Anti-Ulcerogenic Effect of a Whey Protein Isolate and Collagen Hydrolysates Against Ethanol Ulcerative Lesions on Oral Administration to Rats

Abstract

The effect of the administration of a whey protein isolate (WPI) and collagen hydrolysates on ethanol-induced ulcerative lesions was studied in rats. WPI and bovine or porcine collagen hydrolysate (BCH and PCH, respectively) were given to rats by gavage. In acute experiments, (single-dose) physiological saline (10 mL/kg of body weight) was used as the negative control, and carbenoxolone (200 mg/kg of body weight) was used as a positive control. Ethanol (1 mL per 250-g rat) was also given by gavage. These treatments reduced the ulcerative lesion index (ULI) in a range of 40–77%, depending on the dosage. Some mixtures of WPI with either PCH or BCH provided results that suggested synergisms between WPI and the collagen hydrolysates. For example, WPI/BCH (in the proportion of 375:375 mg/kg of body weight) decreased ULI by 64%. The mechanism for mucosal protection involved a decrease in plasma gastrin (∼40%), a significant increase (50–267%) in mucus production, and a reduction in ULI (percentage) when intragastric administrations were performed after in vivo alkylation by N-ethylmaleimide. Results suggest that gastrin, sulfhydryl substances, and some mechanisms related to mucus production are all involved in gastric ulcer protection against ethanol. The collagen hydrolysates (both PCH and BCH) presented a stronger effect on mucus production; on the other hand, the effect of WPI was also dependent on sulfhydryl compounds, resulting in a more protective effect when the two proteins were administered together.

Introduction

T he gastric mucosa is one of the most important tissues in the organism, because of its function, structure, and the pathological processes that can take place in this tissue.¹ It has been accepted that peptic ulcerogenesis (gastric and duodenal) results from an imbalance between infectious agents, such as the bacterium Helicobacter pylori, and/or aggressive chemical agents and stresses versus protective substances, such as proteins, glutathione, other sulfhydryl group-containing substances, mucus, bicarbonate, prostaglandins (E₂, I₂), and certain types of dietary fibers, as well as the blood flow to the mucosal cells.^2

–7

Absolute ethanol is an ulcerogenic agent that destroys mucosal cells upon direct contact,⁵ independently of gastric acidity. Ethanol initially destroys the stomach-protecting layers and finally reaches the mucosal cell surface, causing cell necrosis and liberation of vasoactive mediators, leading to vasoconstriction, edema, and hemorrhage.⁵

Rosaneli et al.^8,9 demonstrated, in rats, the anti-ulcerative properties of a whey protein concentrate, produced in a pilot plant,¹⁰ against two ulcerogenic agents, ethanol and indomethacin. It has been recently shown^11
–13 that one of the most anti-ulcerogenic proteins in bovine milk whey is α-lactalbumin; in contrast, β-lactoglobulin has no antiulcerogenic effect.¹³ Collagen hydrolysates have been reported to be efficiently absorbed¹⁴ and capable of stimulating cell regeneration and extracellular matrix production, as well as cytokine modulation.^15
–17 Interestingly, Santos¹⁸ described the action of collagen, extracted from bovine tendons (type I collagen), on the recovery of duodenal mucosa lesions caused by Phaseolus vulgaris lectins.

A recently published study¹⁹ demonstrated that both bovine and porcine collagen hydrolysates (BCH and PCH, respectively) were effective in protecting gastric mucosa when given to rats prior to intragastric administration of absolute ethanol. Bovine collagen did not present a dose–response correlation in the ethanol model, whereas porcine collagen showed a logarithmic dose–response relationship. BCH decreased the ulcerative lesion index (ULI) by 55%, versus a 61% decrease for PCH at the same dosage (750 mg/kg of body weight). No significant differences were found (P > .05) between the hydrolysate and its fractions. Therefore, the present study focused on combinations of whey protein isolate (WPI) with BCH at the lowest concentration (200 mg/kg of body weight) and with PCH at the best concentration that was previously found (750 mg/kg of body weight).

The main objectives of the present investigation were (1) to comparatively study the anti-ulcerogenic efficacy of two commercial collagen hydrolysates and a WPI on the inhibition of gastric mucosa lesions induced by intragastric administration of absolute ethanol to rats, (2) to investigate the possible synergistic action of various combinations of WPI and collagen hydrolysates on their gastric mucosa protection against ethanol lesions, and (3) to investigate the mechanisms by which WPI and collagen hydrolysates protect the gastric mucosa from ethanol. This study may provide information to suggest the application of WPI and collagen hydrolysates as a nutritive supplement that may contribute to prevent gastric ulcers.

Materials and Methods

Collagen hydrolysates

BCH and PCH were derived from commercial production lines (lots EH4517 and PL-040701, respectively) and were supplied by Gelita (Gelita South America), Cotia, SP, Brazil, in the form of spray-dried powder. The gross percentage composition for both hydrolysates was 91% protein, 0.39% minerals, 0.09% fat, and 1.51% carbohydrate.

Production of WPI

WPI was produced in a pilot plant from defatted and microfiltered milk (1–1.4-μm pore size membrane) and then concentrated by the use of a ceramic membrane (0.1–0.14-μm pore size; aluminum zirconium) to separate casein micelles from milk whey. The milk whey was then concentrated by ultrafiltration (8-kDa cutoff membrane) and dehydrated by freeze-drying.²⁰

Experimental ulcerogenesis

Male Wistar rats (250 ± 12.5 g body weight) from the experimental colony of the State University of Campinas, Campinas, SP, Brazil were used. The number of animals used in each test is specified in Results.

