Department of Biochemistry, Nippon Medical School, 1-1-5 Sendagi, Bunkyo-ku, Tokyo113, Japan
ABSTRACT Inhibitory effects of soybean proteinisolate (SPI) and soybean lectin on the intestinal absorption of nonheme iron were investigated by in vivostudies in rats. Rats fed the SPI-based diet absorbedsignificantly less iron than did control rats fed thecasein-based diet. Supplementing the SPI diets with 8%D-galactose significantly increased the incorporation ofiron into liver ferritin, although D-galactose did not significantly increase iron absorption. Heat treatment ofSPI significantly increased iron absorption. Ascorbatedid not enhance iron absorption in rats fed the SPI-baseddiet. The presence of lectin in an aqueous extract of SPIwas suggested by hemaggiutination activity as well asby ¡mmunoreactivity with soybean lectin antibody.Soybean lectin introduced into ligated segments of theupper small intestine of rats inhibited ferrous iron absorption. This inhibitory effect, especially in the mucosa!uptake, was significantly improved by addition of N-acetyl-D-galactosamine to soybean lectin. Soybeanlectin had no effect on ferric iron absorption. Our resultssuggest that a portion of the reduction in iron absorptionin rats fed SPI may be due to lectins. J. Nutr. 122:1190-1196, 1992.
INDEXING KEY WORDS:
•iron absorption•soybean lectin
soybean protein isolateconcanaualin A
rats
The process of iron absorption in the small intestine consists of several steps, one of which is therelease of nonheme iron from ingested food in thedigestive tract. Many factors in food, either activatorsor inhibitors, can affect the availability of food ironfor absorption. Several reports (1-4) have shown thatsoybeans and soybean products reduce the absorptionof nonheme iron, although there are some contradictory studies (5, 6). Phytates in soybean and theformation of iron-phytate complexes that are poorlysoluble in the intestinal lumen have been suggestedas a major factor in this reduced absorption. However,this must be clarified by more specific evidence (7, 8).
Lectins are common components of a variety oflegumes, and certain lectins conjugate specificmonosaccharides or polysaccharides, causing hemaggiutination (9). Because the consumption of lectinsdisturbs normal growth in humans and experimentalanimals, the effect of lectins on nutrient intakes hasbeen studied (10). Some of these studies demonstratedmorphological injury of the small intestinal mucosacaused by the adhesion of lectin to the mucosal surface. However, this injury could be prevented by thesimultaneous administration of saccharides with specific affinity for the lectins or by the heat treatmentof the foods containing lectins (11, 12).
The present research tested, for the first time toour knowledge, the inhibitory effect of soybeanlectins on iron absorption. We used an in vivo radiola-beled iron absorption technique with rats and foundan inhibition of iron absorption in rats fed a dietcontaining soybean protein isolate (SPI)1 as theprotein source. Inhibition was also observed whenpurified soybean lectin or concanavalin A (ConA) waspreintroduced into ligated intestinal segments.
MATERIALS AND METHODS
Materials. Male Wistar rats weighing 200-300 gwere purchased from Nippon Bio-Supply Center(Tokyo, Japan). This study received clearance fromthe Animal Ethics Steering Committee of NipponMedical School. Rats were fed for 1 wk a commercialnonpurified rat diet (MF, Oriental Yeast Co., Tokyo,Japan) with the following composition (g/kg diet):protein, 246; lipid, 56; mineral, 64; fiber, 31; solublenitrogen-excluded compounds (e.g., carbohydrate),523; water, 83; and 14.98 MJ/kg. Iron-free casein-
based diet was prepared by Oriental Yeast Co. (Table1). Soybean protein isolate (Fujipro R) was obtainedfrom Fuji Seiyu Co. (Osaka, Japan). Fujipro R, prepared from soybeans from Illinois, is a neutralized,sterilized, spray-dried and powdered SPI with the following composition (g/kg): protein [determined by theKjeldahl method (13)], 905; water, 60; ash, 45; fat, 1;trace amount of fiber. The Protein DispersibilityIndex is 92-93. The SPI contained a trace of trypsininhibitor unit as defined by Kakade et al. (14) and nourease activity. The following materials were purchased as noted: 59FeSO4and 59FeCi3 from Du PontNEN (Boston, MA); purified soybean lectin fromHonen Oil Co. (Tokyo, Japan); purified ConA fromPharmacia Fine Chemicals (Uppsala, Sweden) (thepurity of each lectin was demonstrated by HPLC);Bio-Gel P-300 and molecular weight markers fromBio-Rad (Richmond, CA); rabbit anti-soybean lectinantibody and goat anti-rabbit IgG antibody conjugatedto horseradish peroxidase from Sigma Chemical (St.Louis, MO). Unless otherwise stated, all other chemicals were obtained from Wako Chemicals (Tokyo,Japan).
