Use of antimicrobial peptides as a feed additive for juvenile goats

Although antimicrobial peptides (AMPs) have been used as feed additives, only a few studies have examined their use in ruminants. In this study, we evaluated the use of AMPs(recombinant swine defensin and a fly antibacterial peptide were mixed by 1:1) as a medicated feed additive for juvenile goats. Dietary treatments included control groups (group I: 300 g concentrate; group III: 600 g concentrate), and AMP-supplemented groups (group II: 300 g concentrate + 3.0 g AMPs; group IV: 600 g concentrate + 3.0 g AMPs). AMP-treated groups exhibited an increase in bacterial genera, including Fibrobacter, Anaerovibrio, and Succiniclasticum, and the ciliate genus Ophryoscolex; as well a reduction in bacterial genera, such as Selenomonas, Succinivibrio, and Treponema, and the ciliate genera Polyplastron, Entodinium, and Isotricha. The changes in Fibrobacter, Anaerovibrio, Ophryoscolex, Polyplastron, Entodinium, and Isotricha were related to the concentrate. AMP treatment led to increased body weight, average daily weight gain, enzymatic activity (pectinase, xylanase, and lipase), especially in the normal concentrate group, and influence on ruminal fermentation function. In addition, goats treated with AMPs had higher rumen microorganism diversity indices than the control groups. Our results demonstrate that AMPs can be utilized as feed additives for juvenile goats.


Results
Growth performance. The mean initial body weights in groups I, II, III, and IV were 15.54 kg, 15.51 kg, 16.31 kg, and 16.70 kg, respectively. The weights increased to 18.96 kg, 19.93 kg, 21.60 kg, and 22.99 kg, respectively, following 60 days of experimental feeding (Table 1 and Fig. 1). The average daily gain (g) was significantly higher (P < 0.05) in the AMP-supplemented groups (II, IV) than in the control groups (I, III; Table 2).
Ruminal fermentation function. The mean ruminal pH of samples from AMP-treated goats ranged from 6.74 to 6.92, which is within the normal physiological range. No significant difference in ruminal pH was observed between AMP-treated groups and control group (P > 0.05; Table 2).
Total volatile fatty acid (T-VFA) and acetate concentrations increased in goats fed AMPs with normal concentrate (significant difference on day 20), but decreased in goats fed double concentrate compared to the control groups (I, III; P < 0.05). The concentrations of ammonia (significant difference with normal concentrate) and urea nitrogen decreased in AMP-treated groups. The acetate + butyrate-to-propionate ratio decreased in AMP-treated groups; however, significant differences were only observed with double concentrate (P < 0.05). The concentrations of microbial protein (MCP, significant difference on day 60 with double concentrate) and propionate (significant difference on day 20 with normal concentrate) increased in AMP-treated groups.
In addition, all indicators (except ruminal pH and urea nitrogen) were elevated in AMP-treated groups compared with groups I and III; the concentrations of T-VFA, acetate, ammonia, and MCP were significantly increased (P < 0.05). Similarly, the concentrations of propionate, butyrate, and the acetate + butyrate-to-propionate ratio were significantly increased on day 20 (P < 0.05) with double concentrate. Enzyme activity. Pectinase activity appeared to increase in the AMP-supplemented groups (Table 3), and was higher in AMP-supplemented goats than in the control groups (I, III; P < 0.05, except on day 60 with double concentrate). Changes in xylanase, lipase, and amylase activity were associated with concentrate. Xylanase increased with normal concentrate (P < 0.05) and decreased with double concentrate; lipase increased with normal concentrate (P < 0.05) but did not change with double concentrate; and amylase decreased with normal concentrate (P < 0.05) but did not change with double concentrate. No differences in β-glucosidase, carboxymethyl cellulase (CMCase), and protease activity could be detected between AMP-treated and control animals (P > 0.05).
