Response of Turkey Poults to Soybean Lectin Levels Typically Encountered in Commercial Diets. 1. Effect on Growth and Nutrient Digestibility

Response of Turkey Poults to Soybean Lectin Levels Typically Encountered in Commercial Diets. 1. Effect on Growth and Nutrient Digestibility

Fasina, Y O

INTRODUCTION

Soybean lectin (SBL) and protease inhibitors are considered the major antinutrients present in raw soybeans because of the severity of the effects they cause in animals (liener, 1994a). The presence of these antinutrients has prevented the direct use of raw soybeans in poultry diets. Rather, the raw beans are defatted and processed by desolventizing-toasting to produce the high-protein cooked meal used in commercial poultry diets (Mustakas et al., 1981). Although the desolventization-toasting process reduces antinutrients in raw beans to low levels, processed meals still contain residual lectin levels of 0.22 to 0.67 mg/g of meal (Maenz et al., 1999). The desolventized-toasted soybean meal (SBM) may be included at levels of 25 to 50% in poultry diets. At such high dietary inclusion levels, the lectin content of the diet could be high enough to cause antinutritional effects.

Lectins are glycoproteins that possess at least one noncatalytic domain that binds reversibly to specific monoor oligosaccharide receptors on cells without altering the receptor covalent structure (Peumans and Van Damme, 1995). When lectins are ingested by animals, they can be degraded by intestinal digestive enzymes or survive intestinal digestion and bind to enterocytes on the brush border membrane (BBM). Upon binding, lectins may cause antinutritional effects such as disruption of the intestinal microvilli, shortening or blunting of villi, impairment of nutrient digestion and absorption, increased endogenous nitrogen loss, bacterial proliferation, and increased intestinal weight and size (Pusztai, 1993; liener, 1994a,b). These effects culminate in reduced nutrient assimilation by animals.

The deleterious effects of SBL on laboratory mice and rats are well established. Only a few studies have investigated the antinutritional effects of SBL in poultry. Most of these experiments utilized raw soybeans (including genetically low trypsin inhibitor varieties) or processed SBM as lectin sources with chickens as the experimental animal (Huisman, 1991; Mogridge et al., 1996; Douglas et al., 1999). The drawback in these studies is that the detrimental effects reported cannot be totally attributed to SBL because other antinutrients (such as protease inhibitors) present in soybeans can act synergistically with SBL (Jordinson et al., 1996). Recently, a more definitive study in which chicks were fed diets containing lectin-free or raw conventional soybeans was conducted (Douglas et al., 1999). Chicks fed lectin-free soybeans gained more weight (P

During the first 2 wk posthatch, young poultry, particularly turkey poults, inefficiently digest feedstuffs, especially fat (Sell, 1996; Jin et al., 1998). During this starting period, the required dietary protein is usually provided by incorporating high levels of SBM (up to 50% in the diet of turkeys). Therefore, high amounts of residual lectin may be present and may contribute to the suboptimal nutrient utilization that occurs in young poultry. To the best of our knowledge, the antinutritional effects of SBL in turkey poults have never been investigated.

The objective of this study was to investigate the effects of purified SBL on growth and digestive function in turkey poults. Purified SBL was prepared and added to a corn starch-casein-arginine semipurified diet and fed to starting poults over a 2-wk period, during which poult performance, nutrient digestibility, and brush border enzyme activities were evaluated.

MATERIALS AND METHODS

Experimental Design

Two experiments were conducted in accordance with the Guidelines for the Care and Use of Agricultural Animals in Agricultural Research and Teaching (Federation of Animal Science Societies, 1999). Experimental protocols were approved by the University Animal Care and Use Committee of North Carolina State University. For experiment 1,224 female day-old turkey poults (Hybrid Converter strain) were obtained from a commercial hatchery.2 The poults were wing-banded, weighed, and allotted to 1 of 4 dietary treatments on the basis of BW such that the distribution of starting weights was similar for all treatments. Each dietary treatment was assigned to 4 replicate pens, each consisting of 14 poults. The dietary treatments were 1) a casein-corn starch based semipurified control diet containing no SBL (PD), 2) a semipurified diet containing a low (0.024%) lectin level (PDL), 3) a semipurified diet containing a high (0.048%) lectin level (PDH), and 4) a corn-soybean meal diet (SBD) as a second control. The duration of the experiment was 14 d and poults were housed in thermostatically controlled Petersime batteries with raised wire floors, under continuous lighting. Room temperature was kept at 24°C and brooder temperature was maintained at 35 to 38°C. Poults had ad libitum access to feed and water in stainless steel troughs.