The model used absolute ethanol (1 mL per 250-g rat) as the ulcerogenic agent, physiological saline (10 mL/kg of body weight) as a negative control, and carbenoxolone (200 mg/kg of body weight) as a positive control. The experimental samples were given by intragastric intubation in different quantities dissolved in saline. Carbenoxolone was also dissolved in saline and administered by gastric intubation.

Two types of experiments were performed: a single-dose (acute) or a double-dose (repetitive) experiment on two consecutive days. A fluxogram illustrating the general protocol for both single-dose and double-dose experiments is shown in Figure 1. The saline group was subjected to the same procedure applied for the samples. Single-dose (acute) experiments were performed by omitting step 2 of Figure 1. The experimental protocol was approved by the Ethics Committee for Animal Experimentation of the State University of Campinas.

FIG. 1.

Experimental ulcerogenesis protocol. bw, body weight.

Analysis of ulcerative lesions

Ulcerative lesions were quantified by the calculation of a ULI using the sum of the following parameters²¹: loss of normal mucosa morphology, 1 point; mucosa discoloration, 1 point; mucosa edema, 1 point; hemorrhage, 1 point; petechial points (until 9), 2 points; petechial points (>10), 3 points; ulcers (up to 1 mm), n* × 2 points; ulcers (>1 mm), n* × 3 points; perforated ulcers, n* × 4 points (where n* = number of ulcers found).

Percentage reduction in ULI was calculated by the following expression: percentage reduction in ULI = ([mean ULI negative control − mean ULI testing sample]/mean ULI negative control) × 100.

The effective dose for 50% reduction of the ULI (ED₅₀) was calculated by using increasing single doses (200, 250, 750, 1,000, and 1,500 mg/kg of rat body weight). The dose–response relationship was fitted into a logarithmic equation, which permitted the calculation of the ED₅₀ for collagen hydrolysate and WPI.

Gastrin determination

Gastrin was determined according to the protocol proposed in Figure 1 (omitting step 2), using six animals per group. One hour after ethanol administration, blood was collected from ocular plexus, treated with EDTA, and centrifuged (3,000 g for 15 minutes at 4°C), and the plasma was stored (−20°C) until gastrin determination. Gastrin was determined in the rat blood plasma²² using the radioimmunoassay kit supplied by CIS Bio International (Cis Biointernational, Gif sur Yvette, France). Residual radioactivity bound to the tubes was measured for 1 minute in a gamma scintillation counter.

Determination of mucus linked to the gastric mucosa

One hour after administration of ethanol (Fig. 1), given to six animals per group, the mucus glandular region of the stomach mucosa was removed, immersed in 10 mL of 0.1% Alcian blue solution for 2 hours, and then washed (twice) with 10 mL of 0.25 M sucrose solution for 15 minutes and 45 minutes, respectively. The dye complexed with the gastric mucus was extracted (2 hours) with 10 mL of 0.5 M MgCl₂ solution. The extracted material was emulsified with 10 mL of ethyl ether and then centrifuged (1,300 g for 15 minutes at 4°C). The ether phase was discarded, and the absorption of the aqueous phase was read in a spectrophotometer at 598 nm.²³ Linked Alcian blue was quantified (in g of dye/g of original tissue) based in a dye standard curve.

Sulfhydryl substances and the gastric mucosa protection

Rats weighing 250 ± 12.5 g (six animals per group) were subjected to treatments by gastric intubation, according to the protocol of Figure 1. In this series of tests, two controls were used: (1) physiological saline at 10 mL/kg of body weight and (2) saline (10 mL/kg of body weight) plus 10 mg/kg of body weight N-ethylmaleimide (NEM). In both controls, saline (control 1) was given by gastric intubation, whereas NEM (control 2) was administered by subcutaneous injection.

Half of the remaining groups (experimental) received only the testing samples, whereas the other half received the testing samples plus NEM. The remaining procedures were identical to those of the protocol in Figure 1.

Statistical analysis

Results are presented as mean ± SD values. The experimental results were subjected to analysis of variance, and the criteria for critical statistical significance was 5% probability (P < .05). Comparisons of means were made by Tukey's or Dunnett's test.

The dose–response relationship was established by adjustment of the experimental points to a logarithmic correlation equation. The equations for porcine collagen (y = 21.857 ln x − 87.561, R ² = 0.9716) and for WPI (y = 21.652 ln x − 78.969, R ² = 0.9776) were established where y = percentage reduction ULI and x = mg of tested sample.

Results

The mathematical model established for porcine collagen and WPI allowed the calculation of logarithmic equations to obtain an ED₅₀ of 541 mg/kg of body weight for porcine collagen and an ED₅₀ of 386 mg/kg of body weight for WPI, in single-dose experiments from 100 to 1,500 mg/kg of body weight of these proteins. The BCH did not present a dose–response relationship when evaluated in a single dose; therefore the ED₅₀ for bovine collagen could not be determined. However, BCH was tested at different concentrations to evaluate BCH in combination with WPI at the best concentration found for WPI, in order to verify whether BCH presented a dose–response relationship when administrated at double doses; for this the highest concentration of 1,500 mg/kg of body weight was tested. Porcine collagen and WPI, administered at the maximum dose of 1,500 mg/kg of body weight, presented a reduction in ULI of 76% and 80%, respectively, compared to saline (negative control), defined as 100%.

