Early research considered that the amount of B vitamins synthesised by rumen microbes could meet the requirement of ruminants and that vitamins synthesised by some bacteria species could meet the requirements of other species by cross-feeding in the rumen(Reference Wolin, Miller and Hobson1). However, Van Gylswyk et al. found that the supply of certain B vitamins from rumen microbial synthesis was too low to meet the optimal growth of some species(Reference Van Gylswyk, Wejdemar and Kulander2). Recent studies observed that dietary addition of folic acid or pantothenate increased concentration of rumen total volatile fatty acids (VFA), abundance of cellulolytic bacteria and degradabilities of DM, neutral-detergent fibre (NDF) and crude protein (CP) in steers(Reference Liu, Wang and Li3,Reference Wang, Liu and Guo4) .
Riboflavin (RF) is required by all organisms. In the form of flavin mononucleotide and flavin adenine dinucleotide, RF functions in electron transfer reactions related to energy, carbohydrate, lipid and amino acid metabolism(Reference Pinto and Rivlin5). Flavin coenzyme is required for the activation and metabolism of other B vitamins, such as folate, cyanocobalamin and pyridoxine(Reference Pinto and Rivlin5). Rumen microbial RF synthesis was associated with dietary composition and increased with increasing the level of readily degradable carbohydrates and protein(Reference Beaudet, Gervais and Graulet6,Reference Castagnino, Seck and Longuski7) . Apparent ruminal synthesis of RF calculated as the duodenal flow minus its daily intake was negative in cows receiving diets, such as high fibre and low N, alfalfa or orchardgrass silages as the sole forage or different types of dry maize grain(Reference Beaudet, Gervais and Graulet6–Reference Castagnino, Seck and Beaudet8). Santschi et al. noted 99·3 % RF supplemented in diets disappeared before the duodenal cannula of cows(Reference Santschi, Berthiaume and Matte9). These results suggested that rumen microbes might need to use dietary RF to support their growth and nutrient digestion. Studies in vitro reported that RF was essential for the growth of Ruminococcus flavefaciens (Reference Hungate10) and R. albus (Reference Bryant and Robinson11) and that the digestion of cellulose by rumen micro-organisms was stimulated by RF addition(Reference Hall, Cheng and Burrows12). However, there was no reports about the impacts of dietary RF supplementation on ruminal fermentation and bacterial growth in vivo.
On account of the results above, it was speculated that dietary RF supplementation could promote rumen microbial growth and nutrients degradation. Therefore, the objective of current study was to investigate the influences of RF supplementation on ruminal fermentation, nutrients degradation, microbial abundance and urinary excretion of purine derivatives in Holstein bulls.
Materials and methods
Animals and experimental design
The experiment was approved by the Animal Care and Use Committee of Shanxi Agriculture University and conducted at Shanxi Agriculture University experiment station. Eight Holstein dairy bulls (24 (sd 3·4) months of age and 614 (sd 28·9) kg body weight) equipped with a ruminal cannula were randomly divided into four treatments in a repeated 4 × 4 Latin square design with a 4 × 24 d period. The first 14 d of each period were for adaptation and the followed d 15–24 were for data and sample collection. Bulls in control, low RF, medium RF and high RF groups were individually offered basal diets supplemented with RF 0, 300, 600 and 900 mg/d, respectively. Basal diets were formulated according to the National Research Council(13) to meet nutrient requirement of dairy bulls (Table 1). Content of RF in the basal diets was 4·2 mg/kg DM and was measured according to the method described by Santschi et al.(Reference Santschi, Berthiaume and Matte9). Supplementary RF (feed grade, 980 mg RF/g; Guangzhou Yeshang trade Co. Ltd) was mixed into the premix and then mixed with the first third of morning ration before daily feeding. The low supplementation level of RF was estimated as: (RF net requirement – rumen synthesised RF × intestinal absorptivity)/intestinal absorptivity. Daily estimated net requirement of RF for tissues in dairy cow of 650 kg body weight was 95 mg/d, estimated apparent ruminal synthesis of RF was 15·2 mg/kg of digestible organic matter consumed per d and estimated intestinal absorption percentage of RF was 23–25 %(13). All bulls were individually housed in a stall of 2·5 × 3 m, fed at 07.30 and 19.30 hours daily and had free access to clean drinking water.