Experiment 1. Rats were divided randomly intoseven groups and were fed the following diets: 1) thecasein-based diet (control); 2) casein-based diet containing 8% D-galactose at the expense of sugar,- 3)casein-based diet with the rats given 100 umol ofsodium ascorbate with the radioactive iron mixture;4) the SPI-based diet (control), prepared by substituting SPI for casein with an equal protein content; 5)SPI-based diet containing 8% D-galactose at the ex
pense of sugar; 6) SPI-based diet with the rats given100 umol of sodium ascorbate with the radioactiveiron mixture; 7) heated SPI (H-SPI)-based diet (H-SPIwas prepared by mixing SPI with three volumes ofwater (wt/v), heating it in an electric oven at 200°Cfor l h and grinding in a mortar,- the toasted andpowdered H-SPI was substituted for SPI).
Rats were housed individually in metabolic cagesand freely fed their respective diets for 5 d. Afterovernight food deprivation, the rats were again fedtheir test diets for 1 h, and then a radioactive ironmixture (74 kBq of 59FeSO4,2 umol of FeSO4-7H2Oand 2 umol of sodium ascorbate per 100 g body wt, in0.6 mL of 20 g/L fructose) was administered by acatheter to the stomach of each rat within 15 minafter preparation of the mixture. They were fed therespective diets for two more days and then fed thecommercial rat diet (previously described) thereafter,because 93% of total radioactivity in feces collectedfor 7 d appeared within 2 d after administration ofradioactive iron. Feces of individual rats were collected for 7 d. The rats were then anesthetized, killedby drawing blood from the abdominal aorta, and thelivers excised. Radioactivity in total feces and 1 mL ofwhole blood was determined in Aloka AutowellGamma System ARC221 (Aloka, Tokyo, Japan). Totalradioactivity in blood was estimated by assuming thetotal blood volume to be 7.5% of body weight (15).The absorption of iron was calculated as the percentage of the dose detected in blood 7 d after administration and as the percentage of the dose not recovered in feces collected over the 7 d. Radioactiveiron in liver ferritin was determined by immunologi-cally precipitating ferritin from the liver extract withantiserum to rat liver ferritin (16).
Experiment 2. Rats weighing -300 g were dividedrandomly into six groups for testing ferrous or ferriciron absorption. The upper intestinal loops were pre-filled with one of the following test solutions: 1) 0.5mL of saline (control); 2) 1 mg of soybean lectin in 0.5mL of saline; 3) 1 mg of soybean lectin +10 mg of D-galactose in 0.5 mL of saline,- 4} 1 mg of soybeanlectin + 10 mg of N-acetyl-D-galactosamine in 0.5 mLof saline,-5) 1 mg of ConA in 0.5 mL of saline,- and 6) 1mg of ConA +10 mg of methyl-a-D-mannoside in 0.5mL of saline. Ferrous and ferric iron solutions wereprepared by the method of Muir et al. (17). Ferrousiron solution consisted of 74 kBq of 59FeSO4, 0.357umol of FeSU4-7H2O and 7.14 umol of ascorbate per0.5 mL of saline. The pH was adjusted to 7.0-7.4 withNaHCOa. Ferric iron solution consisted of 74 kBq of59FeCl3,0.357 umol of FeCl3 and 357 umol of citratein 0.5 mL of saline, adjusted to pH 7.0-7.4 with 2mol/L NaOH. Animals were anesthetized and 7-cmintestinal segments (measured from duodenum toupper jejunum) were ligated. Blood circulation in theintestinal segments was maintained. The test solution (0.5 mL) was injected at the most proximalpoint of each segment, and the ligature was secured
while carefully withdrawing the needle in order toensure no loss of substance. Thirty minutes later, theferrous or ferric radioactive iron mixture (0.5 mL) wasintroduced into the same segment. Rats were killedby bleeding from the abdominal aorta after 30 min ofradioactive iron administration. Intestinal segmentswere carefully isolated and the serosal surface wasrinsed with cold saline. The radioactivities in thesegments were determined in the gamma counter.This fraction represents the nontransferred iron (A)and includes the resident iron in the mucosal cellsand the iron adhered to the luminal surface. Thesegments were then cut longitudinally and the luminal surface was carefully rinsed with cold saline.The mucosa was scraped off and its radioactivitymeasured (mucosal uptake, B}. Total absorption (T)was calculated as the sum of the transferred iron(total dose, D minus A] plus mucosal iron (B), formulated as T = (D - A] + B.