Experiment 2 was conducted using 96 female day-old turkey poults (Nicholas 88 strain) obtained from a commercial hatchery.2 Poults were wing-banded, weighed, and allotted to 1 of 4 dietary treatments on the basis of BW such that the distribution of starting weights was similar for all treatments. The 4 dietary treatments were the same as those described in experiment 1. Each treatment was assigned 4 replicate groups each consisting of 6 poults. The duration of the experiment was 12 d and poults were housed and raised in conditions similar to experiment 1.

Diet Preparation

The experimental diets (Table 1) met or exceeded the NRC (1994) nutrient recommendations for starting turkey poults. Corn starch and casein were chosen because they would be very low in lectins and certainly free of SBL. Adequate arginine was added to balance the high lysine content of casein. Lard at 10% of the diet was selected to assess the capacity to digest and absorb fat. Ingredients for the semipurified diets (PD, PDL, and PDH) were blended in a mixer and then pelleted with steam3 prior to the addition of purified SBL. Purified SBL was obtained using affinity chromatography (Fasina et al., 2003). A hemagglutination assay using untreated rabbit red blood cells (Lis et al, 1970) revealed that the activity of the purified SBL (35 HU/mg of protein) was similar to that of Sigma grade SBL^sup 4^ (33 HU/mg of protein). One hemagglutinating unit (HU) is defined as the amount of material required to cause a 50% decrease in the absorbance of an erythrocyte suspension within 2.5 h at room temperature (liener, 1955). Purity of the extracted SBL was ascertained using SDS- and nativePAGE as described by Fasina et al. (2003).

Appropriate amounts of SBL were mixed with corn starch and coated onto pellets of the PDL and PDH diets. The SBD was fed in a mash form and contained 0.013% SBL as determined by affinity chromatography. Titanium dioxide5 was incorporated as an indigestible marker into all diets in experiment 1.

Protocol for Experiment 1

Performance Evaluation. Body weight gain, feed consumption, and feed efficiency (FE) calculated as gain:feed ratio (mortality corrected) were recorded on d 7 and 14 of the experiment. Mortality was recorded daily.

Sampling for Brush Border Enzyme Assays, Measurement of Pancreas Weights, and Serum Harvesting. Poults were randomly sampled on d 6 (8 poults per treatment) and 14 (8 poults per treatment). For brush border enzyme assays, poults were killed by cervical dislocation and the small intestine was excised, placed on ice, and divided into 3 segments: duodenum (from the gizzard to the point of entry of the pancreobiliary ducts), jejunum (from the pancreobiliary ducts to the yolk stalk), and ileum (from yolk stalk to the ileocecal junction). Tissue sections (about 4-cm long) were cut from the middle of each segment, flushed with cold saline (0.9% NaCl), weighed, placed in a vial, and stored at -20°C until analyzed.

Pancreas weights were assessed as follows: the pancreas from each bird was quantitatively removed from the duodenal loop, blotted dry, and weighed. Pancreas weight was reported as percentage of BW.

Blood was collected from the poults on d 14 before killing, to measure the levels of circulating SBL-specific antibodies. Poults were bled by cardiac puncture and blood samples were collected in heparinized syringes. Blood samples were centrifuged6 at 1,500 rpm for 15 min. Serum was harvested and stored at -20°C until time to measure the levels of SBL-specific antibodies.

Fecal Collection. Feces were collected at 24-h intervals over 3 consecutive d in wk 1 (d 5 through 7) and wk 2 (d 12 through 14) to ensure that the samples obtained were representative of each pen. Care was taken to avoid spilled feed and feathers during collection. Fecal samples collected each day were kept frozen at -20°C until the end of the experiment. Fecal samples were thawed, dried in a forced draft oven at 75°C for 24 h, allowed to equilibrate with the atmosphere, ground, mixed, and stored in airtight containers until analysis for nutrient digestibility.

