Rumen liquid was collected from fistulated cows grazing the experimental pasture of the Agricultural University of China, Beijing, China, and filtered through a double cheesecloth. Approximately 1 mL of the rumen liquid was spread onto basal medium plates (5 g/L (NH₄)₂SO₄, 1 g/L KH₂PO₄, 1 g/L NaCl, 0.5 g/L MgSO₄·7H₂O, 0.1 g/L CaCl₂, 0.2 g/L yeast extract, 15 g/L agar) containing 2 g/L gossypol as the only source of carbon and was then cultured at 39°C under aerobic condition. The strains were inoculated on Luria-Bertani (LB) medium (5 g/L yeast extract, 10 g/L tryptone, 5 g/L NaCl, 30 g/L sucrose, and 15 g/L agar) and incubated for several days at 39°C to obtain pure cultures. The isolated strain was maintained at 4°C on LB medium.

To observe the colony characteristics of the isolated strain, the dilution plate method was used to obtain single colonies. The LB medium plates were inoculated with the strain by streaking. The isolated colonies were cultured for 6 h and then observed under the light microscope (Olympus BX50, Hamburg, Germany) for their size and shape. The isolated strain was cultured in LB medium for 2 d at 39°C and the thalli were collected. The thalli were resuspended in 2.5% (v:v) glutaraldehyde solution in 0.1 mol/L phosphate buffer saline (PBS), and fixed for 24 h. The thalli were washed three times with PBS. The thalli were then dehydrated by ethanol concentration gradient (10%, 30%, 50%, 70%, 90%; 10 min each), and dehydrated twice in 100% ethanol at 10 min intervals. For scanning electron microscope (SEM) assay, the thalli were washed with 50%, 70%, 90%, and 100% isoamyl acetate (each 3 min), critical point dried, coated with gold/palladium, and observed and photographed with a scanning electron microscope (model Philips SEMXL-20, Philips, Amsterdam, The Netherlands) [19].

The thalli of the isolated strain were inoculated into liquid LB medium and cultured for 24 h at 39°C with shaking, then harvested from the liquid medium by centrifugation. DNA was extracted using the sodium dodecyl sulfate (SDS)-proteinase K method [20] and purified DNA samples were stored at −20°C. The genomic DNA was amplified by polymerase chain reaction (PCR) using universal primers F1: 5′-AGAGTTTGATCCTGGCTCAG-3′ and R1: 5′-ACGGTTACCTTGTTACGACTT-3′ [21] and the PCR reaction was conducted using the following conditions: 5 min at 94°C, followed by 35 cycles of 30 s at 94°C, 30 s at 56°C, and 90 s at 72°C, and a final extension of 10 min at 72°C. The isolated strain 16S rRNA genes were sequenced by Shanghai Huada, China. The product of 16S rRNA genes of the isolates in the present study were compared against National Center for Biotechnology Information (NCBI) database reported sequences with the Basic Local Alignment Search Tool (BLAST) hosted at website of NCBI. A phylogenetic tree was then drawn using the Neighbor joining method [22]. Phylogenetic and molecular evolutionary analyses were conducted at 1000 bootstrap value using Molecular Evolutionary Genetics analysis (MEGA) version 5.0 software (Center of Evolutionary Functional Genomics, Biodesign Institute, Arizona State University, Tempe, AZ, USA) [23].

In order to optimize the fermentation conditions for FG in solid state fermentation (SSF) by the isolated strain, various fermentation conditions (initial moisture content [IMC], initial pH, incubation temperature, inoculum level and fermentation period) were analyzed. Different IMC levels of CSM, 35%, 40%, 45%, 50%, 55%, and 60%, were prepared by adjustment with distilled water. Different initial pH levels of the CSM, 5.5, 6.0, 6.5, and 7.0, were prepared by adjustment with 1 mol/L HCl or 1 mol/L NaOH. The fermentation was carried out at various temperatures (28°C, 30°C, 35°C, 39°C, and 41°C), with different inoculum levels (10³, 10⁴, 10⁵, 10⁶, 10⁷, and 10⁸ cells/g) of solid medium and for different incubation periods (48, 72, 96, and 120 h) to study the influence of these factors on FG.

The CSM, 200 g in each 500 mL conical flask, was inoculated with 20 mL of the isolated strain and 80 mL of double-distilled (dd) water, and incubated at 39°C for 72 h. After fermentation was complete, fermented substrates and the control (200 g CSM with 100 mL dd water) were dried in an oven at 60°C to constant weight [1,7]. Samples were subsequently processed into flour for analysis.