Ingredient and chemical composition of the basal diet
* Contained per kg premix: 1600 mg Cu, 8000 mg Mn, 7500 mg Zn, 120 mg iodine, 20 mg Co, 1640 mg vitamin A, 600 mg vitamin D and 200 mg vitamin E.
† Non-fibre carbohydrate, calculated by 1000 – crude protein – neutral-detergent fibre – fat – ash.
Data collection and sampling procedures
All bulls were weighed individually at the start and every 24 d of the trial to determine changes in body weight. During each data and sample collection period, feed offered and refused for each bull were quantified and sampled daily. Total faeces and urine voided by individual bull in 24 h were gathered by using a harness system fitted with a faecal collection bag and urine collection aprons, weighed and sampled. Faecal sample, a representative of 1/12 daily faecal production, was blended with 10 % tartaric acid solution according to 1/4 of sample wet weight. Urine sample, 1·0 % of daily urine excretion, was put into bottles containing 10 % sulphuric acid solution to ensure pH below 3·0. Feed (offered and refused), faeces and urine samples were stored at −20°C. At the end of the trial, all of the feed and faeces samples were dried (55°C and 48 h), mixed by bull and period and then ground to pass through a 1-mm sieve screen for chemical analysis.
Rumen fluid was sampled on days 23 and 24 of each period. Samples of 200 ml were collected from several sites of the rumen (reticulum, dorsal and ventral sac) in the same proportion at 0, 3, 6, 9 and 12 h after the morning feeding. Ten ruminal samples for each bull per period were collected and used for chemical analysis. Rumen fluid pH was measured using a portable pH meter (PHS-3C, Shanghai Meiyingpu Instrument Manufacturing Co. Ltd) and then filtered using four layers of medical gauze. Filtrate of 5 ml was mixed with metaphosphoric acid (1 ml, 250 g/l) or H2SO4 (1 ml, 20 g/l) and frozen at −20°C for subsequent analyses of VFA and ammonia-N, respectively. Filtrate of 50 ml was placed in liquid N2 and kept at −80°C for analyses of microbial abundance and enzyme activity, respectively.
In situ nutrient degradability
Nutrient degradability of maize silage and concentrate was measured in situ according to the Agricultural and Food Research Council(14). Samples of the air-dried maize silage and concentrate were ground to pass through a 2·5-mm screen. On day 15 of each period, 3·0 g of maize silage and 3·5 g of concentrate were incubated separately in nylon bags (8 × 12 cm, pore size 47 μm) suspended in the rumen of each dairy bull. The duplicated bags were inserted at 2 h after feeding, taken out at 0, 4, 8, 12, 24, 36, 48 and 72 h, washed with cold water and then dried at 65°C for 24 h. Kinetic parameters of DM, CP and NDF degradabilities were calculated using the non-linear regression procedure of SAS(15). Nutrient disappearance percentage at each incubation time of individual bull was calculated by using the equation:
$$y\; = \;a\; + \;b\left( {1\; - \;{e^{ - c(t - L)}}} \right)\;\;{\rm{for}}\;t\;{\rm{ > }}\;L\;{\rm{and}}\;y\; = \;a\;{\rm{when}}\;t\;{\rm{ < }}\;L$$
in which y is the fraction degraded in the time t, a is the soluble fraction, b is the slowly degradable fraction and c is the fractional degradation rate constant at which b is degraded, L (h) is the lag time and t (h) is the time of incubation(Reference McDonald16). Effective degradability (ED) was calculated as ED = a + [bc/(c + k)] × e −k × L , where k is the particulate passage rate, which is 0·058 h−1 and 0·025 h−1 for concentrate and maize silage, respectively(Reference McDonald16).