Protocol for Experiment 2

Body weight gain, feed consumption, and FE calculated as gain to feed ratio, were recorded on d 7 and 12 of the experiment. Mortality was recorded daily. Poults were randomly selected on d 6 (4 poults per treatment) and 12 (4 poults per treatment) of experiment and sampled for brush border enzyme assays, pancreas weight, and serum as previously described in experiment 1. Fat and starch digestibility were not assessed in experiment 2.

Analytical Methods

Nutrient Digestibility. Feed and fecal samples were analyzed for DM, titanium dioxide, fat, and starch content. Dry matter was determined by drying feed and fecal samples overnight in an oven7 at 105°C. Titanium dioxide was determined according to the procedures of Short et al. (1996). Fat content was determined in acidified feed and fecal samples as described by Renner and Hill (1960). Starch content was measured as the amount of glucose released by amyloglucosidase enzyme (Bjorck et al., 1987). Fat and starch digestibility were calculated by expressing the difference between dietary and fecal ether extract or starch content as a percentage of dietary values, relative to the titanium dioxide content of the feed and feces.

Measurement of Brush Border Enzyme Activities. Previously weighed tissue samples from duodenum, jejunum, and ileum were thawed and homogenized with 3 mL of cold 0.9% saline. Homogenization was accomplished using an Ultra-Turrax homogenizer8 at the moderate speed setting for 30 s. Aliquots of homogenates were analyzed for protein concentration (BCA Protein Assay Reagent, Kit 23225(9)) and brush border enzyme (alkaline phosphatase [ALP], maltase, and sucrase) activities. Alkaline phosphatase (EC 3.1.3.1.) activity was determined spectrophotometrically by measuring the rate of paranitrophenol formation, using Sigma4 kit (Sigma Diagnostic Kit 245). One unit of ALP activity was defined as the amount of enzyme that will produce one micromole of para-nitrophenol per minute. Maltase (EC 3.2.1.20) and sucrase (EC 3.2.1.48) activities were assayed colorimetrically using maltose and sucrose as substrates, respectively (Dahlqvist, 1968; Black and Moog, 1978). Activity was expressed as micromoles of glucose released per hour.

Detection of SBL-Specific Antibodies by ELISA. Microtiter plates10 were coated with 100 µL per well of 50 µg/mL SBL in carbonate-bicarbonate buffer, pH 9.6, and incubated overnight at 4°C. The next day, plates were washed (3×) with washing buffer (PBS containing 0.5% Tween-20; PBST).4 After washing, the remaining proteinbinding sites in the coated wells were blocked with dilution buffer (PBS containing 3% nonfat dry milk), and the plates incubated for l h at 37°C. Serum samples were thawed and diluted (1:15) with dilution buffer. After blocking, plates were washed (3×) with PBST, 50 µL/well of each sample was added in triplicate, and the plates were incubated for l h at 37°C. Plates were washed (3x) with PBST and a secondary antibody (goat anti-turkey IgG-peroxidase conjugated)11 was diluted (1:200) with dilution buffer and added to each well. Plates were incubated for l h at 37°C. After incubation, plates were washed (3×) with PBST and 100 µL/well of substrate solution (made from dissolving o-phenylene diamine tablet9 in citrate phosphate buffer) was added. Plates were then incubated in the dark at room temperature for 20 min, and absorbance of the reaction product (diaminophenazine, a yellow compound) was read at 450 nm. The absorbance of diaminophenazine measured in each serum sample was considered an estimate of the total SBL-specific antibody concentration. Thus, we did not construct a standard curve to calculate antibody concentration.