In addition, β-glucosidase and CMCase (except on day 20) activities appeared to be significantly lower in group III compared to group I (P < 0.05); whereas pectinase (except on day 60) and lipase (except on day 20) activities appeared to be significantly higher in group III compared to group I (P < 0.05). No differences in xylanase, amylase, and protease activity could be detected between groups I and III (P > 0.05).
Rumen microorganisms. Bacterial community structure. Following the removal of low-quality reads from sequencing data, we obtained 1,786,781 total reads for bacteria, with an average of 49,632 reads per sample. The identified bacterial phyla and genera are detailed in Tables 4 and 5 and their respective community compositions   Item   Time  point(day)/  Time range   Groups  P-Value   I  II  III  IV  I VS II   III VS  IV  I VS III   II I  II  III  IV  I VS II   III VS  IV  I VS III   II Table 3. Changes of the activity of enzymes in rumen fluid of goats. a Mean significant difference (P < 0.05).
are detailed in Supplementary Fig. S1A and B. Bacteroidetes was the dominant bacterial phylum in all goat rumen samples (expect in group III), accounting on average for 40.85% of the bacterial community. The next seven most abundant phyla were Firmicutes, Proteobacteria, Verrucomicrobia, Fibrobacteres, Tenericutes, Spirochaetes, and Cyanobacteria. At the phylum level, Proteobacteria appeared to significantly decrease (P < 0.05; Table 4) and Fibrobacteres appeared to significantly increase (P < 0.05; Table 4) in the AMP-supplemented groups compared with the control groups (I and III). In addition, Bacteroidetes and Proteobacteria appeared to significantly decrease (P < 0.05) and Firmicutes appeared to significantly increase (P < 0.05) in group III compared with group I.
Fibrobacter and Anaerovibrio appeared to increase in the AMP-supplemented groups although a significant increase was only apparent with normal concentrate. Succiniclasticum appeared to increase (Table 5), whereas Succinivibrio, Selenomonas, and Treponema appeared to decrease in the AMP-treated groups (Table 5) compared with the control groups (I and III). In addition, Prevotella, Anaerovibrio (except on day 20), and Treponema appeared to significantly increase (P < 0.05); whereas Succinivibrio, Selenomonas (except on day 20), and Fibrobacter (except on day 60) appeared to significantly decrease (P < 0.05) in group III compared with group I. No differences in [Paraprevotellaceae]CF231, Butyrivibrio, and Succiniclasticum were observed between groups I and III (P > 0.05).
The Chao1, ACE, Simpson, and Shannon diversity index values of each sample (at the bacterial and ciliate genus level) are shown in Tables 6 and 7, all indices were elevated in the AMP-supplemented groups, especially on day 60. Moreover, all indices were reduced in group III, although these decreases were not statistically significant. These results indicate that AMP supplementation may enhance microbial diversity in the rumen whereas increasing concentrate may reduce it.
Ciliate community structure. A total of 631,179 quality protozoa sequences were obtained from the 36 samples, with an average of 17,532 reads per rumen sample. Although all animal groups were fed the same diet, there was a high level of variation between individuals in terms of ciliate community composition at the genus level and their respective community compositions are detailed in Fig. S2. The only characteristic in common was the dominant role of Polyplastron and Ophryoscolex (Table 8).
Compared with the control groups (I and III), Ophryoscolex appeared to increase in the AMP-supplemented groups (Table 8), although a significant increase was only apparent with normal concentrate. Polyplastron, Entodinium, and Isotricha appeared to decrease in the AMP-supplemented groups, although a significant  I  II  III  IV  I VS II   III VS  IV  I VS III   II    decrease was only observed with normal concentrate. No differences in Diploplastron and Dasytricha were detected between AMP-treated goats and control animals (P > 0.05). Moreover, no differences in Polyplastron, Ophryoscolex, and Isotricha were evident between groups I and III (P > 0.05).

Discussion
Microbial community composition in ruminants has previously been linked with animal production traits 21,22 .
