This study was approved by the Institutional Animal Care and Use Committee, Chung-Ang University, Seoul, Republic of Korea (NO: 2016-00105).

R. sphaeroides belonging to the Korean Collection Type Culture 1434 strain (http://kctc.kribb.re.kr/English/index.aspx, Korea) were cultivated on Van Niel’s medium [11], containing 1.0 g K₂HPO₄, 0.5 g MgSO₄, and 10 g yeast extract, at 20°C for 72 h, resulting in a final total concentration of 10⁹ cfu/mL (viable units). Total bacterial counts were determined using the method of Harris and Sommers [12]. Both in vitro and in vivo experiments were conducted using these cultured microorganisms as a direct-fed microbial (DFM).

To examine the effect of R. sphaeroides supplementation on the characteristics of ruminal fermentation, an in vitro batch culture experiment was conducted using the rumen contents of six ruminally cannulated Holstein dairy cows, collected 2 h post-afternoon feeding. Approximately 1 L of rumen content from each cow were filtered through four layers of muslin, pooled into a Thermos bottle, and immediately brought to the laboratory. Experimental diets consisted of four dietary treatments, including a control diet based on total mixed ration (TMR) with no supplement and three diets supplemented with R. sphaeroides culture at 0.1%, 0.3%, and 0.5% of TMR dry matter (DM) (v/w) in triplicate. The control diet for the in vitro study was identical to the one used for the in vivo study, which was a typical TMR diet for lactating cows. The chemical composition of the basal experimental diet is presented in Table 1. For the in vitro procedure, 100 mL of artificial rumen saliva [13] was placed in serum bottles containing 0.5 g of experimental diets under anaerobic techniques, in triplicate [14]. The filtered rumen fluid was injected (10%, v/v) into the serum bottles with continuously infused O₂-free CO₂ gas, and these serum bottles were crimped with butyl rubber stoppers and aluminum seals before being placed in a 39°C shaking water bath (100 rpm) and incubated for 2, 4, 8, 12, and 24 h. The volume of gas produced was measured at each time point by a pressure detector (model PSGH-28PCCA, DECO Co., Seoul, Korea) connected to a digital pressure transducer (DPT-03, Dail Information Co., Seoul, Korea). The supernatant of each incubation bottle was collected for pH determination and stored at −20°C for ammonia nitrogen (NH₃-N) [15], volatile fatty acids (VFAs) [16], and microbial protein synthesis [17] analyses. To determine DM digestibility, the incubated residues were transferred to a sintered glass crucible, cleaned, oven-dried, and weighed.

Six Holstein lactating cows (body weight 612±27 kg, milk yield 28.4±2.3 kg/d, and parity 2nd) equipped with permanent ruminal cannulae were used in the in vivo experiment, to examine if supplementing diets with R. sphaeroides cultures affected milk production and milk composition, particularly the concentration of CoQ10 in milk. Cows were randomly allocated to one of the two dietary treatments (n = 3): control TMR (C-TMR), identical to that used in the in vitro study, and TMR supplemented (S-TMR) with 0.5% of R. sphaeroides culture (TMR dry matter basis, v/w). Cows were allowed to adapt to the experimental diets for 20 days and at the end of the adaptation to the experimental diets an aliquot of milk samples (30 mL) were then collected at the same time, over three consecutive days and pooled per cow, immediately stored at −20°C, and freeze-dried prior to the CoQ10 analysis. Milk composition was analyzed using 50 mL of milk and Milko-Scan (FOSS-4000, FOSS, Denmark).