Statistical Analysis

Using the GLM procedure (SAS Institute, 1994), data were subjected to ANOVA for completely randomized designs with each replicate pen containing 6 to 14 poults. The experimental unit for all data was pen average. For performance (weight gain, feed intake, and FE), data were analyzed on a weekly and cumulative basis. Significant differences among means were determined with the Duncan option of the GLM procedure (Waller and Duncan, 1969; SAS Institute, 1994). Statements of statistical significance were based upon P

RESULTS

Weight Gain and FE

The cumulative weight gain of poults fed the 3 semipurified diets (PD, PDL, and PDH) was at least similar to those of poults fed the SBD in both experiments (Tables 2 and 3). Poults fed the control PD had a comparable growth rate and FE to poults fed the SBD (Tables 2 and 3). Weight gain was similar for poults fed the PD, PDL, and PDH (Tables 2 and 3). The trend observed for FE in experiment 1 was not similar to that observed in experiment 2. In experiment 1, FE was highest for poults fed the PDL on d 7 (Table 2). By d 14, poults in all treatments had gained similarly, but FE was still significantly higher for poults that consumed the lectin-supplemented diets (PDL and PDH; Table 2). On the contrary, in experiment 2, poults fed the control PD and the lectin-supplemented diets (PDL and PDH) had similar FE values on d 14 (Table 3). In spite of the disparity in FE results between experiments 1 and 2, it remained clear that the FE of poults fed the control PD was at least similar to the FE of poults fed the SBD in both experiments. In addition, the PDL treatment was at least similar to the control PD in weight gain and FE in both experiments.

Mortality

Total mortality at the end of experiment 1 was 14.7% (Table 4). Mortality occurred mainly from d 4 to 6 of the experiment. Mortality appeared to occur evenly in all replicates within each treatment. Mortality was significantly higher for poults fed the semipurified diets (PD, PDL, and PDH) compared with poults fed the control SBD (Table 5). Furthermore, poults in the PDL treatment had higher mortality than poults in the PD and PDH treatments (Table 5). In experiment 2, total mortality was 5.2% and mortality occurred mainly from d 5 to 8 (Table 6). There were no differences in mortality among all treatments (Table 7). In both experiments, poults that died seemed healthy and were not starve-outs before dying. In fact, postmortem evaluation of dead poults revealed no obvious evidence of infection or nutritional disease (Rollins Animal Diagnostic Laboratory, Raleigh, NC; personal communication).

Nutrient Digestibility and Pancreas Weight

Fat and starch digestibility were determined only for experiment 1. The SBL levels included in the PDL and PDH had no effect on fat and starch digestibility. However, the ability of poults to digest starch and particularly fat tended to improve with age in all treatments (Table 8). Fat digested by poults in all treatments improved by 9 to 15% from d 7 to d 14. A lesser magnitude of improvement (3 to 8%) was observed for starch digestibility. Poults in all treatments exhibited a similar capacity to digest fat and starch. There were no differences in the absolute and relative pancreas weights of poults in all treatments. In experiment 1, absolute pancreas weights were 0.39 ± 0.10 on d 6 and 0.90 ± 0.34 on d 14. In experiment 2, pancreas weights were 0.35 ± 0.04 on d 6 and 0.65 ± 0.09 on d 12.

Activity of Brush Border Enzymes

Activities of the brush border enzymes (maltase and sucrase) are presented in Tables 9 to 11. Total activity of each enzyme represented the sum of enzyme activity in the small intestinal segments (duodenum, jejunum, and ileum). Throughout this study, there were no differences in ALP activity.12 On d 6, maltase and sucrase activities were lowest for the PDH treatment compared with the PD and PDL treatments in both experiments (Tables 9, 10, and 11). In contrast, trends in enzyme activities on d 14 in experiment 1 and d 12 in experiment 2 were inconsistent. In experiment 1 (d 14), sucrase activity was significantly higher for the PDH treatment compared with the PD and PDL treatments (Table 10), but in experiment 2 (d 12), there were no observable trends or differences in enzyme levels among treatments.

Concentration of Circulating SBL-Specific Antibodies

A dose-response pattern was observed in the amount of SBL-specific antibodies produced by the poults in response to ingested lectin in experiments 1 and 2 (Table 12). Birds fed the PDH and PDL treatments had a significantly higher SBL-specific antibody concentration in their blood compared with the PD treatment (Table 12).

