Journal of Animal Science, 2019, 4305–4313

doi:10.1093/jas/skz212 Advance Access publication June 28, 2019 Received: 8 June 2019 and Accepted: 26 June 2019 Reproduction

Reproduction Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 Bacterial taxonomic composition of the postpartum cow uterus and vagina prior to artificial insemination1 Taylor B. Ault,* Brooke A. Clemmons,* Sydney T. Reese,*,† Felipe G. Dantas,* Gessica A. Franco,*,† Tim P. L. Smith,‡ J. Lannett Edwards,* Phillip R. Myer,* and Ky G. Pohler*,†,2 *Department of Animal Science, University of Tennessee, Knoxville, TN 37996, †Department of Animal Science, Texas A&M University, College Station, TX 77843-2471, and ‡U.S. Meat Animal Research Center, Agricultural Research Service, United States Department of Agriculture, Clay Center, NE 68933

1The authors would like to thank the Angus Foundation for providing funding (Award A16-1356-001), USDA-NIFA Hatch/Multistate Project W3112- TEN00506—Reproductive performance in domestic animals, Brandon Beavers and the East Tennessee Research and Education Center for housing and caring for the animals, and Dr Liesel Schneider for statistical consulting.

2Corresponding author: [email protected]

ORCiD number: 0000-0002-1980-2105 (Phillip R. Myer).

Abstract The current study characterized the taxonomic composition of the uterine and vaginal bacterial communities during estrous synchronization up to timed artificial insemination TAI( ). Postpartum beef cows (n = 68) were subjected to pre- synchronization step 21 d prior to TAI (day −21), followed by an industry standard 7 Day Co-Synch on day −9 and TAI on day 0. Uterine and vaginal flushes were collected on days −21, −9, and −2 of the protocol and pH was immediately recorded. Pregnancy was determined by transrectal ultrasound on day 30. Bacterial DNA was extracted and sequenced targeting the V1 to V3 hypervariable regions of the 16S rRNA bacterial gene. Results indicated 34 different phyla including 792 different genera present between the uterus and vagina. Many differences in the relative abundance of and genera occurred between resulting pregnancy statuses and among protocol days (P < 0.05). At day −2, multiple genera were present in >1% abundance of nonpregnant cows but <1% abundance in pregnant cows (P < 0.05). Uterine pH increased in nonpregnant cows but decreased in pregnant cows (P > 0.05). Overall, our study indicates bacterial phyla and genera abundances shift over time and may potentially affect fertility by altering the reproductive tract environment.

Key words: cow, fertility, reproductive microbiome, uterus, vagina

Introduction et al., 1986; Wolfenson et al., 2000; Meikle et al., 2004; Rodney et al., 2018). In postpartum cows, the effects of these factors are Reproductive losses cost the beef and dairy industry over 1 billion particularly challenging due to the rapid changes during uterine dollars every year (Bellows et al., 2002). Fertility is affected by involution that must occur after parturition. Uterine involution many factors such as nutrition, heat stress, and parity (Richards involves the shedding of the maternal caruncles and regrowth

© The Author(s) 2019. Published by Oxford University Press on behalf of the American Society of Animal Science. All rights reserved. For permissions, please e-mail: [email protected].