Variable quantities of WPI with BCH were combined based on a previous study¹⁹ showing that 200 mg/kg of body weight of BCH was enough to provide an anti-ulcerative efficacy (percentage ULI) against intragastric ethanol administration. Therefore, based on the individual WPI and BCH efficacious concentrations, various combinations were tested, and the reductions (ULIs) for single-dose (acute) experiments are shown in Table 1.

Table 1.

Effect of Various Combinations of WPI and BCH, in a Single Dose, in Reducing the ULI in Rats

Intragastric intubation (dosage)	Number of animals	ULI (mean ± SD)	% reduction of ULI
Saline^a	21	105.7 ± 25.5	—
Carbenoxolone (200)	16	20.0 ± 5.8***	81.0
BCH/WPI
187/187	6	60.8 ± 14.7***	42.5
375/375	6	37.7 ± 14.1***	64.4
200/175	6	82.4 ± 17.5*	22.0
200/550	6	57.7 ± 17.9***	45.4
200/800	6	39.0 ± 7.5***	63.1
200/1,300	6	47.0 ± 10.3***	55.5

Dosage is in mg/kg of body weight.

Saline dosage of 10 mL/kg of body weight.

By analysis of variance, F _(7,64) = 37.41; ***P < .001.

By Tukey's test, compared with saline: *P < .05, ***P < .001.

Various combinations of BCH/WPI promoted reductions of ULI in the range of 22% for the respective dosage of 200/175 mg/kg of body weight (total dosage, 375 mg/kg of body weight) to 64% for the respective dosage of 375/375 mg/kg of body weight (total dosage, 750 mg/kg of body weight). No statistical differences were found (P > .05) among the various combinations tested, except for the respective combination 200/175 mg/kg of body weight, which was statistically inferior (P < .05) compared to all other combinations.

Table 2 shows the effect of various combinations of mixed PCH and WPI on the percentage reduction of ULI in the rat stomach mucosa, prior to administration of 1 mL of absolute ethanol per 250-g rat. The reduction ranged from 31% to 57% with no statistical differences among the various combinations (Tukey's test), except for the PCH/WPI combination of, respectively, 187/187 mg/kg of body weight, which differed from the others by presenting the lowest percentage of reduction with regard to saline (P < .01). The PCH/WPI combination in the respective dosage of 375/ 375 mg/kg of body weight differed from saline (P < .001) and was one of the best among all combinations, considering the relative cost and availability of the collagen hydrolysate and of the WPI, and an efficacy of 56% reduction in the ULI at a total dosage of 750 mg/kg of body weight. This suggests a positive synergism between PCH and WPI at this dosage because other combinations at the same total dosage showed relatively lower percentage reductions in the ULI. Increasing PCH and decreasing WPI showed a tendency to decrease the ULI reduction. The results presented in Tables 1 and 2 were conducted on the same experimental day, and the same control was used.

Table 2.

Effect of Various Combinations of WPI and PCH, in a Single Dose, in Reducing the ULI in Rats

Intragastric intubation (dosage)	Number of animals	ULI (mean ± SD)	% reduction of ULI
Saline^a	21	105.7 ± 25.5	—
Carbenoxolone (200)	16	20.0 ± 5.8***	81.0
PCH/WPI
187/187	6	72.8 ± 37.0**	31.1
375/375	6	46.3 ± 16.6***	56.2
150/600	6	44.8 ± 17.8***	57.6
300/450	11	59.2 ± 21.6***	44.0
450/300	6	55.0 ± 15.0***	47.9
600/150	6	65.2 ± 28.2***	38.3

Ethanol was given at 1 mL per rat. Intubation dosage is in mg/kg of body weight.

Saline dosage of 10 mL/kg of body weight.

By analysis of variance, F _(7,69) = 21.78; ***P < .001.

By Tukey's test, compared with saline: **P < .01, ***P < .001.

The efficacies of BCH and PCH alone or in combinations with WPI in a double-dose experiment are shown in Table 3. Some interesting comparisons are as follows: a double dose of 1,500 mg/kg of body weight BCH showed the same efficacy as 750 mg/kg of body weight WPI and also as the BCH/ WPI combination of, respectively, 375/375 mg/kg of body weight, emphasizing the higher efficacy of WPI in comparison with the BCH and once again suggesting a synergistic effect of the BCH/WPI combination at the respective dosage of 375/375 mg/kg of body weight. On the other hand, the BCH/WPI combination seemed to perform better than the PCH/WPI combination in reducing the ULI caused by ethanol in the rat stomach mucosa. Both WPI and BCH were tested alone at 375 mg/kg of body weight; the ULI reduction percentages were 49.1% and 51.1%, respectively.¹⁹

Table 3.