DISCUSSION

In this study, a corn starch-casein based semipurified diet (control PD treatment) supplemented with graded levels of affinity-purified SBL (PDL and PDH treatments) was fed to turkey poults to investigate the antinutritional effects of SBL. The use of a purified diet and the affinity-purified lectin eliminated the presence of other antinutrients in the experimental diets (PD, PDL, and PDH). The SBD was included as a second control for comparison with the semipurified control diet (PD). This comparison enabled us to determine whether the control PD was as nutritionally adequate as the conventional SBD.

The poorer performance of poults fed the SBD (compared with the control PD; Tables 2 and 3) was probably due to lower nutrient density of the unpelleted (mash) SBD. However, poults fed the SBD overcame the nutrient density effect during the second week as indicated by their having similar weight gain and FE to poults in the PD treatment. As such, results from experiments 1 and 2 indicate that poults fed the control PD grew as well as poults fed the SBD, and that the control PD was nutritionally satisfactory. The high gain:feed ratio obtained with the semipurified casein-cornstarch-lard diet may be attributed to the high digestibility of this diet relative to the SBD.

Performance of poults in this study was similar to previously reported values. Weight gain of poults in the SBD treatment at the end of each experiment (347 g/poult in experiment 1; 289 g/poult in experiment 2) was comparable to the values previously reported (281 to 302 g/poult) for 13-d-old Nicholas poults (Turner et al., 1999a). Similarly, the FE obtained for poults fed the SBD in experiment 1 (0.957 ± 0.035) was comparable to the average value of 1.064 ± 0.006 previously reported (Turner et al., 1999b).

Soybean lectin levels in our PDL and PDH diets closely approximated SBL levels that could reasonably be encountered in commercial diets for starting poults. The amount of lectin remaining in the SBM after oil extraction, desolventization, and toasting of soybeans is referred to as the residual lectin content. The level of SBL present in a commercial starter diet depends on 1) the residual lectin content of the SBM used to make the diet and 2) the level at which SBM is included in the diet. Typically, SBM is included in turkey poult diets at 30 to 50% level of the diet. A survey that focused on the measurement of residual lectin levels in commercial SBM samples in the US revealed that desolventized-toasted meals contained 0.22 to 0.67 mg of lectin/g of meal and that flash-desolventized meals may contain up to 3 mg of lectin/g of meal (Maenz et al., 1999). The lectin inclusion levels in the PDL and PDH were chosen to simulate the range of lectin levels typically encountered in starting diets that contain commercially produced SBM. The lectin level in the PDL diet (0.024%) corresponds to SBM that is included at 40% level of the diet and contains 0.5 mg of lectin/g of meal. The lectin level in the PDH diet (0.048%) corresponds to SBM that is included at 40% level of the diet and contains 1 mg of lectin/g of meal. Results from this study (Tables 2 and 3) showed that the PDL treatment was at least similar to the control PD in weight gain and FE, therefore indicating that the SBL level in the PDL had no significant detrimental effect on poults less than 2 wk of age.

Mortality recorded in this study would have been attributed to the dietary SBL levels, but the SBL levels in the PDL and PDH treatments did not have a consistent effect on poult mortality (Tables 5 and 7). This implies that the exceptionally high mortality in PDL treatment of experiment 1 (Tables 4 and 5) was a random event. A more plausible explanation for the mortalities that occurred in this study would be the effect of diet type (semipurified vs. practical), because all birds that died in both experiments were from the semipurified diet treatments (PD, PDL, and PDH).