Hemoglobin determination. Blood hemoglobin wasdetermined by a cyanmethohemoglobin method (18).
Bio-Gel P-300 chromatography. One milliliter offerrous iron solution as mentioned in Experiment 2and 1 mL of soybean lectin or ConA solution (1 g/Lsaline) were combined and incubated at 37°Cfor 30
min. The mixture then was applied to a column (1.4 x25 cm) of Bio-Gel P-300 equilibrated with 20 mmol/LTris, 0.1 mol/L NaCl, pH 7.4, and eluted into1-mL fractions with the same buffer. The radioactivity and protein concentration (19) in each fractionwere measured. Molecular weight markers were alsochromatographed on the same column.
Hemagglutinating test. One gram of SPI or H-SPIwas stirred in 20 mL of PBS overnight at 4°Cand
centrifuged at 9000 x g for 15 min. The protein concentrations of the supernatants of SPI and H-SPI were
9.5 and 1.1 g/L, respectively. Serial dilutions of these
extracts were tested for hemagglutinating activitywith 2% suspensions in PBS of rat RBC (20).
Dot immunobinding assay. The procedure was amodification of the methods of Hawkes et al. (21) andJahn et al. (22). In simplified form, purified soybeanlectin, SPI extract, H-SPI extract and PBS as controlwere applied to nitrocellulose sheets (S &. S Membranefilter, Dassel, Germany) and incubated withrabbit anti-soybean lectin antibody or whole rabbitserum. After washing with saline, the sheet was incubated again with goat anti-rabbit IgG antibody con
jugated to horseradish peroxidase. Soybean lectin onthe sheets was identified by the color developed bythe reaction of 3,3'-diaminobenzidine and hydrogen
peroxide.Statistical analysis. The statistical analyses were
performed for differences of means and correlationsamong the values using SPSS/PC+â„¢ Ver. 3.0 (SPSS,Chicago, IL). One-way ANOVA was used to examine
the data for differences within each group. If significant effects were seen, the data were further analyzedby Tukey's or Duncan's multiple range test. Differ
ences were considered significant at P < 0.05. Data areexpressed as means ±SEM.
RESULTS
Effect of diets on body weight and hemoglobinconcentrations. There were no significant differencesin mean values of body weights and hemoglobin concentrations at the end of Experiment 1 between therats fed the SPI-based diet and those fed the casein-based diet (body weight: 252 ±9.6 and 236 ±7.3 g,hemoglobin: 106 ±3 and 119 ±3 g/L, respectively).No significant difference among various diet groupswas observed either in body weight or in hemoglobin.
TABLE 2
Effect of diets on iron absorption in rats fed soybean protein isolate (SPI)- or casein-based diet (Experiment I)1
% absorption of dose±4.1**'a 43.6 ±3.8a±3.7*'a 49.4 ±4.1a
41.5 ±3.9a
64.966.277.4 ±3.9'*'b
4
62.4 ±2.9* *'a64.3 ±3.3 *'a77.9 ±3.1* *'b
54.9 ±3.2"
Values are means ±SEMfor six rats in each group. Asterisks show significant differences (*P < 0.05, "P < 0.01 as assessed by Tukey's test)between the SPI-fed groups and the casein-fed groups for the same treatment. Means in each column with different superscript letters showsignificant differences (P < 0.05 as assessed by Duncan's test; Tukey's test did not reveal significant differences).
