INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE NUTRITIONAL IMPORTANCE of legume seeds has received considerable attention in recent years. In addition to the potentially beneficial effects of bean consumption (17, 45), legumes represent a more economic source of dietary protein than those of animal origin. In particular, faba bean is one of the oldest crops and ranks sixth in production among the different legumes grown in the world. It is widely grown for human consumption throughout the Mediterranean region, in Ethiopia, and in parts of Latin America. In the developed world, such as Europe and North America, it is mainly grown for animal feeding (9). However, the presence of a number of antinutritional factors in the raw seed meal has limited the application of faba beans as an animal foodstuff (40, 46).

The skeletal musculature undergoes rapid and extensive adaptive changes in response to mechanical, nervous, nutritional, hormonal, and pharmacological factors (5, 18, 36), and this has made muscles a focus for the identification of pathophysiological stimuli that influence tissue growth and differentiation. It is well recognized that muscle protein synthesis in young rats is very sensitive to food intake (e.g., Ref. 18). Previous studies have shown a reduced growth performance accompanied by a decreased protein synthesis in gastrocnemius muscles of growing animals that had been fed diets in which raw faba beans [Vicia faba L. (Vf)] were the sole or main source of protein (22, 33). However, the response of metabolically different muscle types has been less extensively investigated. Consequently, the present study in growing rats was undertaken with three objectives: 1) to investigate the effect of raw field bean intake on the fractional rates of protein synthesis (k_s), degradation (k_d), and growth (k_g) of predominantly fast- and slow-twitch hindlimb muscles; 2) to examine, apparently for the first time, the changes of Vf intake on muscle fiber type composition; and 3) to examine any relationships between blood hormonal changes elicited by Vf consumption and muscle protein metabolism and fiber type profiles.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Legume flour analysis. The guidelines of the Association of Official Analytical Chemists were followed in the analyses of the field bean meal (4) (Table 1). Carbohydrate analysis was performed as described by Rubio et al. (40). The amounts of free sugars were estimated by HPLC after extraction with boiling ethanol, and the starch content was determined by enzyme digestion. Samples were treated with dimethyl sulfoxide to solubilize starch. The starch was hydrolyzed with -amylase (EC 3.2.1.1), pullulanase (EC 3.2.1.41), and pancreatin and nonstarch polysaccharides precipitated by ethanol. After separation by centrifugation, the nonstarch polysaccharides were dispersed in 12 M sulfuric acid, diluted to 1 M, and hydrolyzed. The constituent sugars were determined by gas-liquid chromatography as their alditol acetate derivatives. Uronic acids were determined colorimetrically. Fatty acid composition was determined as reported by Hendrikse and Harwood (25). The lipid content of the flour was determined by conversion of fatty acids to the corresponding fatty methyl esters, which, after being dissolved in heptane, were determined by gas chromatographic analysis. The amino acid profile was assessed by HPLC (6). Before acid hydrolysis, cysteine and methionine were oxidized to cysteic acid and methionine sulfone, respectively, to determine them (Table 2).

Animals and diets. Three-wk-old male Wistar growing rats (CIFA, Pamplona, Spain), weighing 75 ± 5 g, were assigned to three different experimental groups of 27 animals each [control (Con), field bean (Vf)-fed, and pair-fed controls (PF; given the control diet in the same amount of food intake displayed by the Vf-fed animals)]. Previous studies have shown that, compared with well-fed animals, rats consuming Vf as the sole source of protein display a reduced food intake (22, 33). To compensate for this effect, the PF group was included. Rats were maintained under controlled conditions of room temperature (20 ± 2°C), relative humidity (50 ± 10%), ventilation (at least 15 complete changes of air/h), and artificial light-dark cycle (light from 800 to 2000). Animals had free access to daily renewed tap water and over the experimental period were fed isoenergetic (16.72 ± 0.49 MJ/kg diet) and isoproteic (11%) diets containing either lactalbumin (Con and PF groups) or raw field bean (Vf-fed group) as the source of protein (Table 3). The digestibility coefficient of lactalbumin and Vf proteins used in the diets was 93 and 86%, respectively (16). Furthermore, diets provided 69% of energy as carbohydrate and 20% as fat. After a 3-day adaptation time, most of the rats were fed over a 15-day period. The experimental periods of the animals destined to determine skeletal muscle k_g comprised 13, 14, 16, or 17 days. All experimental procedures were performed according to institutional guidelines for animal care and use at the University of Navarra.