For the molecular analyses, R. sphaeroides genomic DNA was isolated from the rumen fluid of the cows fed on C-TMR and S-TMR, using a previously described method with minor modifications [18]. Briefly, genomic DNA was extracted by bead-beating (BioSpec Products, Bartlesville, OK, USA) for 4 min at full speed (2,000 strokes/min) in the presence of zirconium beads (weight 0.7 g, diameter 0.1 mm), 282 μL buffer A (NaCl 0.2 M, Tris 0.2 M, ethylenediaminetetraacetic acid 0.02 M; pH 8), 268 μL buffer B (QIAquick 96 PCR purification kit, Qiagen, Hileden, Germany), and 200 μL phenol-chloroform-isoamyl alcohol (25:24:1, pH 8). After centrifugation (16,000×g for 20 min at 4°C), the supernatant was thoroughly mixed with 650 μL buffer PB (Qiagen, Germany), and DNA was purified from the sample using the Qiagen PCR purification kit following the manufacturer’s protocol.

To conduct the denaturing gradient gel electrophoresis (DGGE) analysis, the photosynthetic reaction center M subunit gene, which is specific to photosynthetic bacteria such as R. sphaeroides, was amplified using the primers GC-clamp-557f (5′-CGC ACC TGG ACT GGA C -3′) and 750r (5′-CCC ATG GTC CAG CGC CAG AA -3′) as previously described [19]. The polymerase chain reaction (PCR) was performed in a TaKaRa Bio Ins. PCR Thermal Cycler (Kusatsu, Shiga, Japan) in a 25 μL final volume containing EmeraldAmp GT PCR MASTER mix (TaKaRa Bio Ins., Japan), 1 μL each primer (GC-cramp 557f and 750r), 2 U Taq polymerase (Ex Taq, TaKaRa Bio Inc., Japan), and 1 μL template DNA. Amplification cycles started with denaturation at 95°C for 5 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s, and ended with a final extension at 72°C for 10 min. Amplification success was checked by using 2% agarose gel electrophoresis and by visualizing PCR products in a Gel Doc XR+ system (Bio-Rad, Hercules, CA, USA). Following the purification of PCR products with the QIAquick PCR purification kit (Qiagen, Germany), DGGE was conducted using a D-Code system (Bio-Rad, USA) with 8% (w/v) polyacrylamide gels containing 20% to 80% denaturant gradient in 1×TAE buffer containing a mixture of Tris base, acetic acid and ethylenediaminetetraacetic acid. Equal amounts of purified PCR products were loaded on the gel and electrophoresis was performed at 130 V for 8 h. The gel was stained in 250 mL running buffer (ethidium bromide, Bio-Rad, USA) for 20 min and stained gels were photographed under UV using the Gel Doc XR⁺ documentation system (Bio-Rad, USA).

After electrophoresis, bands of interest were carefully excised with a sterile razor blade under UV illumination and then placed in 100 μL TE buffer containing Tris and ethylenediamintetraacetic acid for 24 h at 4°C. This solution containing DNA was again amplified using the primer pairs mentioned above, purified, and sent to Genotech (Daejeon, Korea) for sequencing. The nucleotide sequences obtained were compared to that deposited in the GenBank database (https://www.ncbi.nlm.nih.gov/genbank/) using the basic local alignment search tool (BLAST) algorithm [20]. Sequence identities were determined based on the highest identity score.

The concentrations of CoQ10 in rumen contents and milk were determined in a high performance liquid chromatograph (920-LC, Varian Inc., Palo Alto, CA, USA) equipped with an ultraviolet detector [21] and a liquid chromatography column (Zobax Eclipse Plus C18, 4.6×100 mm, 5.0 μm packing; Agilent technologies, Santa Clara, CA, USA). Rumen contents (10 mL) from the in vitro study were centrifuged at 500×g for 15 min to remove feed particles; 1 mL of the resulting supernatant was then re-centrifuged at 16,000×g for 20 min to isolate rumen microbial cells. The concentration of CoQ10 in the collected cells and cow milk samples was then analyzed as previously described [22,23].

Ruminal fermentation characteristics and milk composition were subjected to analysis of variance, with diet as the main effect, using the PROC MIXED procedure in the SAS program package [24]. Where necessary, the multiple comparison was performed by Duncan’s multiple range test [25]. Significance of the treatment was tested at 5% level. Orthogonal polynomial contrasts were performed to determine linear and quadratic effects of R. sphaeroides supplement.