The semipurified diets contained about 71 % starch by analysis compared with the SBD (about 61%). The presence of indigestible cellulose and the higher level of corn starch in the semipurified diets compared with the SBD (Table 1) could have contributed to mortality. Corn starch contains low levels of amylopectin (1.1%) and a higher proportion of amylose (20.1%), which is the less digestible starch (White, 1994). Because the intestine of poultry is not fully developed until 10 to 14 d posthatch (Sell, 1985; Sell and Angel, 1990), exposure of the large intestine to considerable amounts of undigested starch and cellulose could result in rapid fermentation that exceeds the absorptive process for short chain fatty acids (Caspary, 1992). Poor absorption of short chain fatty acids could result in acidic fecal pH and increased water and electrolyte loss (Caspary, 1992), which in turn may cause osmotic diarrhea and alter electrolyte balance (Sell and Angel, 1990). Osmotic diarrhea and altered electrolyte balance could culminate in poult mortality. In this study, the feces of poults fed the semipurified diets were watery and sticky, indicating that these poults had some degree of osmotic diarrhea during the first week of life. However, by the end of the first week, poults consuming the semipurified diets (PD, PDL, and PDH) seemed to overcome the detrimental effects of high dietary starch and cellulose as indicated by a considerable reduction in mortality during the second week in both experiments (Tables 4 and 6). This reduction in mortality during the second week may be due to improved development or maturation of the absorptive process for short chain fatty acids in the large intestine to match the rate of fermentation of undigested starch and cellulose.

The higher total mortality in experiment 1 (14.7%) compared with experiment 2 (5.2%) was probably caused by unknown differences in the composition of dietary ingredients. Variation probably existed in the chemical composition of ingredients used for each experiment because different batches of the major ingredients (corn starch, casein, and lard) were used to prepare the diets for each experiment. Another factor that could have influenced mortality levels is that the diets for each experiment were pelleted separately. Thus, it is possible that there were differences in the pelleting conditions (temperature, time, amount of steam) that resulted in higher digestibility of starch present in the semipurified diets used in experiment 2. This could have reduced the degree of osmotic diarrhea experienced by these poults, thereby causing a subsequent reduction in poult mortality.

It is well documented that the ability of young poultry to digest fat increases with age. Therefore, the apparent improvement in the ability of poults to digest fat with age in this study is probably due to the increase in bile salt secretion, lipase activity, and fatty acid binding protein in the intestine with age (Sell et al., 1986; Turner et al., 1999b). In contrast, the ability of poults to digest starch over the 2-wk period did not change much. A similar observation was reported by Noy and Sklan (1995). Noy and her colleague did not find a large change in starch digestion in chicks between 4 and 21 d of age. Moran (1985) and Jin et al. (1998) concluded that the early posthatch increase in pancreatic amylase secretion results in adequate intestinal amylolytic activity for starch breakdown such that chicks (and perhaps poults) are fully competent in starch digestion shortly after hatching.

Lectins are known to increase pancreatic enzyme output and pancreas weight (liener, 1994a,b). Specifically, SBL induces pancreas hypertrophy by binding to small intestinal neuroendocrine cells, thereby releasing cholecystokinin, which in turn stimulates accumulation of polyamines in the pancreas (Grant et al., 1989a,b; Pusztai et al, 1995). These polyamines (putrescine, spermidine, and spermine) are biogenic amines that are involved in promoting cell division, protein synthesis, and tissue growth (Seiler, 1992), therefore causing a subsequent increase in pancreas weight. Grant et al. (1987, 1989b) observed up to 35% increase in pancreas dry weight after feeding rats a lactalbumin-based semipurified diet supplemented with 0.75% SBL. In experiments 1 and 2, trophic changes were not observed in the pancreas weights of poults fed the lectin-supplemented diets. Perhaps the SBL levels used in these experiments (0.024% in PDL; 0.048% in PDH) were too low to induce pancreatic hypertrophy.

Brush border enzymes (such as sucrase-isomaltase, aminopeptidases, and ALP) are synthesized by villiattached enterocytes and inserted into the apical membranes during the process of enterocyte differentiation (Uni, 1999). These enzymes are responsible for the final stages of digestion of macromolecules and are important in regulating the amount of nutrients available for absorption, nutrient transport from intestine, reception of signals into cells, and regulation of cell growth and differentiation (Kenny, 1986; Iji et al, 2001; Sklan, 2001). Because the brush border enzymes are specific for enterocytes and their levels increase as enterocytes mature, they can be used as markers of intestinal (or enterocyte) maturity (Ortega et al., 1995; Uni, 1999). Activities of the brush border enzymes are reduced when the BBM is damaged (Hong et al., 1991). Lectins can bind to their specific carbohydrate receptors on the intestinal surface, crosslink the receptors and cause the disruption of BBM and abnormal development of intestinal microvilli (liener, 1994b). This lectin-induced disruption of the BBM could result in loss of brush border enzymes, and subsequent reduction in enzyme activity.