2See Materials and Methods for details.treatments: control, SPI- or casein-based diet; Ga-8, SPI- or casein-based diet containing 8% D-galactose,-AsA-100, 100 umol of sodium
ascorbate with radioactive iron mixture was given to control rats; H-SPI, heated SPI-based diet.4Not determined.
5g 20 40 60 80 100Fe absorption measured by fecal analysis (% of dose)
FIGURE 1 Correlation between the paired results of 59Feabsorption measured by fecal analysis and blood analysis.Correlation coefficient, r = 0.99; regression equation, Y =1.008X - 1.320. (•)iron-deficient rats (n = 65 ); (o) iron-replete rats (n = 34).
Iron absorption by rats fed the SPI- and casein-based diets for 7 d. Iron absorption was estimated bymeasurement of radioactivity in the feces collectedover the 7 d of the experiment and in blood at the endof this period (Table 2). An excellent correlation wasobserved between the two methods (r = 0.99) (Fig. 1).
Table 2 shows the effect of diet composition oniron absorption. In rats fed the SPI-based diet, ironabsorption was significantly lower than in those fedthe casein-based diet, as measured by either fecal orblood analysis. Supplementing the SPI-based dietswith 8% D-galactose did not significantly increaseiron absorption (Table 2), but significantly increasediron incorporation into liver ferritin in rats fed theSPI-based diet (Table 3). Ascorbate was not effectivein improving iron absorption when given to rats fedthe SPI-based diet. However, a significant stimulationin absorption was observed when ascorbate was addedto the iron material of rats fed the casein-based diet(Table 2). Ascorbate also markedly increased theamount of radioactivity incorporated into liver ferritin in the casein-fed group but not in the SPI-fedgroup (Table 3). Substitution of SPI with H-SPI significantly increased iron incorporation into liver ferritin,though the value did not reach that of the casein-fedcontrols (Table 3).
Hemagglutination activity. Serial dilutions of PBSextracts of SPI or H-SPI were tested for hemaggluti-nation activity against 2% suspensions of red blood
TABLE 3
Effect of diets on incorporation of iron into liver ferritin inrats fed soybean protein isolate (SPI)- or casein-based
diet (Experiment I)
Treatment2 SPI diet Casein diet
% incorporation ofdoseControl
Ga-8AsA-100
H-SPI0.07
±0.01a0.14 ±0.01b0.05
±0.01a0.18 ±0.02b0.48
±0.02* **'a0.48 ±0.03*'*'a1.21
±0.14***'b_3
'Values are means ±SEMof six rats in each group. Asterisksshow significant differences (***P < 0.001 as assessed by Tukey's
test) between the SPI-fed groups and the casein-fed groups for thesame treatment. Means in each column with different superscriptletters show significant differences (P < 0.05 as assessed by Tukey's
test).treatments: control, SPI- or casein-based diet; Ga-8, SPI- or
casein-based diet containing 8% D-galactose; AsA-100, 100 ^mol ofsodium ascorbate with radioactive iron mixture was given tocontrol rats,- H-SPI, heated SPI-based diet.
3Not determined.
cells from rats and the positive reactions were observed. The titer, defined as the reciprocal of thehighest dilution, was 64 for the SPI extract and 16 forthe H-SPI extract (Fig. 2).
Detection of soybean lectin in SPI and H-SPI extracts. The presence of soybean lectin in the extractswas confirmed by a dot immunobinding assay usinganti-soybean lectin antibody; it was visualized by thecolor reaction of the peroxidase conjugated to anti-rabbit IgG antibody. The positive reaction was clearin the SPI extract (Fig. 3). The soybean lectin concentration in the SPI extract was 0.05 mg/L.