4305 4306 | Journal of Animal Science, 2019, Vol. 97, No. 10

of the uterine endometrium, necessary for the cow to develop administered 2 d before TAI (day −2; average of 78 DPP). On the another successful pregnancy (Sheldon et al., 2008). During this day of TAI (day 0), cows were administered 100 μg of GnRH IM. In time, there is a high probability of bacterial contamination that order to reduce variation, a single sire and technician were used is most often cleared by uterine immune cells (Sheldon et al., for breeding. Transrectal ultrasound occurred 30 d following 2006). Colonization of pathogenic that are not cleared TAI for pregnancy diagnosis by observation of embryonic can lead to uterine diseases which may result in decreased heartbeat. Uterine and vaginal flush samples were collected fertility of beef and dairy cows (Földi et al., 2006; Sheldon et al., throughout the synchronization protocol at day −21, day −9, 2009; Potter et al., 2010). The presence of commensal bacteria, and day −2. The perineal area was cleaned and disinfected however, in the reproductive tract of cattle and its effect on prior to flush collections to prevent introduction of external health and fertility is relatively unknown. bacteria into the reproductive tract. For vaginal flush collection, The human reproductive tract microbiome has been well 60 mL of 0.9% sterile saline (Vetivex; pH = 5.6) was expelled by characterized, with a healthy microbiome consisting of low sterile syringe into the vagina and recovered via vaginal lavage. Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 diversity due to the presence of greater than 90% Lactobacillus Following vaginal flush collections, a sterile Foley catheter was species, creating a low pH environment (Rönnqvist et al., placed in the body of uterus to prevent the uterine flush from 2006). Shifts in normal bacterial communities, reducing the contamination. A large speculum was used as well as sterile presence of Lactobacillus, have been associated with bacterial chemise to prevent contamination while placing the catheter vaginosis or preterm birth (Ravel et al., 2011; Moreno et al., 2016). in the uterus. Similar to industry standard method for embryo Although the bovine reproductive microbiome is significantly flushing, sterile saline (180 mL) was flushed through the catheter more diverse than humans, similarities exist with the phyla into the uterus and collected. Uterine and vaginal flush pH was dominating both the uterus and vagina of humans measured by UltraBasic pH meter (Denver Instruments, Arvada, and bovine (Ravel et al., 2011; Laguardia-Nascimento et al., CO) and recorded immediately following collection. Samples 2015; Clemmons et al., 2017). The abundance of Lactobacillus were flash-frozen in liquid nitrogen and stored at −80 ℃ until in the bovine reproductive tract, however, is very low (Swartz analysis. To confirm expected response to the synchronization et al., 2014; Clemmons et al., 2017). Regarding diversity, our protocol, blood samples were collected in BD Vacutainer tubes group previously determined that although the bacterial (Becton, Dickinson and Company, Franklin Lakes, NJ) for diversity is high in the uterus and vagina compared to previous progesterone (P4) analysis at each time point of the protocol, human studies, the number of species significantly decreases along with evaluating the presence or absence of ovarian throughout estrous synchronization leading up to the time of structures by transrectal ultrasound. Twenty (10 pregnant and artificial insemination Ault( et al., 2019). Phylogenetic diversity 10 nonpregnant) cows were selected for sequencing and further of the bacteria also decreased over time, leading to resulting analysis. The criteria for cow inclusion in the study included: nonpregnant and pregnant cows having significant separation 1) presence of a corpus luteum (CL) on day −21, day −9 with P4 in their overall uterine phylogenetic diversity 2 d prior to prior greater than 1 ng/mL, 2) GnRH response on day −9 assumed by to timed artificial insemination (TAI; Ault et al., 2019). The the presence of a CL on day −2, 3) presence of an ovulatory size decreasing shift of bacterial diversity during the estrous cycle follicle on the day of TAI (day 0). may influence the bacterial taxa present in the uterus and vagina, altering their overall phylogenetic bacterial composition DNA Extraction and Sequencing and potentially affecting fertility. Therefore, our objective was to From the selected 20 cows, 120 flushes were collected resulting evaluate differences in bacterial relative abundances between in 10 pregnant and 10 nonpregnant cow samples from 3 cows that successfully established and maintained a pregnancy different time points in both the uterus and vagina. As described to day 30 after TAI and cows that were not pregnant at day 30 in Ault et al. (2019), DNA was extracted and sequenced from all after TAI. We hypothesized the taxonomic composition of the uterine and vaginal flush samples. Once samples were thawed uterus and vagina will differ between those that are pregnant to room temperature (22 °C), 5 mL aliquots were removed and 30 d after TAI versus cows that are not pregnant 30 d after TAI. centrifuged (4,696 × g at 4 °C). Sterile saline (180 µL) was added to the resulting pellet. The Qiagen DNeasy Blood and Tissue Materials and Methods kit (Qiagen, Hilden, Germany) was used for DNA extraction following manufacturer protocol. DNA amplification and library All experimental procedures involving the use of animals preparation was conducted using polymerase chain reaction for the current study was approved and performed under the (PCR) targeting the V1 to V3 hypervariable regions of the 16S rRNA Institutional Animal Care and Use Committee of the University bacterial gene for identification of bacteria. Modified universal of Tennessee, Knoxville. The methods below are similar as primers 27F (5′-Adapter/Index/AGAGTTTGATCCTGGCTCAG) described in Ault et al. (2019). and 519R (5′-Adapter/Index/GTATTACCGCGGCTGCTG) including TruSeq indices and adapters were used to produce Experimental Design sequencing libraries, using Accuprime Taq high fidelity DNA Angus beef cows (n = 68) enrolled in this study, were located at polymerase (LifeTechnologies, Carlsbad, CA). Conditions for the East Tennessee Research and Education Center, averaging PCR reaction were set at an annealing temperature of 58 °C 80 ± 2.6 days postpartum (DPP) and 4.6 ± 0.57 yr old at TAI. No for 30 cycles. Gel electrophoresis was used to confirm quality cows displayed any clinical signs of disease or reproductive and presence of PCR product. Purification and quantification tract abnormalities. Twenty-one days prior to TAI all cows were of resulting PCR product was conducted using AmPure beads administered prostaglandin F2α (Lutalyse, 25 mg as 5 mL) IM as a and Nanodrop 1000 spectrophotometer (Agencourt, Beverly, pre-synchronization step. Nine days before TAI (day −9; average MA and ThermoScientific, Wilmington, DE) as well as real-time of 71 DPP), an industry standard 7 Day Co-Synch Protocol was PCR on LightCycler 480 system (Roche Diagnostics, Mannheim, implemented beginning with gonadotropin-releasing hormone Germany). Libraries were sequenced at the United States Meat (GnRH; Factrel, 100 μg as 2 mL IM injection). An injection of Animal Research Center (USDA-ARS-USMARC, Clay Center, NE) prostaglandin F2α (Lutalyse, 25 mg as 5 mL) IM was then by the Illumina MiSeq platform (Illumina, Inc., San Diego, CA) Ault et al. | 4307