Measuring the activities of ALP, maltase, and sucrase in the small intestinal epithelium enabled us to assess the extent of membrane disruption caused by SBL levels in PDL and PDH. The lower total activities of maltase13 and sucrase in the PDH compared with the PD and PDL during the first week (Tables 9,10, and 11) was possibly due to disruption of the BBM by the SBL added to the PDH. However, the degree of membrane disruption caused by this level (0.048%) of dietary lectin did not compromise the ability of the poults to digest and assimilate nutrients efficiently (Table 8) nor did it affect growth (Tables 2 and 3).

The disparity observed between the trends in enzyme activities in experiment 1 (d 14) and experiment 2 (d 12) could be due to the smaller number of poults sampled per treatment in experiment 2 (4 poults per treatment) compared with experiment 1 (8 poults per treatment). Alternatively, because different strains of poults were used in each experiment (Hybrid Converter strain in experiment 1; Nicholas 88 strain in experiment 2), it could be that poults of each strain responded differently to SBL. With age, birds of different strains could have variability in the number of lectin-specific glycosyl receptors available for lectin binding (Pusztai and Bardocz, 1996). Soybean lectin binds specifically to N-acetylgalactosamine and galactose-containing receptors in the intestinal mucosa (Pereira et al., 1974). Availability of the appropriate glycosyl receptors and binding of the lectin to the BBM are prerequisites for the lectin to alter gut structure and function (Gueguen et al., 1993). Therefore, the lectin-induced increases in disaccharidase activities on d 14 in experiment 1 could be due to availability of higher numbers of SBL-specific glycosyl receptors in Hybrid poults (experiment 1) compared with Nicholas poults (experiment 2) during the second week.

The mechanism by which SBL induced higher disaccharidase activity in poults fed PDH during the second week of experiment 1 is probably based on lectin-induced changes in the process of enterocyte differentiation. Enterocyte differentiation is the process by which immature cells proliferating in the crypts migrate along the villi and differentiate into enterocytes that acquire specialized functions, including nutrient digestion, absorption, and mucin secretion (Uni, 1999). The differentiation process along the crypt-villus axis involves different biochemical events that result in 1) the expression of the microvillus structure to increase villi surface area, 2) the expression of brush border enzymes, and 3) surface expression of carrier molecules and transport mechanisms that enable the transport of digestion products and antigens across the intestine (Smith, 1993). Perhaps SBL directly induced changes in the biochemical events occurring during differentiation in a manner that resulted in the expression of higher disaccharidase activities in the BBM. For instance, Otte et al. (2001) stimulated human (intestine-407) and rat (IEC-6; IEC-18) small intestinal cell lines with phytohemagglutinin and found that the lectin activated the mitogen-activated protein kinase and induced c-fos mRNA expression. The mitogen-activated protein kinase cascade is known to transmit transmembrane signals required for cell growth and differentiation by phosphorylating and activating nuclear binding proteins such as c-fos and c-jun, which in turn modulate transcription of target genes (Seger and Krebs, 1995).

Alternatively, SBL could have indirectly altered the differentiation process by triggering the unregulated release of hormones or peptides that affect enterocyte differentiation. For example, kidney bean lectin (phytohemagglutinin) and SBL are known to bind to neuroendocrine cells of the small intestine, thereby releasing cholecystokinin, which in turn induces an increase in the secretion of pancreatic digestive enzymes (Pusztai, 1991; Jordinson et al., 1996). Some hormones, gastrointestinal peptides (e.g., epidermal growth factor), and cytokines have been implicated as factors that influence the process of enterocyte differentiation (Smith, 1993; Johnson and McCormack, 1994; Uni et al., 2001). Thus, we hypothesize that SBL induced higher disaccharidase activity on d 14 of experiment 1 in poults fed PDH by binding to intestinal enterocytes in these poults, thereby triggering the release of specific hormones or peptides that are capable of altering the differentiation process in a manner that resulted in the expression of higher levels of disaccharidases. Furthermore, the SBL-induced changes during differentiation could have increased the absorptive capacity of the enterocytes such that the increased enzyme levels combined with the increased absorptive capacity culminated in a higher FE for poults fed the PDH on d 14, compared with poults fed the control diet (PD) (Table 2).