Effect of ¡ectinson iron absorption in ligated intestinal segments. The direct effect of soybean lectinand ConA on the intestinal absorption of ferrous ironwas examined in vivo by using ligated segments ofthe upper small intestine in rats (Table 4). A significant reduction of total iron absorption in the pre-treated soybean lectin group and ConA group wasobserved as compared with controls. The addition ofD-galactose or A/-acetyl-D-galactosamine to thesoybean lectin solution did not significantly improvethe transferred iron and total absorption; N-acetyl-D-galactosamine significantly increased the mucosaluptake relative to the soybean lectin group, however.In the ConA-treated group, the addition of methyl-oc-D-mannoside to the ConA solution almost completelyimproved the mucosal uptake and total absorption.For ferric iron, there were no statistical differencesbetween the control values and the soybean lectin-and ConA-treated groups for total absorption,mucosal uptake, or transferred iron (Table 5).
Interaction of ferrous iron with soybean lectin orConA. To examine the possibility of the formation of
FIGURE 2 Hemagglutination activity of soybean protein isolate (SPI)and heated SPI extracts. Agglutination of red bloodcells from rats was assayed in V-bottom microtiter plate by the serial dilution technique. Red blood cells reacted with 1/64(upper) or less dilution of native SPI extract and 1/16 (lower) or less dilution of heated SPI extract.
complexes between ferrous iron and soybean lectin orConA, a mixture of each lectin and the radioactiveferrous iron solution was chromatographed on a Bio-Gel P-300 column. The elution patterns showed nocoincidence for the main peaks of radioactivity andprotein concentration. Approximate molecularweights of 130,000 and 4000 were found for soybeanlectin and 59Fe, respectively (Fig. 4, left panel). In the
case of ConA, two protein peaks appeared, a mainpeak [molecular weight, (MW) 39,000] and a secondary peak (MW 670,000); 59Fe was also separated
into a main peak (MW 4000) and a secondary peak(MW 670,000) (Fig. 4, right panel).
1 3
B
FIGURE 3 Detection of soybean lectin in soybeanprotein isolate (SPI) and heated SPI extractions by dot im-munoassay for rabbit anti-soybean lectin antibody: Î)soybean lectin (10 ng); 2) SPI extract (3 \iL)¡3) heated SPIextract (8 \iL); and 4) PBS (8 H.L)were applied on nitrocellulose sheets, and soybean lectin on the sheets was detectedby enzyme-linked immunosorbent assay labeled by horseradish peroxidase. Top row (A): rabbit anti-soybean lectinantibody; bottom row (B): rabbit serum.
DISCUSSION
Iron absorption in rats assessed by the appearanceof radioactive iron in blood, the excretion in feces andthe iron incorporation into liver ferritin was significantly lower in rats fed the SPI-based diet comparedwith those fed the casein-based diet (Tables 2, 3),indicating that 7 d of feeding the SPI-based dietcaused a reduction in iron absorption.
The valence state of iron and its solubility areimportant factors in the absorption of food iron. As-corbate enhances iron absorption by converting ferriciron to the ferrous state and forming a soluble chelatewith ferrous iron (23). However, our data have shownthat ascorbate in conjunction with the SPI-based dietwas not effective in promoting iron absorption.Hallberg (24), Sayers et al. (25) and Derman et al. (26)also observed the same phenomenon.
In this study we focused on lectins as a factorinvolved in the inhibitory process. Lectins are foundin most legumes and their toxic effects have beenreported (10, 27). Pusztai et al. (28) showed that beanswith a higher content of lectins caused more seriousdamage to the luminal surface of rat intestine thanthose with lesser lectin contents. Lectins have beenshown to conjugate readily with the glycocalyx layerof the microvillous membrane and cause structuraldamage to the luminal surface of the intestine(29-31). An inhibitory effect of lectins on the activityof peptidase and disaccharidase of enterocytes hasalso been reported (31, 32). Lectins also interfere withthe intestinal absorption of D-glucose and L-histidine(10). These toxic effects are prevented by proper heattreatment of lectins (33) or by supplementation withsugars that have specific affinity for the lectins (11).