Chemical analyses
Feeds, refusals and faeces samples were analysed for DM (method 920.36) by drying at 135°C for 3 h, CP (method 984.13) by a Kjeldahl method and ash (method 923.03) by combustion at 550°C for 3 h(17). Organic matter was estimated as DM minus ash. NDF was determined according to the method of Van Soest et al.(Reference Van Soest, Robertson and Lewis18), and heat α-amylase and Na2SO3 were used in the procedure. Acid-detergent fibre was measured according to the Association of Official Analytical Chemists (method 973.18)(17). Allantoin and uric acid were determined using DR3900 spectrophotometer (Shanghai Zhenbao Electromechanical Equipment Co. Ltd) according to the method of the International Atomic Energy Agency(19). Rumen VFA was analysed using GC (GC7890; Shandong Jinpu Analytical Instrument Co. Ltd), and ammonia-N was done by a method of the Association of Official Analytical Chemists(17). Microbial enzyme activities in the rumen fluid were measured using the methods of Agarwal et al.(Reference Agarwal, Kamra and Chaudhary20).
Extraction of microbial DNA and real-time PCR
Isolation of microbial DNA was performed by using the repeated bead-beating method(Reference Yu and Morrison21) from 1·5 ml of homogenised ruminal fluid. The integrity and purity of extracted DNA were checked via agarose gel electrophoresis and NanoDrop 2000 Spectrophotometer (Thermo Scientific, NanoDrop Technologies), respectively. The target microbial primers set sequences are shown in Table 2. For absolutely quantifying the copy number of gene, nine sample-derived DNA standards were prepared by using the regular PCR. The PCR products were purified using the PureLinkTM Quick Gel Extraction and PCR Purification Combo Kit (Thermo Fisher Scientific Co. Ltd) and quantified using a spectrophotometer. Copy number of each sample-derived standard was evaluated according to the PCR product length and mass concentration. The target DNA was quantified by using ten-fold serial dilutions from 101 to 108 DNA copies(Reference Kongmun, Wanapat and Pakdee22). The quantitative PCR assay was carried out in a StepOneTM system (Bio-Rad) in triplicate. The reaction mixture (20 μl) contained 10 µl SYBR Premix TaqTM II (TaKaRa), 2 μl DNA template, 0·8 μl of each primer (10 μmol/μl), 0·4 µl ROX Reference Dye II (TaKaRa) and 6·0 μl nuclease-free water. The conditions of quantitative PCR assay were as follows: one cycle at 50°C for 2 min and 95°C for 2 min for initial denaturation, followed by forty cycles at 95°C for 15 s, and then at annealing temperature for 30 s and extension at 60°C for 1 min.
PCR primers for real time-PCR assay
F, forward; R, reverse.
Statistical analyses
Feed conversion ratio for each bull was estimated as DM intake divided by average daily gain. Data were analysed by the mixed model procedure of SAS (Proc Mixed; SAS 2002)(15) with a repeated 4 × 4 Latin square design to account for effects of square, period within square, bull within square and treatment. The treatment was considered as a fixed effect; square, period within square and bull within square were considered as random effects. Data for ruminal pH, VFA, ammonia-N, microbial enzyme activity and microbiota were summarised by sampling time and then analysed using the same mixed model but with time included as a repeated measure using compound symmetry. Rumen samples were analysed using the average per sampling time. Linear and quadratic orthogonal contrasts were tested using the CONTRAST statement of SAS with coefficients estimated based on the level of RF supplementation. The significant effects for the factors were suggested at P < 0·05, and trends were declared at 0·05 < P < 0·10.
Results
Growth performance, nutrient apparent digestibility and ruminal fermentation
Dietary addition of RF did not affect DM intake of bulls (Table 3). No significant differences were observed for body weight among treatments at the beginning or at the end of the trial. Average daily gain tended to be linearly increased (P = 0·082), and feed conversion ratio was linearly decreased (P = 0·039) with increasing addition level of RF. Increasing supplementation level of RF, total tract digestibilities of DM, organic matter, CP and NDF increased linearly (P < 0·05) but digestibility of acid-detergent fibre was unchanged. Increasing the level of RF addition, rumen pH decreased quadratically (P = 0·041), total VFA concentration increased quadratically (P = 0·039) but molar proportions of butyrate, isobutyrate, isovalerate and valerate were unaffected. Acetate percentage and acetate:propionate ratio increased linearly (P < 0·05), while propionate percentage and ammonia-N concentration decreased linearly (P < 0·05).
