient to cause histologic alterations in mucosal morphology. A variety of lectins (31-34)have been observed to bind to the mucosal surface membranes of both small and large intestine. Evidence for direct PHA adherence to the intestinal wall by immunofluorescent techniques has been obtained in rat studies (34) after crude RKB administration and in work in our own laboratory. Boedecker and Boldt observed that radiolabeled PHA demonstrated saturable binding to ileal brush border membranes and competition with IF-B12 receptor sites (35). Phytohemagglutinin might, therefore, directly alter absorption of nutrients by effects on transport across the intestinal microvillous membrane (36)as well as by changes in intestinal microflora. Phytohemagglutinin is only one of several antinutritional agents present in raw RKB (2):evidence would suggest that it accounts for ?35% of the total growth inhibitory fraction of one variety of Phaseolus vulgaris seed (7). It is not possible to calculate the exact amount of PHA available in RKB from this study since extraction of PHA by affinity chromatography was incomplete and, moreover, not all PHA in RKB may have been bioavailable in the intestine (37).
Specific saccharide inhibitable red blood cell agglutination is often demonstrable in commonly ingested foodstuffs, suggesting that intestinal exposure to dietary lectins in the human diet may be more widespread than supposed (38,39). The toxic effects of hemagglutinins in leguminous foodstuffs are generally eliminated by proper heat treatment (2,6,8),
514
BANWELL ET AL.
GASTROENTEROLOGY Vol.84, No. 3
but they might prevail if destruction is incomplete. Diarrhea1 and toxic effects in humans have been described with ingestion of uncooked kidney beans (40-42).
The present study has also shown that a portion of the purified PHA fed to these animals resisted proteolytic degradation during passage through the intestinal tract and was identifiable in the feces based on its similarity to purified PHA on affinity chromatography and gel immunodiffusion. Jaffe et al. also noted that fecal material from animals fed RKB demonstrated agglutinating activity (6). Resistance to intestinal proteolytic activity may, therefore, be an important feature of PHA (16,17)and other lectins (38,40)that permits their survival throughout the small and large intestine. Phytohemagglutinin and RKB ingestion caused weight loss and growth failure that occurred in association with diminished dietary caloric intake in the preliminary experiments conducted without pair feeding. Reduced caloric intake is known to be a major contributory factor in the weight loss associated with intestinal malabsorption (43).The pair feeding experiments indicated, however, that RKB and PHA were directly responsible for malabsorption. Furthermore, comparable weight loss induced in animals by a starvation diet did not of itself impair nitrogen or vitamin B12 absorption. The relevance of these findings to human disease must await further study. Adherence of bacteria to mucosal surfaces is an important factor in the pathogenicity of several animal (44) and human enteric pathogens (45,46).Lectins may enhance bacterial Lectins other than PHA, adherence (24,28,47,48). and even nonlectin dietary proteins, may have similar antinutritional effects in experimental animals (2-4).There are also several human intestinal diseases in which small intestinal dysfunction may be related to ingested dietary proteins or changes in small intestinal bacterial flora, such as cow’s_milk (49) and soy-protein intolerance (SO), chronic diarrhea1 disease of infancy (51),celiac disease (52),and tropical Sprue (53). Phytohemagglutinin-induced small intestinal bacterial overgrowth and malabsorption may provide a model for certain features of these diseases.
References Patwardhan VN. Pulses and beans in human Clin Nutr 1962;11:12-30. Liener I. Significance for humans of biologically in soybeans and other food legumes. J Am 1979;56:121-9. Rackis JJ. Biological and physiological factors Am Oil Chem Sot 1974;51:161A-74A. Jaffe WG. Hemagglutinins. Liener IE, ed. Toxic plant foodstuffs. New York: Academic Press,
nutrition.
Am J
active factors Oil Chem Sot in soybeans. constituents 1969.