with the 2 × 300, v3 600 cycle kit. For sequencing controls, PhiX observed OTU value < 10 were removed from analysis. Bacterial was used as a positive control and sterile saline, used for flush taxa abundances were non-normally distributed and could not collections, as a negative control. be transformed to achieve a normal distribution. Nonparametric ANOVA in SAS 9.4 was used for all bacterial abundance data Sequence Reading and Analysis with the independent variables day of protocol or pregnancy Quantitative Insights Into Microbial Ecology (QIIME) version status. Significance was determined by Wilcoxon exact test 1.9.1 (Caporaso et al., 2010) and Mothur version 1.36.1 (Schloss for comparisons between pregnancy statuses and Kruskal– et al., 2009) programs were used to process resulting sequence Wallis test for comparisons among days of the protocol, with reads. Quality filtering was conducted on the Galaxy server to the Benjamini–Hochberg FDR multiple test correction (Weiss retain all sequences with quality scores ≥ Q 25. Adapter and et al., 2017). Days postpartum and age were not included in index sequences were trimmed from sequences and sequences final model as they were determined to have no significant Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 < 300 bp were removed from analysis. Using the usearch61 effect. Correlations were performed in SAS 9.4 using Pearson command (Edgar, 2010), chimeric sequences were identified and correlation. Significance was determined by P ≤ 0.05 for all removed. Identified chloroplast and mitochondria sequences analyses. were removed from analyses. Samples were then subsampled down to ~30,000 sequences to remove sequence depth bias. In order to perform taxonomic classification, operational Results taxonomic units (OTUs) were first selected with a pairwise identity threshold of 97% using UCLUST in QIIME. Sequence Information assignment then occurred using UCLUST and the Greengenes As reported in Ault et al. (2019), 10,448,316 total sequences v13_8 16S rRNA database (Caporaso et al., 2010). resulted after chimera removal and quality control. An average of 92,463 ± 2,976 sequences were present per sample Progesterone Assay with number of sequences ranging from 42 to 232,160 among Progesterone RIA was performed according to the previously individual samples. described protocol (Pohler et al., 2016) using a double-antibody RIA kit (MP Biomedicals, Santa Ana, CA). A standard curve Phylum Level Taxonomic Composition (ranging from 0 to 50 ng/mL) was generated to determine Among all samples, 34 total phyla were detected. Figure 1 sample concentrations and in-house controls were used for presents the phyla relative abundance in the uterus and vagina quality control. Inter- and intra-assay CV were less than 10%. of pregnant or nonpregnant cows over time.