We have shown that lectins are present in SPI bydemonstrating that extracts from SPI have hemagglu-tination activity against red blood cells of rats andimmunoreactivity to anti-soybean lectin antibody
each column with different superscript letters show significantdifferences (P < 0.05 as determined by Tukey's test). SBL = soybean
lectin.2See Materials and Methods for treatment details.
(Fig. 2, 3). The content of soybean lectin in SPI isestimated to be 1 mg/kg from the SPI-extract lectinconcentration. Although the H-SPI extract showedhemagglutination activity, soybean lectin was not im-munologically detected in the extract; H-SPI mayhave a very small amount of soybean lectin. Moreover, our present results suggest that the lectins inSPI may take part in the inhibition of iron absorption,i.e., rats fed the D-galactose-supplemented SPI-baseddiet or the H-SPI diet incorporated significantly moredietary iron into liver ferritin than did their controls(Table 3). The accumulation of more radioactive ironin ferritin (i.e., iron storage protein) reflects the factthat the experimental animals absorbed and retainedmore iron than did the controls. The value of ironincorporation into ferritin of rats fed the H-SPI dietdid not reach that of the casein-fed controls (Table 3).We could interpret this result as indicating that asmall amount of lectin that remained in the H-SPImight inhibit iron absorption. The prior treatment ofthe luminal surface of the ligated intestine withlectins significantly inhibited the absorption offerrous iron compared with controls (Table 4). Thereduced absorption by treating with ConA could becompletely recovered by adding methyl-a-D-manno-side. The observations suggest that the competitivebinding of the lectins to the glycocalyx or free saccha-rides (e.g., methyl-a-D-mannoside) on the luminalsurface may be a contributing factor in the inhibitoryprocess. On the other hand, D-galactose or JV-acetyl-D-galactosamine did not facilitate a complete recoveryfrom the total reduced absorption resulting fromsoybean lectin treatment, though N-acetyl-D-galac-tosamine significantly improved the reduced mucosaluptake. Supplementing the SPI diet with D-galactosealso did not significantly improve iron absorption
FIGURE 4 Gel filtration patterns of ferrous iron-lectinmixtures. The samples were prepared as described and applied to a column of Bio-Gel P-300 (1.4 x 25 cm)equilibrated with 20 mmol/L Tris-HCl in 0.1 mol/L NaClsolution, pH 7.4. Each 1-mL fraction was assayed for protein(solid line) and for radioactivity (dotted line). Arrows showthe locations of molecular markers by Bio-Rad gel filtrationstandard: 2) thyroglobulin (670,000); 2) IgG (158,000), 3)ovalbumin (44,000); 4} myoglobin (17,000); 5) vitamin B-12(1350). Con A = concanavalin A.
(Table 2). Therefore, some factors besides the competitive binding of the lectin to saccharides may also beinvolved in the inhibitory effect of soybean lectin,whereas that of ConA may primarily result from saccharides. Soybean lectin and ConA did not affectferric iron absorption as much as they did ferrous ironabsorption (Table 5). Ferrous and ferric iron may beabsorbed differently in rat intestine, as mentioned byMuir et al. (17).
TABLE 5
Effect of lectins on ferric iron absorption across ligated ratintestine (Experiment 2?
Treatment2Mucosal Transferred Totaluptake iron absorption
% absorptionofControlSBLSBL
+ D-galactoseConAConA
+ methyl-a-D-mannoside5.2
±0.33.0±0.64.0±0.52.3±0.22.5
±0.518.2
±2.815.7±1.514.3±0.7
15.7 ±0.911.9
±1.7dose23.3
±3.018.7±2.018.2±0.9
18.0 ±0.814.5
±1.8
'Values are means ±SEM of five rats in each group. SBL
soybean lectin.2See Materials and Methods for treatment details.