The CP content was determined by the Kjeldahl method, according to the State Standard of the People’s Republic of China (2010). Amino acids were assayed by the Biotechnology Center of Anhui Agriculture University according to the AOAC method [24], using a Hitachi L-8800 instrument (Hitachi, Japan).

Statistical analyses were completed using SPSS 14.0 (SPSS, Chicago, IL, USA). FG and TG differences between experimental and control group were considered significant when p<0.05. For statistical evaluations of the determined FG and TG, the date was analyzed for significant differences by analysis of variance using approximate tests.

A single strain, named Rumen Bacillus Subtilis (RBS), was isolated from the rumen liquid on basal medium plates. The strain grew well on LB medium containing sucrose (Figure 1A) but was also observed after cultivation for 3 d on sugar-free basal medium containing 2 g/L gossypol as the sole carbon source (Figure 1D). The diameter of the colonies on LB plates was larger than the colonies on gossypol plates. The appearance of colonies on the LB plates was rough and dirty white with wrinkly surfaces and irregular shapes (petal shaped or nearly circular) while the surface of colonies on the gossypol plates was smooth. The strain was rod-shaped by light microscopy and SEM observations confirmed this result (Figures 1B, 1C). The bacteria were positive by Gram-staining (Figure 1B).

The genomic DNA of strain RBS was successfully extracted by the SDS-proteinase K method. DNA was amplified with a bacteria-specific primer pair, which generates an amplicon of about 1.4 kb and is used for the analysis of a range of bacterial strains. Using purified DNA from the RBS isolate as template, a PCR amplicon with a size of around 1.5 kb was obtained (Figure 2). Sequence analysis showed the size was 1,433 bp (Figure 3). The 16S rDNA sequences of RBS were aligned using the NCBI-BLAST database, indicating that it is 98% homologous to the sequence of Bacillus subtilis strain GH38 (Figure 4). The presumption is that the RBS belongs to Bacillus subtilis.

Various culture times were tried to investigate their effect on the reduction of the FG level in CSM (Figure 5A). The FG content of CSM was 180.4 mg/kg after 48 h incubation with the RBS strain. With the fermentation course prolonged, the FG content decreased to 104.5 mg/kg by 72 h; this level was then maintained with longer incubation times. Thus, the optimum fermentation time was taken to be 72 h. Considering temperature, the lowest content of FG (98.2 mg/kg) was attained at 39°C (Figure 5B). The content of FG was higher at 28°C to 25°C or 41°C than at 39°C, so the incubation temperature chosen was 39°C. The effect of initial pH on FG level after SSF is shown in Figure 5C. Different pH levels of fermentation had a very significant effect. The FG level was 251.9 mg/kg at pH 5.5, and 93.2 mg/kg at pH 6.5. The FG content was increased relative to pH 6.5 at higher pH. In terms of IMC, the minimum FG was 68.5 mg/kg, observed at the moisture level 50% (Figure 5D). Other IMCs resulted in a higher FG level. The FG content decreased with increasing inoculum from 10³ to 10⁵ cell/g solid medium, while there were no significant changes at the levels 10⁶ to 10⁸ cell/g (p>0.05) (Figure 5E). The optimized inoculum level chosen was 10⁷ cell/g.

Retention times of the FG absorption peak in control were consistent with the fermented CSM, 6.578 to 6.664 min (Figure 6A, 6B). The content of FG in the control was 289.8 mg/kg, while fermented CSM contained 61.25 mg/kg. Thus, the FG content in fermented CSM decreased 78.86% relative to the control (p<0.01) (Figure 6C).

Retention times of the TG absorption peak in control were consistent with the fermented CSM, 6.290 to 6.663 min (Figure 6D, 6E). The content of total gossypol in the control was 2.52 g/kg, while in fermented CSM the content was 1.29 g/kg. Therefore, the TG content decreased 49% on fermentation relative to the control (p<0.01) (Figure 6F).

The CP content of the control was 357.1 g/kg, while in the fermented substrate the corresponding value was 374.9 g/kg (Table 1) that was improved markedly by 17.8 g/kg compared to the control. Amino acids levels, except for Cys, were increased in the fermented substrate. The contents of essential amino acids in the fermented CSM were also increased. Especially the levels of Arg, Met, and Lys in the fermented CSM were improved greatly, by 1.82, 1.35, and 0.94 g/kg respectively, compared with the control.