In vitro evaluation of protein digestibility in the abomasum and small intestine of ruminants

The investigation included studies with microbial protein, 0,5 % formaldehyde (HCHO) treated protein in soybean meal, and ruminally undegradable feed proteins in cotton seed and sunflower seed cakes. Microbial mass was separeted with centrifugations from the rumen contents of roughage-fed sheep. Undegradable feed proteins were prepared with the nylon bag technique. The proteins were digested in solutions of pepsin-HCI and trypsin-chymotrypsin. The average digestibility for microbial protein was 80.7 %. The corresponding values for the undegradable proteins in cottonseed cake and sunflower seed cake were 83.7 % and 82.7 %. In the incubation with pepsin-HCI, pH in the solution affected significantly (P < 0.001) the digestibility of HCHO-treated protein in soybean meal. The average digestibilities at pH 1.5—2.0, pH 2.5—3.0, and pH 3.1—4.0 were 97.2, 76.0, and 71.0 %, respectively.


Introduction
In vitro tests of protein digestibility have mainly been used for non-ruminants.In the studies for ruminants Kowalczyk et al.   (1977, 1978 a, b) used abomasal fluid from a fistulated animal and they also evaluated protein digestibilities with rats.Multi-enzyme tests were used when the digestibilities of ru- minally undegradable feed proteins (in sac- co) were analyzed.In nylon bag studies in vitro methods would most obviously be rele- vant, because the digestibility of undegradable protein has to be estimated from a very small sample.The digestibility of microbial protein separated from the rumen contents could also be determined (e.g.Bergen et al.   1967).
The aim of the present study was to inves- tigate possibilities to use in digestibility anal- yses a method which included protein treatments with pepsin-HCI and trypsin-chymotrypsin.This method was chosen, because it was reported e.g. by Mozersky and Panet-  tieri (1983) that the methods based on a pH drop in the incubations (e.g.Hsu et al.  1977), would not be applicable to proteins of which the in vivo digestibility is not known.The present study included investigations mainly with the protein in microbial mass, which was separated from the rumen con- tents of sheep.The digestibility of ruminally undegradable protein in sunflower seed cake, cottonseed cake, and formaldehyde treated protein (0.5 °7o treatment level) in soybean meal was also studied.

In vitro technique
The technique was modified using pro- cedures described in the literature, mainly those by Akeson and Stahman (1964), and  Anon (1970).The procedure was divided into two phases, incubations with pepsin-HCI and with trypsine and chymotrypsine.In the first phase samples of 500 mg were incubated at + 39°C in pepsin-HCI solution for 16 hours.Pepsin (Merck 7190, 1 : 10000, 2000 FIP-U/g) concentration in the solution of 0.05 N HCI was 0.1 °/o.The ratio of the solution to the sample protein was 0.7 ml/mg protein.The amount of pepsin-HCI was hence changed according to the amount of protein in the sample, because Van Bruc- hem and Van't Klooster (1980) suggested that proteins stimulated abomasal secretions of acid in vitro.Incubations were made in glass tubes (volume 100 ml) with rounded bottoms and the tubes were carefully shaken at the beginning of the incubations.The pH in the incubations was always kept under pH 2.0.
In the second phase contents of the tubes were centrifuged at the speed of 400 g for 10 min.Supernatant in the tube was carefully discarded and the sediment was neutralized (pH 7.0) using washings with phosphate buf- fer (pH 7.0).The contents of the tubes were centrifuged as described above and the sedi- ment was taken for further analyses.
The solutions of trypsine and chymotrypsine were added into the tubes so that the amounts were 10 ml and 1 ml, respectively.In addition, 9 ml of the buffer was transfer- red into the tubes.
The incubations in the second phase were made (at pH 7) at + 39°C for 24 hours.After this the samples were centrifuged as de- scribed earlier.The supernatants in the tubes were carefully discarded and the sediments (undigestible protein of the sample) were transferred for the analysis of protein N.