We have also confirmed that no lectin-iron complexes are formed, because each main peak of radioactivity and lectin protein was clearly separated in thegel filtration patterns (Fig. 4). In the case of soybeanlectin, a peak of protein (MW 4000) was observed inthe position of the main peak (MW 4000) of radioactivity. Because its molecular weight was too small tobe soybean lectin (MW 120,000) or its subunit (MW30,000) (34), the elution pattern therefore suggestedthat there was no complex formation. As for ConA,the secondary protein peak coincided with the peak ofradioactivity at the void volume position. This overlapping peak (>MW 670,000) seems to be a polymer ofConA (MW 52,000) (34) or ferrous iron in neutralsolution (MW 3000) (35). Even though the polymer ofConA bound iron, the amount would be too small tobe an important inhibitor of iron absorption in theligated intestine (Table 4), because radioactivity inthe secondary peak was only 7% of the total.
In this study, two lectins, soybean lectin andConA, inhibited iron absorption in the ligated intestine of rats. Some part of the reduction in ironabsorption from SPI may be due to lectins in the SPI.
LITERATURECITED
1. Cook, J. D., Morck, T. A. & Lynch, S. R. (1981) The inhibitoryeffect of soy products on nonheme iron absorption in man.Am. J. Clin. Nutr. 34: 2622-2629.
2. Hallberg, L. & Rossander, L. (1982| Effect of soy protein onnonheme iron absorption in man. Am. }. Clin. Nutr. 36:514-520.
3. Thompson, D. B. & Erdman, J. W., Jr. (1984) The effect of soyprotein isolate in the diet on retention by the rat of iron fromradiolabeled test meals. J. Nutr. 114: 307-311.
4. Gillooly, M., Torrance, J. D., Bothwell, T. K, MacPhail, A. P.,Derman, D., Mills, W. & Mayet, F. (1984) The relative effect ofascorbic acid on iron absorption from soy-based and milk-based infant formulas. Am. J. Clin. Nutr. 40: 522-527.
5. Steinke, F. H. & Hopkins, D. T. (1978) Biological availabilityto the rat of intrinsic and extrinsic iron with soybean proteinisolates. J. Nutr. 108: 481^89.
6. Rois, E., Hunter, R. E., Cook, J. D., Smith, N. J. & Finch, C. A.(1975) The absorption of iron as supplements in infant cerealand infant formulas. Pediatrics 55: 686-693.
7. Simpson, K. M., Morris, E. R. & Cook, J. D. (1981) Theinhibitory effect of bran on iron absorption. Am. J. Clin. Nutr.34: 1469-1478.
8. Hallberg, L., Rossander, L. & Skanberg, A.-B. (1987) Phytatesand the inhibitory effect of bran on iron absorption in man.Am. J. Clin. Nutr. 45: 988-996.
9. Sharon, N. & Lis, H. (1972) Lectins: cell-agglutinating andsugar-specific proteins. Science (Washington, DC) 177:949-959.
10. Liener, I. E. (1986) Nutritional significance of lectins in thediet. In: The Lectins: Properties, Functions, and Applicationsin Biology and Medicine (Liener, I. E., Sharon, N. & Goldstein,I. J., eds.), pp. 527-552. Academic Press, Orlando, FL.
11. Lorenzsonn, V. & Olsen, W. A. (1982) In vivo responses of ratintestinal epithelium to intraluminal dietary lectins. Gas-troenterology 82: 838-848.
12. Morck, T. A., Lynch, S. R. &. Cook, J. D. (1982) Reduction ofthe soy-induced inhibition of nonheme iron absorption. Am. J.Clin. Nutr. 36: 219-228.
13. Ballentine, R. (1957) Determination of total nitrogen and am
monia. Methods Enzymol. HI: 984-995.14. Kakade, M. L, Rackis, J. J., McGhee, J. E. & Puski, G. (1974)
Determination of trypsin inhibitor activity of soy products: acollaborative analysis of an improved procedure. Cereal Chem.55: 376-382.
15. Whittaker, P., Mahoney, A. W. & Hendricks, D. G. (1984)Effect of iron-deficiency anemia on percent blood volume ingrowing rats. J. Nutr. 114: 1137-1142.
16. Yoshino, Y., Manis, J. G. &. Schachter, D. (1968) Regulation offerritin synthesis in rat liver. J. Biol. Chem. 243: 2911-2917.