Determination of muscle k_s, k_d, and k_g. The k_s was determined by the flooding dose method of Garlick et al. (20) as modified by Martínez (32). On the last day of the experimental period, food was withdrawn from the rats 6 h after onset of the dark cycle. Thus animals had a 6-h feeding period and were killed in the morning after an 8-h fast. Ten rats from each group were intraperitoneally injected with 1 ml/100 g body wt of a solution containing 1.84 MBq/ml of L-[ring-2,3,4,5,6-³H]phenylalanine (DuPont) combined with 150 mM unlabeled phenylalanine (Sigma, St. Louis, MO). After exactly 10 min, the animals were decapitated, and the trunk blood was immediately collected. Soleus (Sol) and extensor digitorum longus (EDL) muscles were quickly and bilaterally removed, weighed, and frozen in liquid nitrogen. Muscles weighing <50 mg were pooled with those of the contralateral hindlimb of the same animal.

Histochemical assessment of muscle fiber types. Transverse sections (5 mm) were taken from the midbelly of Sol and EDL muscles of five rats per group. Care was taken to ensure that the same region in each muscle type was taken from all animals. The muscle pieces were oriented for transverse sectioning, mounted on cork, covered with Tissue-Tek (Miles, Diagnostics Division), immediately immersed in isopentane (2-methyl butane) cooled by liquid nitrogen, and stored at 80°C until they were assayed. Serial cross sections of the muscle samples were cut at 10-µm thickness with a cryostat (Reichert-Jung, Slough, UK) kept at 20°C and were stained for Ca²⁺-activated myofibrillar ATPase (mATPase) after preincubation at various pH levels (24, 28). Muscle fibers were classified on the basis of differences in staining intensity for mATPase after alkaline (pH 9.4) preincubation. Type I (slow-twitch oxidative) fibers stain light for mATPase, whereas type II (fast-twitch) fibers stain dark. The type II fibers were further subclassified by acid preincubation at pH 4.6 and 4.2, according to their staining reaction, into IIa, IIb, and IIc fibers. At least 200 fibers per muscle were examined, and the percentage of frequency of each fiber type was calculated. The cross-sectional area (CSA) of each fiber was determined from nondehydrated sections with the use of computerized planimetry, with the system calibrated by a stage micrometer immediately before measurement.

Hormonal analysis. Blood for measurement of hormone concentrations was obtained after an 8-h overnight fast. Serum insulin was determined by RIA by using a commercially available kit for rats (Amersham; intra- and interassay coefficients of variation were 5.1 and 13.1%, respectively). A double-antibody RIA method was used to measure plasma glucagon (Diagnostics Products; intra- and interassay coefficients of variation were 8.1 and 9.3%, respectively). Quantitative determination of both serum thyroxine and 3,5,3',-triiodothyronine (T₃) levels was also performed by using RIA kits (CIS Bio International; intra- and interassay coefficients of variation were 3.2 and 3.9% for thyroxine and 3.3 and 5.0% for T₃, respectively). Serum thyrotropin (TSH) assessment was carried out by using a commercially available RIA kit for rats (Amersham; intra- and interassay coefficients of variation were 4.8 and 13.2%, respectively). Finally, a third RIA commercial kit, specifically designed for rodents, was used for the quantitative determination of serum corticosterone (CCT) (DRG Instruments; intra- and interassay coefficients of variation were 4.4 and 6.5%, respectively).