Statistical Analysis Between resulting pregnant and nonpregnant cows All analyses were separated between uterine and vaginal data Firmicutes was the most abundant phyla across all samples, and did not compare the 2 environments. Samples with an but no significant differences in the relative abundance was

Figure 1. Relative abundance of phyla. A total of 34 phyla were detected among all samples. Samples were grouped by day, uterus or vagina, and pregnancy status. O represents nonpregnant cows and P represents pregnant cows. 4308 | Journal of Animal Science, 2019, Vol. 97, No. 10

observed between pregnant and nonpregnant cows in the uterus genera in the uterus at day −2 in pregnant cows with relative or vagina. The most significant differences in relative abundance abundance greater than 1%, other than the undetermined genus of phyla between resulting pregnant and nonpregnant cows of the family Ruminococcaceae, included the classified genera occurred at day −9 in the uterus with (P = 0.05), of Ureaplasma, Helcococcus, Bacteroides, 5-7N15, and Oscillospira, (P = 0.05), (P = 0.05), as well as unassigned genera of order Burkholderiales, order (P = 0.009), and unassigned taxa (P = 0.03; Table 1). Only 3 Clostridiales, family Bacteroidaceae, family Lachnospiraceae, phyla were significantly different at day −21 and day −2 in the order Bacteroidales, and family Rikenellaceae. In nonpregnant uterus between pregnant and nonpregnant cows with Thermi cows, Corynebacterium had the greatest abundance in the (P = 0.005), (P = 0.03), and FBP (P = 0.05) at day uterus at day −2 (9.62 ± 4.97%) and a significantly lower relative −21, and Actinobacteria (P = 0.005), (P = 0.02), and abundance (0.50 ± 0.36%) in the uterus of pregnant cows unassigned taxa (P = 0.009) at day −2 (Table 1). In the vagina, (P = 0.003). Genera with relative abundance greater than 1% in the only 2 phyla were significantly different with uterus at day −2 of nonpregnant cows that significantly differed Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 (P = 0.01) and Fusobacteria (P = 0.03) at day −9 and Tenericutes from pregnant cows, other than Corynebacterium, included (P = 0.03) and Acidobacteria (P = 0.05) at day −2 between resulting classified generaStaphylococcus (P = 0.0002), Prevotella (P = 0.03), pregnant and nonpregnant cows (Table 1); no significant Microbacterium (P = 0.01), Butyrivibrio (P = 0.05), Ralstonia (P = 0.01), differences between pregnant and nonpregnant cows in phyla and unassigned genera from family Alcaligenaceae (P = 0.0001), relative abundance were detected at day −21. and family Comamonadaceae (P = 0.02; Table 3). Interestingly, these genera in resulting nonpregnant cows with greater than Days of the TAI protocol 1% relative abundance had a relative abundance less than 1% The most abundant phyla, Firmicutes, significantly decreased in resulting pregnant cows (Table 3). Fewer differences in the over time in the uterus of pregnant (P = 0.04) and nonpregnant relative abundances of genera occurred between pregnant and (P = 0.0002) cows as well as in the vagina of nonpregnant cows nonpregnant cows in both the uterus and vagina at day −21 over time (P = 0.03; Table 2). When observing the change in phyla (Supplementary Tables 1 and 4) and day −9 (Supplementary relative abundance over time of the synchronization protocol, Tables 2 and 5) compared to day −2 in the uterus (Table 3; numerous differences occurred in the uterus and vagina of Supplementary Table 3) resulting pregnant and nonpregnant animals (Table 2). Days of the TAI protocol Genus Level Taxonomic Composition The most abundant genera detected, undetermined from the At the genus level, 792 total genera were detected among all family Ruminococcaceae, significantly changed over time samples. Figure 2 presents the genera relative abundance in in the uterus and vagina of both pregnant and nonpregnant the uterus or vagina of resulting pregnant or nonpregnant cows animals (P < 0.05). Numerous other genera relative abundances over time. significantly changed over the duration of the protocol (Supplementary Tables 7–10). Between resulting pregnant and nonpregnant cows Relative abundances for each significantly different genera Circulating Progesterone Concentrations and between pregnant and nonpregnant cows in the uterus and vagina Relative pH Correlation to Bacteria Phyla at each day of the protocol are listed in Supplementary Tables As reported in Ault et al. (2019), day −21 (2.34 ± 0.62 ng/mL vs. 1–6. The undetermined genus of the family Ruminococcaceae 2.2 ± 0.74 ng/mL), day −2 (6.11 ± 1.81 ng/mL vs. 4.76 ± 1.77 ng/ had the greatest abundance of all samples, differing between mL), and day 0 (0.83 ± 0.31 ng/mL vs. 0.31 ± 0.09 ng/mL) pregnant and nonpregnant cows only at day −21 in the vagina. circulating progesterone concentrations were not significantly The most differences in genera relative abundance between different between resulting pregnant and nonpregnant animals pregnant and nonpregnant cows, occurred at day −2. The as expected. At day −9, however, progesterone was greater in