Effect of Various Treatments (Double Doses) of WPI and BCH or PCH in Reducing the ULI in Rats

Intragastric intubation (dosage)	Number of animals	ULI (mean ± SD)	% reduction of ULI
Saline^a	11	113.9 ± 23.0	—
Carbenoxolone (200)	8	22.4 ± 4.7***	80.4
BCH (200)	6	64.5 ± 20.5**	43.4
BCH (1,500)	5	25.8 ± 10.8***	77.4
PCH (750)	8	70.1 ± 17.3**	38.4
WPI (750)	6	29.2 ± 16.1***	74.3
BCH/WPI (375/375)	6	27.2 ± 13.6***	76.1
PCH/WPI (300/450)	6	65.7 ± 19.4**	42.4

Ethanol was given at 1 mL per rat. Intubation dosage is in mg/kg of body weight.

Saline dosage of 10 mL/kg of body weight.

By analysis of variance, F _(7,42) = 28.00; ***P < .001.

By Tukey's test, compared with saline: **P < .01, ***P < .001.

Table 4 presents the results of gastrin analysis in the plasma of rats intubated with the controls (saline or carbenoxolone), WPI, PCH, BCH, and combinations of WPI with either PCH or BCH. The percentage decrease in plasma gastrin in relation to saline was approximately 40% for carbenoxolone, WPI, and BCH with no statistical differences among them. Carbenoxolone and BCH, both at 200 mg/kg of body weight, were equally efficient in decreasing plasma gastrin by about 39%.

Table 4.

Effect of WPI, BCH, PCH, and Combinations of WPI with Either PCH or BCH on Decreasing Plasma Gastrin in Rats Intubated with Absolute Ethanol

Intragastric intubation (dosage)	Gastrin (pg/mL of plasma)	% decrease in gastrin comparing with saline
Saline^a	146 ± 60.4	—
Carbenoxolone (200)	89.5 ± 19.4**	38.8
WPI (750)	90.2 ± 10.3**	38.3
PCH (750)	84.2 ± 6.8**	42.4
BCH (200)	88.0 ± 11.8**	39.8
BCH/WPI (375/375)	94.3 ± 15.0**	35.5
PCH/WPI (150/600)	77.0 ± 8.5***	47.3

Ethanol was given at 1 mL per rat. Intubation dosage is in mg/kg of body weight. Gastrin production is in mean ± SD values (n = 6 rats per group).

Saline dosage of 10 mL/kg of body weight.

By analysis of variance, F _(7,35) = 5.56; ***P < .001.

By Dunnett's test: **P < .01, ***P < .001.

The combination of PCH/WPI at, respectively, 150/600 mg/kg of body weight was the most efficient treatment (47% reduction in gastrin) and differed from all the other treatments according to Tukey's test (P < .001), compared with saline (negative control).

The gastric mucus-stimulating power of carbenoxolone, WPI, PCH, BCH, and combinations with either PCH or BCH with WPI are presented in Table 5.

Table 5.

Effect of Carbenoxolone, WPI, BCH, PCH, and Combinations of WPI with Either PCH or BCH on Stimulating Gastric Mucosa Production of Mucus in Rats Intubated with Absolute Ethanol

Intragastric intubation (dosage)	Mucus (μg of Alcian blue-linked mucus/g of mucosa tissue)	% increase relative to saline
Saline^a	64.7 ± 10.0	—
Carbenoxolone (200)	192.9 ± 70.8**	198.0
WPI (750)	173.9 ± 55.9**	168.5
PCH (750)	167.3 ± 40.4**	158.5
BCH (200)	102.8 ± 40.8*	58.9
BCH/WPI (375/375)	98.0 ± 23.0*	51.5
PCH/WPI (150/600)	237.7 ± 33.0***	297.4

Ethanol was given at 1 mL per rat. Intubation dosage is in mg/kg of body weight. Mucus production is in mean ± SD values (n = 6 rats per group).

Saline dosage of 10 mL/kg of body weight.

By analysis of variance, F _(6,22) = 8.60; ***P < .001.

By Tukey's test: *P < .05, **P < .01, ***P < .001.

Carbenoxolone at the dosage of 200 mg/kg of body weight increased mucus in the gastric mucosa by 198%, compared with a 58.9% increase for BCH, at the same dosage. On the other hand, WPI increased the mucus concentration by 168.5% compared with an increase of 158.5% for PCH at 750 mg/kg of body weight. Comparing two combinations of WPI with either PCH or BCH, PCH/WPI (respectively, 150/600 mg/kg of body weight) increased mucus in the mucosa of 267%, whereas BCH/WPI (respectively, 375/375 mg/kg of body weight) promoted an increase of only 51.5%, compared to the saline control. The comparisons between the various dosages and agents suggest that carbenoxolone was the strongest mucus-stimulating treatment, as expected. WPI and PCH were equally effective at the dosage of 750 mg/kg of body weight; in contrast, BCH/WPI (respectively, 375/375 mg/kg of body weight) did not differ from 200 mg/kg of body weight BCH alone, suggesting that the 375/375 mg/kg of body weight combination of BCH/WPI exerted a negative synergism regarding the mucus production, whereas PCH/WPI (respectively, 150/600 mg/kg of body weight) strongly stimulated production, suggesting a positive synergism between the two components.

The relative importance of the free sulfhydryl groups, which contain substances that protect the gastric mucosa against absolute ethanol-induced ulcerative lesions, was demonstrated by the in vivo alkylation of SH groups with NEM. Table 6 illustrates the percentage decrease of ULI in the rat stomach mucosa by the various samples studied, in the absence or in the presence of the alkyllating agent.

Table 6.