Effects of riboflavin (RF) supplementation on DM intake (DMI), average daily gain (ADG), feed conversion ratio (FCR), nutrient digestibility and rumen fermentation in Holstein bulls
(Mean values with their standard errors)
VFA, volatile fatty acids.
* Control, low RF (LRF), medium RF (MRF) and high RF (HRF) with addition of RF 0, 300, 600 and 900 mg/d, respectively. Values are presented as means (n 32).
† The P values of time for ruminal pH, total VFA and ammonia-N were 0·034, 0·008 and 0·046, respectively. The P values of time for VFA molar percentage and acetate:propionate, and the treatment × time interactions for the rumen variables were not significant (P > 0·05).
Rumen nutrient degradation
For concentrate DM, the soluble fraction (a) and slowly degradable fraction (b) were unchanged but the degradation rate (c) and ED increased linearly (P < 0·05) with increasing RF addition (Table 4). For concentrate CP, the soluble fraction (a) and slowly degradable fraction (b) decreased linearly (P = 0·001), degradation rate (c) increased linearly (P = 0·001) but ED was unchanged with increasing RF addition.
Effects of riboflavin (RF) supplementation on in situ ruminal digestion kinetics and effective degradability (ED) of concentrate and maize silage in Holstein bulls
(Mean values with their standard errors)
* Control, low RF (LRF), medium RF (MRF) and high RF (HRF) with addition of RF 0, 300, 600 and 900 mg/d, respectively. Values are presented as means (n 32).
† Parameters were calculated from the fitted equation y = a + b(1 − e −c(t–L)) for t > L, where y = percentage of DM disappearance from the nylon bag at time t, a = soluble fraction, b = slowly degradable fraction, c = fraction rate constant at which b is degraded, L = lag time (h) and t = time of incubation (h). ED was calculated using equation a + bc/(c + k), where k = 0·058 h−1 for concentrate and k = 0·025 h−1 for maize silage.
For maize silage DM, the soluble fraction (a) and ED increased linearly (P < 0·05) but the slowly degradable fraction (b) and degradation rate (c) were not affected by the increased level of RF. For maize silage NDF, the soluble fraction (a) and degradation rate (c) increased linearly (P < 0·05), ED increased quadratically (P = 0·010) but the slowly degradable fraction (b) was unchanged.
Rumen microbial enzyme and microbiota
Increasing dietary RF addition linearly increased (P < 0·05) rumen cellulase activity but did not impact activities of α-amylase and protease (Table 5). Populations of total bacteria, fungi, protozoa, dominant cellulolytic bacteria (R. albus, R. flavefaciens and Fibrobacter succinogenes), Prevotella ruminicola and Ruminobacter amylophilus linearly increased (P < 0·05), but Butyrivibrio fibrisolvens was unaltered with increasing RF supplementation.
Effects of riboflavin (RF) supplementation on rumen microbial enzyme activity and microbiota in Holstein bulls
(Mean values with their standard errors)
* Control, low RF (LRF), medium RF (MRF) and high RF (HRF) with addition of RF 0, 300, 600 and 900 mg/d, respectively. Values are presented as means (n 32).
† The P values of time for microbial enzyme activity and microbiota were significant (P < 0·05). The treatment × time interactions for microbial enzyme activity and microbiota were not significant (P > 0·05).
‡ Units of enzyme activity are: carboxymethyl cellulase (μmol glucose/min per ml), cellobiase (μmol glucose/min per ml), xylanase (μmol xylose/min per ml), pectinase (μmol D-galactouronic acid/min per ml), α-amylase (μmol glucose/min per ml) and protease (μg hydrolysed protein/min per ml).