Table 1. Percent relative abundance of significant phyla (mean ± SEM) between nonpregnant and pregnant cows in the uterus and vagina at each day1,2

Type Day Phylum Nonpregnant Pregnant P-value

Uterine −21 Thermi 0.003 ± 0.001 0.025 ± 0.014 0.005 Acidobacteria 0.002 ± 0.001 0.011 ± 0.006 0.029 Uterine −9 Bacteroidetes 18.257 ± 0.636 16.449 ± 0.832 0.047 Actinobacteria 2.346 ± 0.717 3.800 ± 0.698 0.047 Spirochaetes 0.029 ± 0.004 0.026 ± 0.009 0.049 Unassigned 0.003 ± 0.001 0.010 ± 0.003 0.027 Fusobacteria 0.002 ± 0.001 0.010 ± 0.009 0.009 Uterine −2 Actinobacteria 14.403 ± 7.360 1.338 ± 0.761 0.005 Fibrobacteres 0.271 ± 0.213 0.041 ± 0.034 0.018 Unassigned 0.009 ± 0.004 0.001 ± 0.001 0.009 Vaginal −9 Fusobacteria 0.019 ± 0.015 ND 0.029 Verrucomicrobia 0.005 ± 0.003 0.019 ± 0.005 0.010 Vaginal −2 Tenericutes 2.547 ± 1.195 1.010 ± 0.167 0.032 Acidobacteria 0.009 ± 0.005 0.002 ± 0.001 0.053

1Significance determined by P ≤ 0.05. No significant differences were detected at day −21 in the vagina. 2ND: not detectable at current depth. Ault et al. | 4309