Effect of Non-Protein Sulfydryl Group Alkylation Against Anti-Ulcerogenic Activity of BCH, PCH, and WPI in Various Amounts and Combinations Against Ulcerative Lesions Caused by Absolute Ethanol

Intragastric intubation (dosage)	ULI	% decrease in ULI ^a
Saline^b	120.5 ± 30.3	—
Saline + NEM^c	144.3 ± 36.1	—
BCH (200)	64.2 ± 23.4**	46.7
+ NEM	135.3 ± 24.9^NS	12.3
BCH (1,500)	63.2 ± 7.2**	47.5
+ NEM	122.7 ± 38.9^NS	14.9
PCH (750)	69.0 ± 14.7**	42.7
+ NEM	100.8 ± 21.6*	30.1
WPI (750)	45.2 ± 9.9***	62.5
+ NEM	125.7 ± 24.9^NS	12.9
WPI/PCH (600/150)	56.2 ± 19.2**	53.4
+ NEM	73.5 ± 17.5***	49.0
WPI/BCH (375/375)	68.3 ± 28.3**	43.3
+ NEM	87.8 ± 22.0**	39.1

Ethanol was given at 1 mL per rat. Intubation dosage is in mg/kg of body weight. ULI is in mean ± SD values (n = 6 rats per group). NS, difference not significant.

Comparisons for compound alone are to saline control; comparisons for compound plus NEM are to saline plus NEM.

Saline dosage of 10 mL/kg of body weight.

Saline intubation (10 mL/kg of body weight) followed by NEM (10 mg/kg of body weight) by subcutaneous injection.

By analysis of variance, F _(13,79) = 10.83; ***P < .001.

By Tukey's test: *P < .05, **P < .01, ***P < .001.

By eliminating the protecting effect of the sulfhydryl groups by NEM alkylation, the percentage decrease in the ULI dropped significantly. Interestingly, the combinations of WPI with collagen hydrolysates were less sensitive to alkylation, suggesting a lower collagen dependence of sulfhydryl compound protection than with WPI treatment alone.

Discussion

Ethanol is an ulcerogenic agent that acts directly on the mucosa as an aggressive lesion-promoting substance and is also capable of stimulating stomach acid secretion, increased gastric mobility, decreased blood flow in the mucosa, decreased prostaglandin production, and increased acethylcholine and intracellular calcium levels. Some of these factors are involved in gastric secretion under various types of stress.²⁴ A previous study⁸ demonstrated that double doses (1,000 mg/kg of body weight) of a WPC, produced in a pilot plant,¹⁰ protected the rat gastric mucosa from absolute ethanol-induced lesions by decreasing the ULI by approximately 75%, compared to a saline control.

More recent work^12,13 has shown that one of the most protective whey proteins against ethanol gastric mucosa lesion is α-lactalbumin. Furthermore, Santos¹⁸ showed that bovine tendon type I collagen caused recovery from mucosa duodenal lesions caused by P. vulgaris lectins. Castro et al.¹⁹ demonstrated that both BCH and PCH inhibited the ethanol ULI in rats, by approximately 50–60%, at a dosage of 750 mg/kg of body weight. BCH did not present a dose–response correlation in the ethanol model, whereas PCH showed a logarithmic dose–response relationship. BCH decreased the ULI by 55%, compared to the 61% decrease observed for PCH at the same dosage (750 mg/kg of body weight).

In the present report, double doses (750 mg/kg of body weight) of a WPI (Table 3) decreased the ULI by 74%, whereas the same dose of PCH decreased the ULI by 38%. However, BCH, at a lower dose of 200 mg/kg of body weight, decreased the ULI by 43%; in contrast, the 1,500 mg/kg of body weight dose decreased the ULI by 77%. Thus, BCH appears to be more effective than PCH in protecting the rat gastric mucosa from ethanol-induced lesions. The reasons for this have not yet been clarified. Therefore the basic physicochemical composition, degree of hydrolysis, amino acid profile, total and free ammonium content, and isoelectric focusing were determined in order to detect differences. For the amino acid profile (in g/100 g of protein) some differences were found, as serine (0.8), tryptophan (0.21), arginine (0.8), and glycine (0.4) were lower in BCH, and cysteine (0.23) and isoleucine (0.34) were higher in BCH, compared to PCH.²⁵

The main difference between these two sources of collagens is their isoelectric range, which is higher for porcine collagen. The isoelectric pH range varied from 4.6 to 7.2 for the bovine samples and from 6.6 to 8.6 for the porcine collagen hydrolyzed samples. Gel permeation chromatography showed a mean molecular mass for BCH of around 3.0 kDa, whereas porcine samples presented a lower mean molecular mass than bovine samples (2.2 kDa).²⁵

The total ammonium in the PCH was much higher (about 6,000 mg/kg) than in the BCH (about 3,000 mg/kg). The degree of amidation (content of amide glutamine and asparagine) is an indirect proof of the alkaline pretreatment during the production process of the collagen hydrolysates. As expected, the bovine hydrolysates showed a significantly lower degree of amidation (15%) than the porcine hydrolysates (30%).²⁵

Interestingly, the combination of BCH/WPI (respectively, 375/375 mg/kg of body weight), at double doses, provided the same protection as 750 mg/kg of body weight WPI and 1,500 mg/kg of body weight BCH, whereas the PCH/WPI combination (respectively, 300/450 mg/kg of body weight), at double doses, resulted in a lower protection (42%, vs. 76% for the respective WPI/BCH combination 375/375 mg/kg of body weight). These data suggest a synergistic action between WPI and BCH that does not appear to exist between WPI and PCH. This finding is even more significant considering the fact that bovine collagen is cheaper and more available than WPI.