Statistical analysis. The means ± SE are reported for all measurements. Data were analyzed by one-way ANOVA followed by Duncan's tests for multiple comparison. Pearson's product-moment correlation coefficient (r) was used to study the relationship between blood hormone concentrations and muscle protein metabolism data. Simple linear regression analysis was carried out as an extension of correlation analysis as it examines the relationship between one explanatory and one response variable, or the tendency of one variable to change with the other (21). All analyses were performed with the SPSS/Windows version 6.1.3 (SPSS, Chicago, IL) and the StatView/Apple Macintosh version 4.01 non-FPU (Abacus Concepts, 1992-1993) statistical packages. Differences were considered significant with a P value at the 5% level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Weight changes. After the end of the experimental period, both Vf-fed and PF rats showed a statistically significant (P < 0.001) growth retardation accompanied by decreased food intake (Table 4). The gain-to-feed ratio of rats fed on the diet containing Vf meal as the only source of protein was inferior to that achieved by rats in the ad libitum Con group (P < 0.001). Despite the similar food intake between the rats in the Vf-fed and PF groups, inclusion of Vf meal in the diet had a profound effect on the nutritional performance of the animals with a reduced gain-to-feed ratio (P < 0.01), leading to poor weight gain (P < 0.05), compared with PF rats. No differences were observed in the total intake of essential amino acids between Vf-fed and PF groups. However, the legume-fed animals consumed almost 41% less sulfur-containing amino acids than did the PF rats on the lactalbumin-based diet (Table 5).

Muscle protein turnover. Compared with Con animals, EDL muscles of Vf-fed rats displayed a statistically significant (P < 0.01) reduction in k_s. The fact that k_g was even more reduced in the EDL muscles of these animals resulted in a statistically significantly (P < 0.01) reduced k_d (Table 6). On the contrary, no significant differences in the growth rates of EDL muscles of Vf-fed and PF rats were evident after the end of the experimental period. The diet restriction imposed on the PF group resulted in a decrease (P < 0.05) in k_d of Sol muscles compared with that of the legume-fed rats. In addition, k_s values in the slow-twitch muscles of PF rats were significantly (P < 0.05) decreased compared with those displayed by Vf-fed rats.

Fiber type analysis. As shown in Fig. 1, fiber type frequency data suggest that hindlimb muscles do not attain a static pattern of fiber type composition after a 15-day period of consuming Vf as the sole source of protein. Compared with the PF rats, consumption of the Vf diet had a significant impact on the fiber type profiles of Sol muscles. The percentage of type IIc fibers significantly increased (P < 0.01) and the number of type I fibers decreased by 17% (P < 0.05) in the Sol of Vf-fed rats. On the contrary, the fiber type profile of EDL muscles of rats on the legume-based diet and of the PF rats showed no significant differences (Fig. 2). In EDL muscles of Vf-fed and PF rats, a statistically significant increase (P < 0.01) in type I and IIc fibers concomitant with a decrease (P < 0.05) in type IIb fibers was observed, compared with the Con group. Concerning the CSA of the fibers, Figs. 3 and 4 show that no statistically significant differences in any of the fiber types of both the Sol and EDL muscles, respectively, in legume-fed and PF animals were evident.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Fiber type composition of soleus muscles of growing rats in control (Con), field bean (Vf)-fed, and pair-fed (PF) groups. Values are means ± SE of 10 animals/group (1-way ANOVA followed by Duncan's tests for multiple comparison). ** Significantly different from Con group, P < 0.01; significantly different from Vf-fed group: # P < 0.05, ## P < 0.01.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Fiber type composition of extensor digitorum longus muscles of growing rats in Con, Vf-fed, and PF groups. Values are means ± SE of 10 animals/group (1-way ANOVA followed by Duncan's tests for multiple comparison). Significantly different from Con group: * P < 0.05, ** P < 0.01.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Fiber cross-sectional area (in µm2) of soleus muscles of growing rats in Con, Vf-fed, and PF groups. Values are means ± SE of 10 animals/group (1-way ANOVA followed by Duncan's tests for multiple comparison). ** Significantly different from Con group, P < 0.01.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Fiber cross-sectional area (in µm2) of extensor digitorum longus muscles of growing rats in Con, Vf-fed, and PF groups. Values are means ± SE of 10 animals/group (1-way ANOVA followed by Duncan's tests for multiple comparison). Significantly different from Con group: * P < 0.05, ** P < 0.01.