Table 2. Percent relative abundance of significant phyla (mean ± SEM) among protocol days in the uterus and vagina of nonpregnant and pregnant cows1,2

Type Status Phylum Day −21 Day −9 Day −2 P-value

Uterine Nonpregnant Firmicutes 61.301 ± 4.691 74.312 ± 2.225 36.020 ± 5.970 0.001 5.444 ± 2.101 3.336 ± 1.646 27.676 ± 9.456 0.013 Tenericutes 2.264 ± 0.541 0.932 ± 0.161 7.395 ± 6.402 0.027 Verrucomicrobia 0.221 ± 0.058 0.015 ± 0.003 0.115 ± 0.034 0.011 Fusobacteria 0.015 ± 0.005 0.002 ± 0.001 0.027 ± 0.015 0.033 Thermi 0.003 ± 0.001 0.069 ± 0.024 0.032 ± 0.013 0.043 OD1 0.001 ± 0.001 ND 0.014 ± 0.008 0.008 FBP ND 0.003 ± 0.002 ND 0.047 Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 Uterine Pregnant Firmicutes 65.182 ± 3.127 69.480 ± 3.272 45.544 ± 9.283 0.043 Actinobacteria 3.709 ± 0.992 3.800 ± 0.698 1.338 ± 0.761 0.011 Lentisphaerae 0.514 ± 0.081 0.452 ± 0.064 0.255 ± 0.148 0.014 Fibrobacteres 0.301 ± 0.126 0.054 ± 0.016 0.041 ± 0.034 0.004 Spirochaetes 0.134 ± 0.035 0.026 ± 0.009 0.127 ± 0.079 0.037 0.034 ± 0.005 0.057 ± 0.015 0.018 ± 0.010 0.015 Unassigned 0.010 ± 0.004 0.010 ± 0.003 0.001 ± 0.001 0.006 WPS-2 0.005 ± 0.003 0.002 ± 0.001 ND 0.044 0.002 ± 0.001 ND ND 0.015 OD1 ND ND 0.005 ± 0.003 0.047 Vaginal Nonpregnant Firmicutes 81.663 ± 1.533 79.483 ± 2.291 70.667 ± 3.842 0.032 Proteobacteria 1.248 ± 0.377 0.898 ± 0.283 6.626 ± 3.384 0.045 TM7 0.105 ± 0.034 0.025 ± 0.008 0.285 ± 0.078 0.001 Thermi 0.001 ± 0.001 0.016 ± 0.006 0.016 ± 0.007 0.033 Spirochaetes 0.051 ± 0.019 0.016 ± 0.005 0.126 ± 0.035 0.004 Acidobacteria ND 0.001 ± 0.001 0.009 ± 0.005 0.004 0.037 ± 0.018 0.008 ± 0.003 0.081 ± 0.011 0.005 Verrucomicrobia 0.180 ± 0.138 0.005 ± 0.003 0.175 ± 0.065 0.001 Fibrobacteres 0.043 ± 0.018 0.019 ± 0.008 0.121 ± 0.048 0.034 OD1 ND ND 0.071 ± 0.050 0.027 Vaginal Pregnant Proteobacteria 0.959 ± 0.152 1.048 ± 0.354 1.968 ± 0.358 0.014 TM7 0.084 ± 0.023 0.046 ± 0.019 0.317 ± 0.118 0.020 Planctomycetes 0.014 ± 0.006 0.006 ± 0.003 0.076 ± 0.016 0.001 Verrucomicrobia 0.079 ± 0.038 0.019 ± 0.005 0.202 ± 0.094 0.026 Unassigned 0.012 ± 0.003 0.003 ± 0.001 0.003 ± 0.001 0.018 SR1 0.012 ± 0.011 ND ND 0.047 0.002 ± 0.001 ND ND 0.015