The SBL-specific antibodies detected in the serum of poults in all treatments in both experiments (Table 12) were probably produced in response to SBL present in the digestive tract, and to SBL that was endocytosed and entered the systemic circulation (Shanahan, 1994). The antibody response elicited was such that the SBLspecific antibody levels increased with increasing levels of dietary SBL (Table 12). This is an indication that the poults used in this experiment were sensitive to the graded levels of SBL used in this study, and that SBL remained active in the digestive tract. We did not expect to detect SBL-specific antibodies in the serum of poults fed the control PD because there was no SBL in that diet. As such, the circulating SBL-specific antibodies present in the serum of poults in the PD treatment were probably elicited by other antigenically related proteins, which entered the poults from dietary or nondietary (e.g., inhaled antigens) sources (Tchernychev and WiIchek, 1996).

A number of authors have implicated SBL in causing antinutritional effects such as reduced nutrient digestibility, hypertrophy of pancreas, increased endogenous nitrogen loss, and reduced growth in animals (Pusztai, 1993; liener, 1994a; Grant et al, 2000). For instance, Grant (1989) reported that inclusion of 0.7% SBL in a lactalbumin-based diet for rats resulted in a considerable reduction in BW. Schulze et al. (1995) observed that supplementing a basal diet with 0.016 and 0.096% SBL increased endogenous nitrogen flow at the terminal ileum of pigs by 30 and 47%, respectively. However, SBL did not reduce weight gain or feed intake of poults when present at levels up to 0.024% of the diet in this study. In addition, the 0.024% SBL level did not significantly alter brush border enzyme activities nor did it compromise the integrity of the BBM. We did not observe most of the known detrimental effects of lectins (especially at the 0.024% dietary level), probably because animal species (turkey poults vs. pigs and rats) respond differently to ingested lectins (Pusztai and Bardocz, 1996). Furthermore, the SBL levels used in this study (0.024% in PDL; 0.048% in PDH) were probably lower than the amount needed to cause various antinutritional effect(s) in turkey poults.

In summary, this study investigated the antinutritional effects of SBL in young turkey poults at dietary levels that approximate those encountered in commercial poult diets. We incorporated the affinity-purified SBL at 0.024 or 0.048% level of a corn starch-casein based semipurified diet to produce the following experimental diets: a corn starch-casein based (lectin-free) semipurified control diet (PD), a semipurified diet containing 0.024% SBL (PDL), and a semipurified diet containing 0.048% SBL (PDH). The use of affinity-purified SBL and a semipurified diet eliminated the presence of other antinutrients in these experimental diets. As measured by ELISA, levels of circulating SBL-specific antibodies present in the serum of poults at the end of each experiment indicated that the SBL incorporated into our experimental diets remained active in the digestive tract. Results obtained indicated that throughout this study, the SBL level in the PDL (0.024%) did not cause any detrimental effect regardless of poult strain differences, whereas the SBL level in the PDH (0.048%) reduced the activities of brush border enzymes during the first week. We therefore concluded that dietary SBL levels up to 0.024% are safe for turkey poults up to 2 wk of age. This study is the first to determine a safe margin of SBL that should be allowed in diets for turkey poults. The 0.024% lectin level in the PDL diet corresponds to commercial soybean meal that contains 0.5 mg of lectin/g of meal, and is included at 40% of the diet. Thus, soybean products intended for feeding to turkey poults should be routinely monitored to contain less than 0.5 mg of lectin/ g of meal.

ACKNOWLEDGMENTS

The authors of this paper express their appreciation to Carole Morris, Annette Israel, Debbie Ort, Riswana AIi (all of Department of Poultry Science, NC State University), and Dawn Abbott (Department of Animal and Poultry Science, University of Saskatchewan, Canada) for the technical support rendered during this study.

Copyright Poultry Science Association Sep 2004

Provided by ProQuest Information and Learning Company. All rights Reserved.