Urinary excretion of purine derivatives
Urinary excretion of allantoin and total purine derivatives linearly increased (P < 0·05) with increasing dietary provision of RF (Table 6). Uric acid excretion was not affected by RF addition.
Effects of riboflavin (RF) supplementation on urinary excretion of purine derivatives (PD) in Holstein bulls
(Mean values with their standard errors)
* Control, low RF (LRF), medium RF (MRF) and high RF (HRF) with addition of RF 0, 300, 600 and 900 mg/d, respectively. Values are presented as means (n 32).
Discussion
DM intake was unaltered, and so the tendency towards increase in average daily gain should be attributed to the stimulation by supplemented RF to nutrients digestibility in the rumen. Studies in cows indicated that the reason of the increase in performance with B vitamins addition, such as folic acid, vitamin B12 and pantothenic acid, was the improvement in metabolic efficiency(Reference Graulet, Matte and Desrochers23,Reference Wang, Liu and Li24) . However, dietary RF would be used or degraded by rumen micro-organisms, and only approximately 1 % of the supplemented RF could reach the small intestine and be absorbed(Reference Santschi, Berthiaume and Matte9). Therefore, the amount of RF available for absorption was probably not increased by dietary supplementation(Reference Santschi, Berthiaume and Matte9). Likewise, Majee et al. noted DM intake was not affected but milk yield tended to be increased when supplementing a B-vitamin blend including RF in dairy cow diets(Reference Majee, Schwab and Bertics25).
The increase in total tract digestibilities of DM, organic matter and NDF was in accordance with the elevation in ruminal degradabilities of DM and NDF, reflecting a stimulatory effect of RF supplementation on rumen bacterial growth and enzymatic activity. The decrease of rumen pH was related to the positive effect of RF supplementation on total VFA concentration. When rumen pH was below 6·0, the growth of predominant cellulolytic bacteria and degradation of structural carbohydrate would be inhibited(Reference Russell and Wilson26). The lower rumen pH was observed in bulls consuming 600 mg RF/d and was 6·22. The increment in total VFA concentration and acetate percentage was in accordance with the elevation in the soluble fraction, degradation rate and ED of maize silage NDF, indicating that rumen VFA and acetate production as well as the rate and extent of NDF degradation were elevated by RF addition. Noziere et al. reported that rumen VFA percentage and concentration could be used to reflect the production of VFA and that rumen acetate level was positively correlated with NDF digestibility(Reference Nozière, Glasser and Sauvant27). Rumen propionate concentration was calculated as total VFA concentration multiplies propionate molar percentage and was not affected (25·2, 24·1, 23·2 and 23·6 mm for bulls in control, low RF, medium RF and high RF groups, respectively) by RF addition. This was in accordance with the unaffected activity of α-amylase with RF supplementation. That butyrate percentage that was not affected by RF supplementation was in agreement with the unaffected population of B. fibrisolvens which degrades cellulose and hemicellulose to butyrate and is the main producer of rumen butyrate(Reference Russell and Rychlik28). Valerate, isobutyrate and isovalerate are end-products of ruminal true protein degradation(Reference Dijkstra29). The absence of changes of valerate, isobutyrate and isovalerate percentages was consistent with the unmodified concentrate CP degradability and was likely associated with the unaltered activity of protease with RF supplementation.