In the present study, we found that Bacteroidetes was the dominant phylum in all samples (except group III), followed by Firmicutes, Proteobacteria, and Verrucomicrobia. This structure is similar to the rumen bacterial community of sheep inferred from multiplex 454 Titanium pyrosequencing 23 . At the genus level, Prevotella, known as an abundant member of the rumen microbiome 24 Table 8. Influence of diet and AMPs on proportion of ciliates genera. a mean significant difference (P < 0.05).
include organisms that are important cellulose and hemicellulose-degraders; this indicates a rumen bacterial community highly oriented towards fibre degradation. Polyplastron and Ophryoscolex were the most abundant ciliate genera in this study; the protozoal community composition is similar to that of the A type (dominated by Polyplastron, Ostracodinium, Dasytricha, and Entodinium) 27 . However, many studies have identified Entodinium as the predominant protozoal group in ruminants [28][29][30][31] . This discrepancy may be due to diet. In this study, forage grass was the main fodder and xylanase and glucanase activities of Polyplastron and Ophryoscolex are much higher than those of Entodinium 27 . In addition, high-throughput sequencing technology could also affect the true composition of rumen ciliates. Kittelmann et al. 32 reported that smaller-celled genera, such as Entodinium, Charonina, and Diplodinium, tended to be underrepresented, while larger-celled genera, such as Metadinium, Epidinium, Eudiplodinium, Ostracodinium, and Polyplastron, tended to be overrepresented using the pyrosequencing approach. Antimicrobial peptides possess broad-spectrum antimicrobial activity and have been used as a new type of feed additive in animal husbandry. A number of recent studies have suggested that dietary supplementation containing an antimicrobial peptide, such as lactoferricin and the lactoferrampin fusion peptide, potato protein, cecropin AD, or antimicrobial peptide P5, reduced the total numbers of aerobes while simultaneously enhancing the total amount of anaerobes and beneficial lactobacilli, thus improving growth performance in weanling pigs [33][34][35][36] . In this study, we have shown that dietary supplementation with AMPs improved growth of juvenile goats under two types of concentrate conditions. These results suggest that AMPs can be used to promote growth performance in goats. This is consistent with the finding of Yoon et al. 36 who observed an improvement in the average daily gain and feed efficiency of weanling pigs fed diets supplemented with antimicrobial peptide-A3. Similarly, Jin et al. 35,37 observed an improvement in the average daily gain(ADG) of weanling pigs fed diets supplemented with antimicrobial peptides from Solanum tuberosum. Antimicrobial peptides beneficially affect host animals by improving their intestinal balance and creating gut microecological conditions [38][39][40] . In this study, we found that Proteobacteria were significantly decreased in the AMP-supplemented groups, while Fibrobacteres were significantly increased. This may be due to the fact that Fibrobacteres are anaerobic bacteria 41 , whereas Proteobacteria consist of aerobic bacteria that are mostly pathogenic 42 ; the antibacterial peptide could have inhibited the pathogenic bacteria while enhancing the total amount of anaerobes 17 . Dietary supplementation with AMPs has the potential to increase bacterial genera, such as Fibrobacter, Anaerovibrio, Succiniclasticum, and the ciliate genus Ophryoscolex, while reducing bacterial genera, including Selenomonas, Succinivibrio, and Treponema, and ciliate genera such as Polyplastron, Entodinium, and Isotricha. However, changes in Fibrobacter, Anaerovibrio, Ophryoscolex, Polyplastron, Entodinium, and Isotricha were related to the amount of concentrate that no significant different in the double concentrate group. Of these, Fibrobacter 43,44 , Treponema 45 , Ophryoscolex 46 , and Polyplastron 47 are cellulose-degrading microbes and Succiniclasticum 48 , Entodinium, and Isotricha 47 are starch-degrading microbes. Selenomonas and Succinivibrio degrade both starch and cellulose and Anaerovibrio 49 are fat-degrading bacteria. Therefore, we hypothesize that the increase in the relative abundance of Fibrobacter and Ophryoscolex in the normal concentrate group was due to an increase in xylanase and pectinase activities. Similarly, the decrease in the relative abundance of Isotricha and Entodinium was caused by a decrease in amylase activity in the normal concentrate group; whereas the increase in the relative abundance of Anaerovibrio was due to an increase in lipase activity in the same group.