J of
5. Kakade ML, Evans RJ. Growth inhibition of rats fed navy bean fractions. J Agric Food Chem 1965;13:450-2. 6. Jaffe WG, Vega Lette CL. Heat-labile growth-inhibiting factors in beans (Phaseolus vulgaris). J Nutr 1968;94:203-10. 7. Evans RJ, Pusztai A, Watt WB, Bauer DH. Isolation and properties of protein fractions from navy beans (Phaseolus vulgaris) which inhibit growth in rats. Biochim Biophys Acta 1973;303:175-84. 8. Muelenaere HJH. Effect of heat treatment on the haemogglutinating activity of legumes. Nature 1964;201:1029-30. 9. Olsnes S, Pihl A. Abrin, ricin, and their associated agglutinins. In: Cuatrecasas P, Greaves LC, eds. Receptors and recognition: the specificity and action of animal, bacterial and plant toxins. Vol 1. New York: Halsted Press, 1976:131-73. 10. Goldstein IJ, Hayes CE. The lectins: carbohydrate-binding proteins of plants and animals. Adv Carbohydr Chem Biothem 1978;35:127-340. 11. Egorin MJ, Bachur SM, Felsted RL. Leavitt RD, Bachur NR. Phaseolus vulgaris isolectin binding to human erythrocytes. J Biol Chem 1979;254:894-8. 12. Ling NR, Kay JE. Lymphocyte stimulation. New York: American Elsevier Publishing Co. Inc., 1975. RN, Sane1 FT, Felsted RL, Bachur NR. 13. Egorin MJ, Odders Inhibitory effects of phytohemagglutinin isolectins Lq and E, on L1210 cells. Cancer Res 1978;38:1677-87. of the phytohe14. Felsted R, Leavitt RD, Bachur NR. Purification magglutinin family of proteins from red kidney beans (Phaseolus vulgaris) by affinity chromatography. Biochim Biophys Acta 1975;405:72-81. and biochemical 15. Leavitt R, Felsted R, Bachur N. Biological properties of Phaseolus vulgaris isolectins. J Biol Chem 1977;252:2961-6, of 16. Hewitt D, Coates ME, Kakade ML, Liener IE. A comparison fractions prepared from navy (Haricot) beans (Phaseolus vulgaris. L.) for germfree and conventional chicks. Br J Nutr 1973;29:423-35. DJ, Burgess CD. Further observations on the 17. Jayne-Williams toxicity of navy beans (Phaseolus vulgaris) for Japanese quail (Coturnix. Coturnix Japonica). J Appl Bacterial 1974;37:149169. DJ. Influence of dietary jack beans (Canavalia 18. Jayne-Williams ensiformis) and of concanavalin A on the growth of conventional and gnotobiotic Japanese quail (Coturnix. Coturnix Japonica). Nature 1973;243:150-1. A, Clarke EMW, King TP, Stewart JC. Nutritional 19. Pusztai evaluation of kidney beans (Phaeolus vulgaris): chemical composition, lectin content and nutritional value of selected cultures. J Sci Food Agric 1979;30:843-8. DE. Disc electrophoresis of 20. Reisfeld RA, Lewis UJ, Williams basic proteins and peptides on polyacrylamide gels. Nature 1962;195:281-3. 21. Welkos SL, Toskes PP, Baer H. Importance of anaerobic bacteria in the cobalamin malabsorption of the experimental rat blind loop syndrome. Gastroenterology 1981;80:313-20. 22. Rodgers JB, Fondacaro JD, Kot J. The effect of synthetic diether phospholipid on lipid absorption in the rat. J Lab Clin Med 1977;89:147-52. 23. Dahlquist A. Method for assay of intestinal disaccharidase. Anal Biochem 1964;7:18-25. GG, Pietraszek J, McGinnis J. Effect of diet on 24. Untawale adhesion and invasion of microflora in the intestinal mucosa of chicks. Proc Sot Exp Biol Med 1978;159:276-80. RA, Rout WR, Toskes PP. Jejunal brush border 25. Giannella injury and impaired sugar and amino acid uptake in the blind loop syndrome. Gastroenterology 1974:67:965-74. 26. Wilson AB, King TP, Clarket EMW, Pusztai A. Kidney bean (Phaseolus vulgaris) lectin induced lesions in rat small intestine. 2. Microbiological studies. J Comp Path01 1980;90:497602.