In order to better understand how WPI and the collagen hydrolysates protect the gastric mucosa from the ethanol-induced lesions, changes in gastrin and gastric mucus and the relative importance of sulfhydryl substances were considered. The data in Table 4 show a significant decrease in gastrin, ranging from about 35% to 47% compared to physiological saline, as a consequence of the gastric intubation of the various substances studied. Gastrin is a peptide hormone synthesized by endocrine cells located in the antrum region of the stomach and duodenum, entering the bloodstream by the portal system. Its production is stimulated mainly by free amino acids and peptides from foods reaching the stomach. The main action of gastrin is the stimulation of acid secretion by the parietal cells of the stomach. The lowering of the gastrin concentration in the blood plasma suggests a drop in acidity in the stomach content after gastric intubation of WPI and the collagen hydrolysates.²⁶

In addition to the stomach pH, gastrin liberation is also inhibited by a series of peptides of endogenous origin; the most important of thesse is somatostatin.²⁷ Somatostatin is produced by the D cells, localized in the antral mucosa very near to the G cells, which are responsible for the production of gastrin. Pharmacological administration of somatostatin is a potent inhibitor of gastric acid secretion and has a direct action on the parietal cells and an indirect suppression of gastrin liberation. Other substances such as prostaglandins, secretin, glucagon, calcitonin, and an intestinal vasoactive peptide may inhibit the liberation of gastrin.²⁸

Published work on the anti-ulcerogenic properties of α-lactalbumin, using the ethanol induction model,^12,13 showed that a single oral dose of 200 mg of α-lactalbumin/kg of body weight promoted a very high increase in prostaglandin E₂ in the rat stomach mucosa. In a healthy stomach and duodenum there is a balance between the potential of gastric acid and pepsin to damage gastric mucosal cells and the ability of these cells to protect themselves from injury.²⁹ Disruption of the balance results in a breakdown of the normal mucosal defense mechanisms.³⁰ Several mechanisms are believed to be important in protecting the gastric mucosa from damage, including mucus formation, mucosal blood flow, cell renewal, and bicarbonate production.³¹

The extent of ethanol-induced damage in the rat gastric mucosa correlated with degranulating cell content,³² a source of neuropeptides and inflammatory mediators, including histamine and leukotrienes. Ethanol-induced ulcers were not prevented by antisecretory agents, such as cimetidine, but were inhibited by agents that enhanced mucosal defense factors such as prostaglandin E₂.⁵ A recent publication³³ reported that malnourished rats, born from females maintained on a low protein diet (60 g of protein/kg of diet), from the first day of pregnancy until the end of lactation, were less susceptible to ulceration of the gastric mucosa in the ethanol model of gastric ulcer, compared with rats from mothers fed on a normal protein diet (170 g of protein/kg of diet). Mucus production and prostaglandin E₂ formation were increased in malnourished rats, and these animals exhibited a lower number of ulcers in acute experiments than normal rats (P < .05). This difference was not seen in subchronic gastric ulceration (14 days). In the current study, administration of BCH at double doses showed that a higher concentration of protein increased the percentage of ULI, in contrast to results from Paula et al. ³³

The increases in mucus production by single doses of the various protecting agents used in the present study (Table 5) ranged from 51.5% for the BCH/WPI combination (respectively, 375/375 mg/kg of body weight) to 267% for the PCH/WPI combination (150/600 mg/kg of body weight). WPI alone (750 mg/kg of body weight) increased mucus production by 169%, and PCH, at the same dose, presented an increase of 158%. The PCH/WPI combination (respectively, 150/600 mg/kg of body weight) showed a very high efficiency for mucus production stimulation. Mucus stimulation of 74% was also found for acute treatment with purified α-lactalbumin¹³ at a dosage of 200 mg/kg of body weight.

Gastric mucus appears in three forms: soluble mucin present in pancreatic secretion, insoluble mucus (adherent) covering the cells of the mucosa, and the mucus originated from secretory cells located between apical cells.³⁴ The gastric epithelium is covered by a continuous viscoelastic mucus gel layer, which forms a physical barrier between the gastric lumen and the surface of apical cells.³⁵ The structural characteristics of this barrier are primary indicators of its physiological function, and changes in its composition have been identified in gastrointestinal pathologies. The high-molecular-weight mucins are responsible for the viscoelastic properties of the mucus barrier; they are widely expressed in epithelial tissues and are characterized by a variable number of repeated peptide sequences that are rich in serine, threonine, and proline and carry large numbers of O-linked oligosaccharide chains. Secreted and membrane-associated forms of mucins have been found based on their function as extracellular viscous secretions or viscoelastic polymer gels or located as membrane anchoring molecules in the glycocalyx.^35,36

In a study of collagen absorption,¹⁴ around 5–10% of ingested protein was absorbed as peptides of 3–3.5 kDa at 6 hours after ingestion. Therefore, it is possible that collagen may provide peptides enriched with proline to regulate mucin production in the mucosa. The bovine collagen amino acid profile presented 0.5% more proline than PCH (data not shown); it may be speculated that, in the presence of an adequate pool of peptides, the synthesis of mucins and mucus production could occur independently of prostaglandin stimulation.