Moreover, the fermentation products of Fibrobacter, Anaerovibrio, Treponema, Selenomonas, Ophryoscolex, Polyplastron, and Isotricha are acetate, propionate, and succinate; the fermentation product of Succinivibrio is succinate; and the fermentation products of Butyrivibrio are acetate and butyrate. Therefore, an increase in the relative abundance of Fibrobacter, Anaerovibrio, Ophryoscolex in the normal concentrate group may have caused an increase in acetate; whereas a decrease in the relative abundance of Treponema, Selenomonas, Polyplastron, and Isotricha in the double concentrate group may have led to the decrease in acetate. Lack of any variation to the relative abundance of Butyrivibrio prevented a change in butyrate. Acetate, propionate, and butyrate are the main components in VFAs, accounting for 95% of the total volatile matter content 50 . A change of the acetate + butyrate-to-propionate ratio is related to rumen fermentation mode. Thus, changes in acetate can cause alterations to T-VFA content (increase with normal concentrate and decrease with double concentrate) and the acetate + butyrate-to-propionate ratio (significant decrease with double concentrate). These results indicate that the effects of AMPs on rumen fermentation function and rumen microorganisms in goats were related to the amount of concentrate. It is possible that increased dosage causes similar changes in the double concentrate groups. However, additional studies will be needed to thoroughly elucidate these changes. The alpha diversity indices were elevated in the AMP-supplemented groups in this study, especially on day 60; indicating that AMP supplementation could increase microbial diversity in the rumen.
Previous studies 51,52 have demonstrated the importance of concentrate supplementation in goat growth and productivity. In this study, we found that the ADG increased with increasing concentrate amount. This result is consistent with the findings of Salim et al. 53 who reported that feeding grazing goats with concentrate supplement may optimize growth performance. The main reason for this may be changes in the rumen bacterial composition of ruminants driven by the amount of dietary concentrate 54 . In the present study, Firmicutes replaced Bacteroidetes as the dominant phylum in group (III) and Proteobacteria were significantly fewer compared to group I. This is in good agreement with data reported by Liu et al. 48 who reported that Firmicutes increased with a high concentrate diet. Similarly, Wetzels et al. 55 observed that Proteobacteria decreased and Firmicutes increased with increasing concentrate doses because of the ability of many Firmicutes to easily degrade fermentable carbohydrates. In terms of bacterial and ciliate genera, Prevotella increased with increasing concentrate amount, as reported also by Khafipour et al. 56 and Metzler-Zebeli et al. 57 . Prevotella is one of the most abundant genera in the rumen of goats because these bacteria possess highly diverse functions, in particular following a high-grain feeding regime. Anaerovibrio (on day 60) and Treponema increased significantly, whereas Succinivibrio, Fibrobacter (on day 20), Selenomonas (on day 60), and Diploplastron (on day 60) decreased significantly with increasing concentrate dosage. Therefore, the higher relative abundance of Prevotella and Anaerovibrio was due to augmented pectinase and lipase activities, which led to a further increase in T-VFA, ammonia, acetate, and MCP, and ultimately to enhanced goat growth performance.
Moreover, all indices were lower in group III compared to group I, indicating that bacterial diversity depended on dietary concentrate dosage. Similarly, Lillis et al. 58 reported that bacterial diversity was affected to a greater degree by a 90:10 than a 50:50 concentrate:forage ratio.
In summary, this study demonstrates that dietary supplementation with AMPs has beneficial effects on the growth performance, ruminal fermentation function, enzymatic activity, and rumen morphology of juvenile goats; and that these effects are related to concentrate amount. Therefore, AMPs could potentially be used as feed additives for juvenile goats on commercial farms. The detailed mechanism(s) by which AMPs promote growth of juvenile goats and improve their rumen microbial community structure require further clarification.