The responses of carboxymethyl cellulase, cellobiase, xylanase and pectinase activities were in accordance with the changes in populations for cellulolytic bacteria (R. albus, R. flavefaciens and F. succinogenes), fungi and protozoa and should be the reason of the increase in acetate molar percentage and ruminal NDF degradability with RF supplementation. The increase in fibrolytic microbes would cause the level of rumen acetate to increase(Reference Nozière, Ortigues-Marty and Loncke30). Dietary fibre was degraded to acetate by bacteria, fungi and protozoa through the secretion of fibrolytic enzymes(Reference Wang and McAllister31). Fungi can degrade plant lignocellulosic tissues which are resistant to bacteria or protozoa, and approximately 10 % of the VFA production and 30 % of fibre digestion could be attributed to protozoa in the rumen(Reference Orpin32). In addition, RF supplementation increased the populations of P. ruminicola and Rb. amylophilus, and this also contributed to an increase in NDF digestibility. Fondevila & Dehority observed that digestion of cellulose increased when fibrolytic bacteria were co-cultured with amylolytic bacteria compared with fibrolytic species alone(Reference Fondevila and Dehority33). The responses of microbial population and cellulase activity observed suggested that dietary supplementation with RF was essential for rumen microbes. RF, in the form of flavin adenine dinucleotide and flavin mononucleotide, functions in an array of cellular electron transport processes including the metabolisms of carbohydrate, lipid and amino acids, cell signalling as well as protein folding, thereby playing a crucial role in cellular proliferation and growth(Reference Pinto and Rivlin5,Reference Yates, Evans and Powers34) . Moreover, studies in vitro observed that RF addition promoted cellulose digestion and the growth of R. albus and R. fiavefaciens (Reference Hungate10–Reference Hall, Cheng and Burrows12) and that addition of some B vitamins including RF stimulated the growth of protozoa(Reference Bonhomme35).
Activities of α-amylase and protease were unaltered with RF addition and were in agreement with the unchanged B. fibrisolvens population. Rumen B. fibrisolvens is considered as one of the most important bacteria in the hydrolysis of starch and protein(Reference Cotta36,Reference Wallace and Brammall37) . The results suggested that RF might not be an essential nutrient for B. fibrisolvens. Likewise, Gill & King observed that RF addition did not affect the growth of B. fibrisolvens in vitro (Reference Gill and King38). Since B. fibrisolvens was ingested by protozoa more rapidly than other bacteria in the rumen(Reference Bonhomme35), the change of B. fibrisolvens population should also be affected by the increase of protozoa population with RF supplementation. Therefore, the effect of RF addition on B. fibrisolvens needs to be verified by further study.
Rumen ammonia-N comes from feed protein degradation and the digestion of bacteria by protozoa(Reference Firkins, Yu and Morrison39). In contrast to the unchanged CP degradability and increased protozoa population, ammonia-N concentration decreased with RF supplementation. The results indicated that more ammonia-N might be incorporated into microbial protein, as reflected by the higher urinary total purine derivatives excretion. The increment in rumen total VFA concentration suggested that more available energy and carbon source were provided to microbes to synthesise protein. Moreover, rumen ammonia-N content was sufficient to support the maximum microbial protein synthesis. The lower ammonia-N (8·33 mg/100 ml) was observed in bulls added medium RF and was above 5·0 mg/100 ml, a level considered to be suitable for microbial protein production(Reference Firkins, Yu and Morrison39). The increase in total purine derivatives excretion represented an increase in the supply of microbial N to the intestine(Reference Verbic, Chen and MacLeod40) and should be the cause for the increment of total tract CP digestibility with RF supplementation.
Conclusion
Dietary supplementation with RF had no influence on DM intake but tended to increase average daily gain in bulls. Supplementation with RF had positive impacts on nutrient digestion, rumen total VFA production, microbial growth and enzymatic activity in bulls. Dietary supplemented RF was mainly required for rumen fibrolytic microbial growth and feed fibre digestion and had the potential to stimulate microbial protein production. No further increase in performance and rumen total VFA concentration was observed when increasing addition level of RF from 600 to 900 mg/d in bulls.
Acknowledgements
The authors thank the staff of Shanxi Agriculture University dairy calves unit for care of the animals. All authors read and approved the manuscript.
This work was supported by a grant from Key Research and Development project of Shanxi Province (201903D211012 and 201903D221001) and Key Research and Development project of Jinzhong City of Shanxi Province (Y192009).
C. W. and Q. L. designed the experiment. H.-M. W., J. Z., C. L. and G. G. conducted the experiment. W.-J. H., Y.-L. Z., S.-L. Z. and C.-X. P. collected and analysed data. H.-M. W., C. W. and Q. L. wrote and revised the manuscript.
The authors declare that no conflicts of interest exist.