Sulfhydryl compounds such as glutathione, protein sulfhydryls, and cysteine have been reported to protect the gastric mucosal from injury⁶ as confirmed by some more recent work.^8,9,13 In the present study (Table 6), the percentage decrease in ULI noticed after a single-dose intubation (acute treatment) with WPI, PCH, BCH, or their combination ranged from 43% to 62% in the absence of SH alkylation (NEM) and from 12% to 49% after pretreatment with NEM. The importance of alkylation was more evident for WPI (richer in SH groups) than for collagen or the combination of WPI/ collagen (Table 6).

Sulfhydryl substances are naturally present in the mucosa and are subject to in vivo alkylation, whcih diminishes mucosal protection. Furthermore, WPI is rich in cysteine and also has the property of stimulating glutathione synthesis in various body tissues, including the gastrointestinal mucosa.^37,38 Nevertheless, an acute single-dose experiment is not expected to cause a great stimulation in reduced glutathione synthesis. The WPI/BCH combination, in the respective proportion of 375/375 mg/kg of body weight, presented the best result, decreasing the ULI by 64%.

Some studies in the literature correlated the distribution of glutathione and its precursor amino acids (cysteine, glycine, and glutamate) with the expression of their respective amino acid transporters in a specific animal and/or tissue.³⁹ In spite of a very low content of sulfurous amino acids, BCH and PCH have a high content of glycine, which is a precursor of glutathione, and it could lead to glutathione synthesis contribution.³⁹

Therefore, the hypothesis that collagen hydrolysates may contribute to an increase in reduced glutathione synthesis, mainly when administered in combination with WPI, cannot be discounted.

The collagen hydrolysates, both porcine and bovine, presented a stronger effect on mucus production. WPI effect was more dependent on sulfhydryl compounds, resulting in a more protective effect when the two proteins were administered together.

Footnotes

Acknowledgments

The authors acknowledge the financial support given by Gelita South America, Cotia, SP, and the CPQBA (State University of Campinas) for providing the technical support.

Author Disclosure Statement

No competing financial interests exist.

References

Melo

JRC

, Castro

, Lanna

JCBD

, Guimarães

, Sobrinho

JMDL

. Anatomia do estômago e anatomia do desenvolvimento. Gastroenterologia Clínica, Vol. I, 3^rd ed. Dani

, Castro

. Guanabara Koogan S.A.: Rio de Janeiro, 1993; 449–459.

Kontereck

. Mechanisms of gastroprotection. Scand J Gastroenterol, 1993; 174,Suppl:15–28.

Abdel-Salam

DME

, Czimmer

, Debreceni

, Szalesanyi

, Mòzsik

. Gastric mucosal integrity: gastric mucosal blood flow and microcirculation. An overview. J Physiol, 2001; 95:105–127.

Allen

, Flemström

, Garner

, Navilaasko

. Gastroduodenal mucosal protection. Physiol Rev, 1993; 73:823–857.

Robert

, Nezamis

, Lancaster

, Hanchar

. Cytoprotection by prostaglandins in rats. Prevention of gastric necrosis produced by alcohol, HCl, NaOH, hypotonic and thermal injury. Gastroenterology, 1979; 77:761–767.

Szabo

, Nagy

, Pavebani

. Glutathione, protein sulfhydryls and cysteine proteinases in gastric mucosal injury and protection. Clin Chim Acta, 1992; 206:95–105.

Oates

, Hakkinen

. Studies on the mechanism of ethanol-induced gastric damage in rats. Gastroenterology, 1988; 94:10–21.

Rosaneli

, Bighetti

, Antônio

, Carvalho

, Sgabieri

. Efficacy of a whey protein concentrate on the inhibition of stomach ulcerative lesions caused by ethanol ingestion. J Med Food, 2002; 5:221–228.

Rosaneli

, Bighetti

, Antônio

, Carvalho

, Sgabieri

. Protective effect of bovine whey protein concentrate on the ulcerative lesions caused by subcutaneous administration of indomethacin. J Med Food, 2004; 7:309–314.

10.

Zinsly

, Sgabirei

, Jacobucci

, Pacheco

MTB

, Baldine

VLS

. Produção piloto de proteínas de leite bovino: composição e valor nutritivo. Braz J Food Technol, 2001; 4:1–8.

11.

Matsumoto

, Shimokawa

, Ushida

, Toida

, Hayasalva

. New biological function of bovine alpha-lactalbumin: protective affect against ethanol and stress-induced gastric mucosal injury in rats. Biosci Biotechnol Biochem, 2001; 65:1104–1111.

12.

Ushida

, Shimokawa

, Matsumoto

, Toida

, Hayasalva

. Effects of bovine α-lactalbumin on gastric defense mechanisms in naive rats. Biosci Biotechnol Biochem, 2003; 67:577–583.

13.

Mezzaroba

LFH

, Carvalho

, Ponezi

, Antônio

, Monteiro

, Possenti

, Sgabieri

. Antiulcerative property of bovine α-lactalbumin. Int Dairy J, 2006; 16:1005–1012.

14.

Oesser

, Adam

, Babel

, Seifert

. Oral administration of ¹⁴C labelled gelatin hydrolysate leads to an accumulation of radioactivity in cartilage of mice (C57/BL) J Nutr, 1999; 129:1891–1895.