Hormonal concentrations. Hormone-induced modifications observed after the conclusion of the experimental period reveal that Vf-fed rats exhibited a significant decrease in serum insulin (P < 0.05) and TSH (P < 0.01) levels, together with an increase in plasma glucagon (P < 0.05) and T₃ (P < 0.01) concentrations, compared with both ad libitum Con and PF rats (Table 7). No significant changes in blood hormone concentrations were observed between the Con and PF groups, except for the statistically significant (P < 0.01) increase in CCT levels of PF animals. Both Vf-fed and PF rats experienced an increase in CCT concentrations; however, the increment in CCT values displayed by rats fed on the legume-based diet was significantly higher (P < 0.05) than that observed in PF rats.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A considerable amount of experimental work on protein metabolism in mammals has been performed in immature growing animals. These show high rates of protein turnover compared with the net rate of growth, giving the animal the scope for large changes in response to acute stimuli (18, 19). The rate of muscle growth reflects both the adequacy of the diet to provide substrates for growth as well as its ability to evoke a regulatory response that activates anabolic and/or catabolic processes. The present study shows that, compared with Con animals, rats fed the legume-containing diet exhibited a significant decrease in weight gain and muscle mass at the end of the 15-day experimental period. The observed growth retardation, which agrees with previous findings (1, 22, 33), cannot be ascribed only to a decreased food intake, as the effects in the PF group are not as dramatic as those of Vf-fed rats. Thus, although the legume diet was equalized in energy and protein content with that of the controls, weight gains and gain-to-feed ratios obtained in the faba bean group were inferior to those of the ad libitum Con and PF rats. In this context, the quality of milk-derived and legume proteins has to be taken into account. It is well known that bean proteins have a low content of sulfur-containing amino acids (12). Because of the amino acid profile of Vf, rats fed on the legume diet consumed a considerably smaller amount of sulfur amino acids than did the PF rats. Furthermore, it has been reported that the lower efficiency of protein utilization in rats fed on legume-containing diets is due to lower net nitrogen absorption from the small intestine (41).

The protein synthesis and degradation measurements are supportive of the changes observed in muscle masses. Therefore, field bean intake appears to shift protein turnover in EDL muscles toward impaired growth, despite decreasing proteolysis, by significantly reducing protein synthesis. Furthermore, the data herein show that the effects on Sol muscles seem to be specific to Vf-fed rats because they are still evident when compared with PF rats, whereas the effects on EDL muscles seem to be due to diet intake restriction because no differences are evident between the Vf-fed and the PF groups.

The differential response observed in muscle protein metabolism may be explained by the hormonal changes, which take place in the animals consuming the Vf diet. Bean constituents have been reported to influence and modulate the endocrine status (39). Diets rich in leguminous seeds are associated with improved control of carbohydrate metabolism (27). Reduced insulin concentration occurs as a consequence of delayed and diminished postprandial glucose absorption. Decreases in insulin-to-glucagon ratios have been observed after the feeding of soy protein (23). Of the starchy foods, legumes generally elicit the lowest postprandial responses of glucose and insulin (27). These characteristics have promoted the use of whole beans for the dietary management of diabetic patients.