Analysis of protein N in the samples
The sample was transferred into the Kjeldahl bottle and incubated in 80 ml of dis- tilled water, at + 70°C for 1.5 hours.The contents of the bottles were carefully stirred 4-5 times during the incubation.
After this incubation 1.5 ml of A1 2 (S0 4 ) 3 K 2 S0 4 -12H,0, 15 ml of the solution of copper sulphate (156.4 g CuS0 4 -5H 2 0 in 1000 ml of distilled water), and 15 ml of 2.5 % NaOH were added into the bottles.When the contents of the bottles were cool and clear they were squeezed through an N-free filter paper, and the sample on the filter paper was washed with 750 ml of warm, distilled water.The sample and the paper were carried to N determination which was made using the Kjeldahl method but exclud- ing CuS0 4 as a catalyst.Blank tests includ- ing filter paper and reagents were used in all determinations.The method described is based on the procedure used at the Department of Animal Husbandry, University of Helsinki.
Separation of microbial mass from the rumen contents Microbial mass was separated from the ru- men contents of fistulated sheep in the hay or hay and grass silage (65 : 35 on DM basis) diets.Rumen contents were collected from different parts of the rumen, squeezed through a cheese cloth, and the filtrate was centrifuged at 400 g for 5 min.The sediment was discarded and the filtrate was further centrifuged over 30 000 g at + 2°C for 20 min.The sediment was carefully separated and stored frozen for the analyses.

Samples of feed proteins
The samples of the formaldehyde treated protein in soybean meal, and ruminally un- degradable (in sacco) proteins in sunflower seed cake, and cotton seed cake were prepared by Dr. Torben Hvelplund, National   Institute of Animal Science, Denmark (see Hvelplund 1983).

Studies with microbial protein
Composition of the microbial mass Fresh, separated microbial mass was used in incubations.Samples were not dried be- fore incubation, because Wallace (1983)   among others suggested that for instance freeze drying might affect digestibility.Crude protein (Nx6.25)content in the ran- dom-sampled (n = 7) mass was 37.1 ± 1.0 % in fresh weight.The proportion of protein N in total N was 86.7 ± 0.7 % (n = 7).

Effect of HCI and the enzymes on digestibility
According to our results HCI alone gave a poor digestibility value for microbial pro- tein (Table 1).Moreover, the combination of HCI and pepsin was not sufficient eith- er, and it could be suggested that one of the enzymes in the pancreatic fluid is needed for a proper protein digestion.However, pepsin-HCI might have more importance in the digestion of the protein of plant origin when the sample is less digestible owing to the fibrous structure etc.On the other hand, Rojahn and Wagner (1961) ob- served the difference of 9 %-units only in the digestibility of barley protein when the estimation was made either with pepsin-HCI (lower values) or with pepsin-HCI and pancreatin.
Those incubations in which a stronger HCI was used, tended to give higher digest- ibilities for microbial protein.Korte (1979)   wondered whether results of this kind could be caused by a better digestibility of nucleic acids in microbial cells.However, our results emphasized the importance of the enzymes of the pancreatin in the di- gestion of microbial protein.

Digestibility of microbial protein
When the digestibility of microbial pro- tein was determined with the pepsin-HCI and trypsin-chymotrypsin, the mean of the digestibility was 80.7 % (see    Storm & orskov (1982), Hvelp- lund (1983), Storm et al. (1983), and Wal- lace (1983).In some of the incubations di- gestibility of microbial protein differed quite clearly from the mean value.This type of va- riation was also found by Buchmann (1979)  between in vitro incubations.In the present study the coefficient of the variation (CV-%) of the results between the incubations was 3.7 %, being higher than reported by Buch-  mann (1979).
In the experiments reviewed from the liter- ature there was also a great variation in the results between different investigations.This can be partly explained by the different ex- perimental techniques.However, Bergen et  al. (1967) and Wallace (1983) showed great differences in protein digestibility between bacterial strains and therefore part of the differences between different studies and incubations (present study) might be explained by the different composition of the bacterial mass digested.Ecpecially the increase in the proportions of Gram + (Wallace 1983) or cellulolytic bacteria (Bergen et al. 1967) in microbial mass can decrease digestibility of microbial protein.In our technique, occa- sional difficulties in squeezing the small sample in different phases of the work con- tributed also most evidently to the variation of the results.