15.

Oesser

, Seifert

. Stimulation of type II collagen biosynthesis and secretion in bovine chondrocytes cultured with degraded collagen. Cell Tissue Res, 2003; 311:393–399.

16.

Nakajima

, Muroya

, Tanabe

R-i

, Chikuni

. Positive effect of collagen V and VI on triglyceride accumulation during differentiation in culture of bovine intramuscular adipocytes. Differentiation, 2002; 70:84–91.

17.

Das

, Banerjee

. Effects of stress on the antioxidant enzymes and gastric ulceration. Mol Cell Biol, 1993; 125:115–125.

18.

Santos

CDI

. Ação do Colágeno extraído de tendões bovinos (colágeno tipo I) na recuperação de lesões da mucosa duodenal induzidas por lectinas de feijão (Paseolus vulgaris) [Dissertação de Mestrado] Departamento de Ciência da Nutrição. Universidade Estadual de Campinas: Campinas, SP, Brasil, 1992.

19.

Castro

, Sgarbieri

, Carvalho

, Tinti

, Possenti

. Protective effect of collagen derivates on ulcerative lesions caused by oral administration of ethanol. J Med Food, 2007; 10:154–158.

20.

Maubois

, Olivier

. Extraction of milk protein. Food Proteins and Their Applications. Marcel Dekker: New York, 1997; 579–595.

21.

Gamberini

, Skorupa

, Souccar

, Lapa

. Inhibition of gastric secretion by a water extract from Baccharis triptera Mart. Mem Inst Osvaldo Cruz, 1991; 86,Suppl 2:137–139.

22.

Slingerland

, Carderalli

, Burrows

, Miller

. The utility of serum gastrin levels in assessing the significance of low serum vitamin B12 levels. Arch Intern Med, 1984; 144:1167–1168.

23.

Corne

, Morrisey

, Woods

. A method for the quantitative estimation of gastric barrier mucus. J Physiol, 1974; 242:116–117.

24.

Robert

. Effects of drugs on gastric secretion. Physiology of the Gastrointestinal Tract, 2^nd ed. Johnson

. Raven Press: New York, 1987; 1071–1088.

25.

Castro

, Bouhallab

, Mollé

, Sgarbieri

. Fractionation, characterization of collagen hydrolysates. J Chromatogr Sci 2009[Epub ahead print].

26.

Walsh

. As funções fisiológicas e patológicas da gastrina. Conceitos Básicos em Gastroenterologia. SmithKline: São Paulo, 1987; 1–56.

27.

Philip

, Domschke

, Domscke

. Inhibition by somatostatin of gastrin release and gastrin acid responses to meals and to pentagastrin in man. Scand J Gastroenterol, 1977; 12:261–265.

28.

Polack

, Bloom

. Regulatory peptides of the stomach. Gastric Secretion. Basic and Clinical Aspects. Konturek

, Domschke

. Thieme-Stratton: New York, 1981; 1–9.

29.

Bagchi

, Milnes

, Williams

, Balmoori

, Ye

, Stahs

, Bagchi

. Acute and chronic stress-induced oxidative gastrointestinal injury in rats, and protective ability of novel grape seed Proonthorya nidin extract. Nutr Res, 1999; 19:1189–1199.

30.

Forsell

. Gastric mucosal defense mechanisms: a brief review. Scand J Gastroenterol Suppl, 1988; 155:23–28.

31.

D'Souza

, Dhurne

. Gastric cytoprotection. Indian J Physiol Pharmacol, 1991; 35:88–98.

32.

Diel

, Szabo

. Dose-dependent effects of linear and cyclic somatostatin on ethanol-induced gastric erosions: the role of mast cells and increased vascular permeability in the rat. Regul Pept, 1986; 13:2735–2743.

33.

Paula

ACB

, Gracioso

, Toma

, Bezerra

, Saad

MAJ

, De Lucca

IMS

, Carneiro

, Souza Brito

ARM

. Is gastric ulceration different in normal and malnourished rats? Br J Nutr, 2005; 93:47–52.

34.

Mojzis

, Hegedusova

, Mirossay

. Role of mucus in ischemia/reperfusion-induced gastric mucosa injury in rats. Physiol Res, 2000; 49:441–446.

35.

Flemström

, Hällgren

, Nylander

, Engstrand

, Wilander

, Allen

. Adherent surface mucus gel restrict diffusion of macromolecules in rat duodenum in vivo. Am J Physiol Gastrointest Liver Physiol, 1999; 277:G375–G382.

36.

Corfield

, Myerscough

, Longman

, Sylvester

, Arul

, Pignatelli

. Mucins and mucosal protection in the gastrointestinal tract: new propects for mucins in the pathology of the gastrointestinal disease. Gut, 2000; 47:589–594.

37.

Sgarbieri

. Propriedades fisiológicas funcionais das proteínas do soro de leite. Rev Nutr, 2004; 17:397–409.

38.

Marshall

. Therapeutic applications of whey protein. Altern Med Rev, 2004; 9:136–55.

39.

Lim

, Li

, Marc

, Jacobs , Kistler

, Donaldson

. Mapping of glutathione and its precursor amino acids reveals a role for GLYT2 in glycine uptake in the lens core. Invest Ophthalmol Vis Sci, 2007; 48:5142–5151.