March
1983
27. Gibbons RJ, Dankers I. Lectin-like constituents in foods which react with components of serum, saliva and Streptococcus mutans. Appl Environ Microbial 1981;41:880-8. 28. Toskes PP, Giannella RA, Jervis HR, Rout WR, Takeuchi A. Small intestinal mucosai injury in the experimental blind loop syndrome. Gastroenterology 1975;68:1193-203. 29. King TP, Pusztai A, Clarke EMW. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 1. Light microscope studies. J Comp Path 1980;90:585-95. 30. Lorenzsonn V, Olsen WA. In vivo responses of gut intestinal epithelium to intraluminal dietary lectins. Gastroenterolog! 1982;82:838-48. 31. Freeman HT, Lotan R, Kim YS. Application of lectins for detection of goblet cell glycoconjugate differences in proximal and distal colon of the rat. Lab Invest 1980;42:405-12. 32. Etzler ME, Branstrator ML. Differential localization of cell surface and secretory components in rat intestinal epithelium by use of lectins. J Cell Biol 1974;62:329-43. 33. Freeman HJ, Etzler ME, Garrido AB, Kim YS. Alteration in cell surface membrane components of adapting rat small intestinal epithelium. Studies with lectins after massive proximal jejunoileal resection and jejunoileal transposition. Gastroenterology 1978;75:1066-72, 34. King TP. Pusztai A, Clarke EMW. Immunocytochemical localization of ingested kidney bean (Phaseolus vulgarus) lectins in rat gut. Histochem J 1980;12:201-8. 35. Boedecker EC, Boldt DH. Effect of plant lectins on the binding of human intrinsic factor-vitam:n B12 complex (IF-B,,) to isolated guinea pig brush border membranes (B.B.M.). Gastroenteroiogy 1975;68:866. 36. Kim YS, Brophy EJ, Nicholson JA. Rat intestinal brush border membrane peptidases. J Biol Chem 1976;251:3206-12. 37. Brady PG, Vannier AM, Banwell JG. Identification of the dietary lectin, wheat-germ agglutinin, in human intestinal contents. Gastroenterology 1979;75:236-9. 38. Nachbar MD, Oppenheim JD. Lectins in the United States diet: a survey of lectins in commonly consumed foods and a review of the literature. Am J Clin Nutr 1980;3:2338-45. 39. Nachbar MS, Oppenheim JD, Thomas JO. Lectins in the U.S. diet: isolation and characterization of a lectin from the tomato
RED KIDNEY
BEAN
LECTINS
AND
MALABSORPTION
515
(Lycopersicon esculentum). J Biol Chem 1980;255:2056-61. 40. Freed DLJ, Buckley CH. Mucotractive effect of lectin. Lancet 1978:i:585-6. 41. Public Health Laboratory Service. Unusual outbreak of food poisoning. Br Med J 1976;2:1268. 42. Faschingbauer H, Kofler L. ijber Gifwirkung von Rohen Bohnen and Bohnenkeimlingen. Klin Wochenschr 1929; 42:1069-72. 43. Bray GA, Barry RE, Benfield JR, Castelnuova-Tedesco P. Rodin J. Intestinal bypass for obesity decreases food intake and taste preference. Am J Clin Nutr 1976;29:779-83. 44. Smith H. Microbial surface in relation to pathogenicity. Bacterial Rev 1977;41:475-500. 45. Ofek I, Beachey EH. Bacterial adherence. Adv Int Med 1980;25:503-32. 46. Giannella RA. Pathogenesis of acute bacterial diarrhea1 disorders. Annu Rev Med 1981;32:341-57. 47. Pistole TG. Interaction of bacteria and fungi with lectins and lectin-like substances. Annu Rev Microbial 1981;35:85-112. 48. Gilboa-Garber N. Mizrahi L. Interaction of mannosephilic lectins of Pseudomonas aeroginosa with luminous species of marine enterobacteria. Microbios 1979;26:31-6. 49. Kuitunen P, Visakorpi JK, Savilahti E. Pelkonen P. Malabsorption syndrome with cow’s milk intolerance. Clinical findings and course in 54 cases. Arch Dis Child 1975;50:3516. 50. Ament ME, Rubin CF. Soy protein-another cause of the flat intestinal lesion. Gastroenterology 1972;62:227-34. 51. Challacombe DN, Richardson JM. Rowe B, Anderson CM. Bacterial microflora of the upper gastrointestinal tract in infants with protracted diarrhea. Arch Dis Child 1974; 49:270-7. 52. MacDonald WL, Bradborg LL. Flick AL, Thier JS, Rubin CE. Studies of celiac sprue. IV. The response of the whole length of the small bowel to a gluten-free diet. Gastroenterology 1964:47:573-89. 53. Gorbach SL, Banwell JG, Jacobs B, et al. Tropical sprue and malnutrition in West Bengal. 1. Intestinal microflora and absorption. Am J Clin Nutr 1970:23:1545-58.