17. Muir, W. A., Hopfer, U. &.King, M. (1984) Iron transport acrossbrush-border membranes from normal and iron-deficientmouse upper small intestine. J. Biol. Chem. 259: 4896-4903.
18. International Committee for Standardization in Haematology(1967) Recommendations for haemoglobinometry in humanblood. Br. J. Haematol. 13: (suppl.) 71.
19. Lowry, O. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.(1951) Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193: 265-275.
20. Gartner, T. K., Gerrard, J. M., White, J. G. & Williams, D. C.(1981) Fibrinogen is the receptor for the endogenous lectin ofhuman platelets. Nature (Lond.) 289: 688-690.
21. Hawkes, R., Niday, E. & Gordon, J. (1982) A dot-immuno-binding assay for monoclonal and other antibodies. Anal. Bio-chem. 119: 142-147.
22. Jahn, R., Schiebler, W. & Greengard, P. (1984) A quantitativedot-immunobinding assay for proteins using nitrocellulosemembrane filters. Proc. Nati. Acad. Sci. U.S.A. 81: 1684-1687.
23. Conrad, M. E. & Schade, S. G. (1968) Ascorbic acid chelates iniron absorption: a role for hydrochloric acid and bile. Gastroen-terology 55: 35-45.
24. Hallberg, L. (1981) Bioavailability of dietary iron in man.Annu. Rev. Nutr. 1: 123-147.
25. Sayers, M. H., Lynch, S. R., Jacobs, P., Charlton, R. W., Both-well, T. H., Walkers, R. B. & Mayet, F. (1973) The effects ofascorbic acid supplementation on the absorption of iron inmaize, wheat and soya. Br. J. Haematol. 24: 209-218.
26. Derman, D. P., Ballot, D., Bothwell, T. H., Macfarlane, B. J.,Baynes, R. D., MacPhail, A. P., Gillooly, M., Bothwell, J. E.,Bezwoda, W. R. & Mayet, F. (1987) Factors influencing theabsorption of iron from soya-bean protein products. Br. J. Nutr.57: 345-353.
27. Torres-Pinedo, R. (1983) Lectins and the intestine. J. Pediatr.Gastroenterol. Nutr. 2: 588-594.
28. Pusztai, A., Clarke, E.M.W., King, T. P. & Stewart, J. C. (1979)Nutritional evaluation of kidney beans (Phaseolus vulgaris}:chemical composition, lectin content and nutritional value ofselected cultivars. J. Sci. Food Agrie. 30: 843-848.
29. Higuchi, M., Suga, M. & Iwai, K. (1983) Participation of lectinin biological effects of raw winged bean seeds on rats. Agrie.Biol. Chem. 47: 1879-1886.
30. Jindal, S., Soni, G. L. & Singh, R. (1984) Biochemical andhistopathological studies in albino rats fed on soybean lectin.Nutr. Rep. Int. 29: 95-106.
31. Oliveira, A. C., Vidal, B. C. & Sgarbier, V. C. (1989) Lesions ofintestinal epithelium by ingestion of bean lectins in rats. J.Nutr. Sci. Vitaminol. 35: 315-322.
32. Kim, Y. S., Brophy, E. J. & Nicholson, J. A. (1976) Rat intestinal brush border membrane peptidase. H. Enzymatic properties, immunochemistry, and interactions with lectins of twodifferent forms of the enzyme. J. Biol. Chem. 251: 3206-3212.
33. Higuchi, M., Tsuchiya, I. & Iwai, K. (1984) Growth inhibitionand small intestinal lesions in rats after feeding with isolatedwinged bean lectin. Agrie. Biol. Chem. 48: 695-701.
34. Goldstein, I. J. & Porez, R. D. (1986) Isolation, phys-icochemical characterization, and carbohydrate-binding specificity of lectins. In: The Lectins: Properties, Functions, andApplications in Biology and Medicine (Liener, I. E., Sharon, N.& Goldstein, I. J., eds.), pp. 35-247. Academic Press, Orlando,FL.
35. Terato, K., Fujita, T. & Yoshino, Y. (1973) Studies on ironabsorption. I. The role of low molecular polymer in ironabsorption. Am. J. Dig. Dis. 18: 121-128.