Studies with microbial protein and feed protein
Effect of pH and incubation period During the incubations the pH in the con- tents of the tubes was carefully followed.
It is doubtful, however, whether pH in the contents of the abomasum and small intestine remains so stable in vivo.In the experi- value, pH 2.9, for steers and the pH varied form 2.50 to 3.17.Moreover, he also sug- gested that pH in the small intestine can vary being 5.69 at the beginning of the intestine and approaching pH 7.3 at the terminal end of the small intestine.
In order to study the pH effect, the pH in pepsin-HCI incubations was adjusted to three different levels (Tables 2 and 3).How- ever, any clear changes in the digestibility of undegraded feed proteins and microbial protein were not observed when the pH varied from 1.5 to 4.0.Ambrose and Snyder (1964) also found only a limited effect on protein digestibility caused by a change in the pH from 1.2 to 2.5.The digestibility of the for- maldehyde treated feed protein was significantly (P < 0.001) decreased when a pH higher than 1.5-2.0 was used.These results demonstrated the importance of acidic con- ditions for the release of formaldehyde-protein complex as suggested for instance by Ferguson et al. (1967) although on the basis of their results as low a pH as in our study would not be necessary.
The effect of the pH on the incubations with trypsin-chymotrypsin was not studied.Johnson et al. (1983) reported that a pH higher than 6.9 could change the activity of some enzymes of the pancreatin.At least this was the case with a-amylase.However, ac- cording to the paper of McAllan (1981) the effect of pH in vivo may not be significant in controlling the activity of trypsin and chymotrypsin.
When protein digestibilities in vitro are applied to in vivo a critical point might be the length of the incubation period, e.g. the retention time of the feed particles in the abomasum and in the small intestine of the ruminant.In pepsin-HCI incubations length of the incubation may not be important if the activity of pepsin is reasonably high (e.g.AO AC-standard).In that case Ambrose and   Snyder (1964) did not report clear differ- ences in digestibilities evaluated after the periods of 2,4, 6, or 16 hours.
In the literature there is not very much information available about the retention time of feed particles in the abomasum and small intestine.In one of the publications Colucci et al. (1982) reported retention times from 11 to 17 hours for concentrate and roughage in the omasum, abomasum, and small intestine.According to these findings an incubation period of 24 hours in our study was too long.
When the incubation period was shortened to 18 hours, which is still reasonable regard- ing the practical work, there were very small and non-significant changes in the digestibility values of microbial and feed proteins (Table 2).
The digestibilities of undegradable proteins in cotton seed cake and sunflower seed cake were, however, generally higher than reported in vivo by Hvelplund (1983).It is possible that the in vitro conditions created optimum circumstances for the activity of enzymes and therefore higher digestibilities were obtained.
In conclusion, the in vitro technique is a reasonable method in evaluation of protein digestibility in the lower digestive tract of a ruminant when it is used in studies with microbial protein or nylon bags.However, the analysis has to be carefully made if HCHO- treated feeds are studied.
It is clear that the described technique is relatively complicated and laborous, but it is more simple than the in vivo techniques generally available.However, direct compar- isons between in vitro and in vivo determina- tions have to be made, before in vitro results can be applied to in vivo conditions.

Table 1 .
The effect of different factors on the digestibility-% of microbial protein in vitro.

Table 3 .
Statistical parameters of the data in table 2 (analysis of variance, factorial experiment).