Materials and Methods
Ethics statement. All  Animal handling. Twenty-four, approximately four-month old, non-castrated Chuanzhong black goats, of average weight (16.17 ± 0.72 kg), were acclimated for 7 days prior to the experiment. All goats were caged and randomly allotted to four dietary treatment groups: I-normal concentrate group (300 g concentrate [per head per day]), II-normal concentrate and antimicrobial peptide group (300 g concentrate + 3.0 g AMPs), III-double concentrate group (600 g concentrate), and IV-double concentrate and antimicrobial peptide group (600 g concentrate + 3.0 g AMPs).
The diet included concentrate (Table 9) and forage (fresh grass). The groups were composed of three replicate pens with 2 goats each, animals were maintained in a house with free access to water, and fed twice daily (at 09:00 and 18:00); the animals maintained their normal herd behaviour.
Sampling and DNA extraction. Rumen fluid samples were collected using a stomach tube on days 0, 20 and 60, prior to morning feeding; the first part of the rumen fluid was discarded to prevent saliva interference. Three goats were selected from each group for sampling(one goats per pen). Rumen pH was measured immediately after collection using a portable pH meter (Model PHB-4, Shanghai Leica Scientific Instrument Co., Ltd., Shanghai, China). Solid feed particles were removed from the rumen fluid by filtration through 4 layers of cheesecloth. Samples were stored at −80 °C for later analysis. Microbial genomic DNA was extracted from rumen samples using a stool DNA kit (OMEGA Bio-Tek, Norcross, GA, USA), according to the manufacturer's instructions.
Ruminal fermentation function and enzyme activity analysis. Samples were prepared for VFA analysis and chromatography according to Luo et al. 59 . The concentration of NH 3 -N was analysed using visible-light spectrophotometry (Scientific BioMate 3 s, Thermo). NH 4 Cl standards were prepared according to Broderick and Kang 60 . Microbial protein (MCP) in the rumen was analyzed by trichloroacetic acid protein precipitation 61 . The activities of CMCase, xylanase, pectinase and β-glucosidase were measured using commercially available ELISA kits (R&D Systems, Minneapolis, MN, USA). Protease activity was measured as follows: a reaction mixture containing 1 mL casein and 4 mL protease enzyme was incubated for 4 h at 38 °C; at this point, the reaction was stopped by adding 10% trichloroacetic acid. The sample was then centrifuged at 3500 × g for 15 min. Next, 1 mL of supernatant was removed and mixed with 5 mL 0.4 M Na 2 CO 3 and 1 mL Folin-Ciocalteu's phenol solution and incubated on the laboratory bench for 15 min. The hydrolysed protein was measured using visual-light  spectrophotometry at 680 nm. Concentration and activity of lipase and amylase were measured using commercially available kits (NanJing JianCheng Bioengineering Institute, Nanjing, China).
Data analysis. Sequence reads were processed and analysed using QIIME pipeline software (version 1.8.0).
Chimeric sequences were removed to generate high quality sequences. High-quality sequences were divided and aligned into Operational Taxonomic Units (OTUs) with 97% sequence similarity using the QIIME pipeline software. The highest abundance sequences were compared with template regions in the Greengenes database (Release 13.8, http://greengenes.secondgenome.com/) (bacterial) and NCBI (http://www.ncbi.nlm.nih.gov) database (Ciliate protozoal), and were used to acquire taxonomic information for each OTU and species composition. Alpha diversity indices (including the Simpson index and Shannon index) were obtained using QIIME pipeline software. R software was used to analyze microfloral population structures. The results of these various analyses are expressed as means ± standard error of the mean (SEM). Statistical comparisons were made using paired sample t test via a commercially available statistical software package (SPSS 19.0, Business Machines Corporation, Armonk, NY, USA). Differences among treatments were regarded as significant at P < 0.05.