Protein metabolism and hormonal status are affected by the amino acid composition and/or proportionality of the diet. Analyses of the Vf amino acid profile revealed that the lysine/arginine ratio was low, compared with animal proteins (0.57 in Vf vs. 1.22 in lactalbumin). In this context, low lysine/arginine ratios have been associated with stimulation of glucagon and the inhibition of insulin secretion (42). The present study shows that, after a relatively short experimental period (15 days), intake of Vf as the sole source of protein is followed by a significant decrease of insulin levels together with an increase in glucagon concentrations.

It has been observed that fiber type composition of the muscle is clearly an important factor in determining the magnitude of response to plasma insulin concentrations, with only mixed or fast-glycolytic fiber muscles showing lowered k_s values under conditions of decreased insulin levels (5). The lack of an observed effect on the Sol may rest with the fact that this slow-twitch muscle increases both the number and affinity of its insulin receptors in response to hypoinsulinemia (2). It has been also shown that elevated plasma glucagon concentrations result in decreased rates of protein synthesis in plantaris and gastrocnemius muscles, but not in Sol (37), and that hyperglucagonemia during decreased insulin concentrations accelerates protein catabolism (35).

In addition to the effects of insulin and glucagon on muscle protein metabolism, the influence exerted by the increased CCT concentrations observed in Vf-fed rats also requires consideration. The diet-related stress associated with the intake of either a poor-quality protein in the legume-fed group or the restraint feeding in the PF group probably accounts for the increased CCT concentrations observed in both Vf-fed and PF animals. The major muscular effects of glucocorticoid elevation have been attributed primarily to a decrease in the rate of protein synthesis, concomitant with increased protein catabolism (30). It is interesting to note that, whereas Vf-fed rats showed a statistically significant decrease in plasma insulin together with an increase in glucagon, T₃, and CCT concentrations, PF animals, however, showed only a significant elevation in CCT levels. This different hormonal response between the Vf-fed and PF groups may be responsible for the more pronounced catabolic effect observed in rats fed the bean-containing diet.

The mechanisms underlying the selective morphological adaptations of different fiber types in response to nutritional stimuli still remain unclear. Inclusion of field beans as a dietary protein source induces a type I-to-type IIc transformation in the Sol, whereas a lower percentage of type IIb, which was compensated for by an increased percentage of IIc and I fibers, occurred in the EDL. Our study suggests the possibility of participation via hormone-induced changes. In this sense, heterogeneity between fast- and slow-twitch muscles in terms of insulin response and sensitivity has been demonstrated, so that slow-twitch muscles show a higher degree of insulin sensitivity and a greater maximal response to insulin than do the fast-twitch muscles (26, 43).

Forsythe (14) has been able to show a causative relationship between thyroid hormone increase and vegetable protein intake. Furthermore, the thyroid state of the animal has the potential for determining the phenotype of skeletal muscle myosin. T₃ stimulates the development of fast-twitch fiber characteristics. This is particularly evident in slow-twitch muscles, such as the Sol, where T₃ has been shown to induce the replacement of type I with type II fibers (8, 44). Thus the statistically significant increase in T₃ may account, at least in part, for the muscle fiber type shift observed in the Sol of Vf-fed rats.

It has been well documented that fast-twitch fibers are more affected than slow-twitch fibers in response to glucocorticoids (e.g., Refs. 3, 13). Selective atrophy of type IIb fibers has been reported to be associated with increased sensitivity to circulating glucocorticoids (10) because of an upregulation of cytosolic glucocorticoid receptors (31). The results of the present study point to the possibility of a corticoid-induced selective atrophy of type IIb fibers in EDL muscles for both the Vf-fed and PF animals. Together with the atrophy of type IIb fibers in EDL muscles, a decrease in the CSA of these fibers takes place, which contributes to the reduction in muscle mass observed.

In conclusion, the existence of different adaptive and regulatory mechanisms operating distinctly on fast- and slow-twitch muscles of growing rats fed a diet for 15 days containing Vf as the sole source of protein has been observed. The present data strongly support the possibility that muscle-specific changes in both protein metabolism and fiber type composition depend, at least in part, on the hormonal changes that followed and were induced by field bean intake.