1Significance determined by P ≤ 0.05. 2ND: not detectable at current depth. nonpregnant cows than pregnant cows (P = 0.01; 7.58 ± 1.30 ng/ vs. 6.29 ± 0.11), or day −2 (6.69 ± 0.14 vs. 6.58 ± 0.14). Regardless mL vs. 3.83 ± 0.68 ng/mL), but no day × status was observed. In of pregnancy status, vaginal pH changed significantly overtime addition, correlation of circulating progesterone concentration (P < 0.0001) with the lowest relative pH on day −9 (6.30 ± to the relative abundance of phyla indicated significance with 0.07) compared to day −21 (6.90 ± 0.15; P = 0.0002) and day −2 Firmicutes. In the vagina of nonpregnant cows, as circulating (6.63 ± 0.10; P = 0.0009). Additionally, pH values were tested for progesterone concentration decreased, the relative abundance correlation with genera abundances in the uterus and vagina. of Firmicutes increased at both day −9 (r = −0.79, P = 0.02) and In the uterus, genera Mogibacterium (r = −0.25, P = 0.057) and day −2 (r = −0.81, P = 0.004). At day −2, however, an increase unassigned genera from the family Lachnospiraceae (r = −0.35, in progesterone was significantly correlated to an increase P = 0.008) and family Clostridiaceae (r = −0.27, P = 0.04) were in relative abundance of Proteobacteria in the vagina of negatively correlated with pH. As the abundances of these nonpregnant cows (r = 0.70, P = 0.03). genera increased, uterine pH decreased. In the vagina, however, Regardless of the day of the TAI protocol, the uterus had a genera Clostridium (r = 0.28, P = 0.04), unassigned genera from the lower relative pH, with means ranging from 5.86 ± 0.11 to 6.06 ± family Lachnospiraceae (r = 0.31, P = 0.02), and 2 different genera 0.09, compared to the vagina with means ranging from 6.31 ± from the unassigned from the order Clostridiales (r = 0.30, 0.09 to 7.04 ± 0.17. In the uterus, no effect of day or status was P = 0.02 and r = 0.36, P = 0.007) were positively correlated to pH detected on relative uterine pH. Although no significant day× with an increase in abundance associated with an increase in status occurred, uterine relative pH numerically increased in pH. resulting nonpregnant cows from 5.86 ± 0.11 at day −21 to 6.06 ± 0.09 at day −2 and decreased in resulting pregnant cows from Discussion 6.04 ± 0.20 at day −21 to 5.92 ± 0.13 at day −2. In the vagina, no overall effect of status or day × status on relative pH was The uterine and vaginal bacterial communities of cattle have detected (P > 0.05) between resulting nonpregnant and pregnant been determined to contain greater diversity compared to the cows at day −21 (7.04 ± 0.17 vs. 6.76 ± 0.24), day −9 (6.31 ± 0.09 human reproductive microbiome (Swartz et al., 2014; Ault et al., 4310 | Journal of Animal Science, 2019, Vol. 97, No. 10 Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020

Figure 2. Relative abundance of genera. A total of 792 genera were detected among all samples. The top 30 most abundant genera are presented with the Other category including all genera below the top 30 most abundant genera detected. Samples were grouped by the day, uterus or vagina, and pregnancy status. O represents nonpregnant cows and P represents pregnant cows

Table 3. Significantly different genera between nonpregnant and multiple genera that have previously been determined to be pregnant cows at day −2 in the uterus with a relative abundance commonly pathogenic differed between resulting pregnant and 1 greater than 1% nonpregnant cows. These genera were determined to be present in the uterus at day −2 with abundances greater than 1% in cows Genus Nonpregnant Pregnant P-value that did not become pregnant, but these same bacteria were Corynebacterium 9.623 ± 4.974 0.499 ± 0.358 0.003 present in less than 1% in cows that became pregnant. Clemmons Family Alcaligenaceae 9.491 ± 9.407 0.013 ± 0.011 0.001 et al. (2017) found Corynebacterium as the most abundant genus Staphylococcus 3.283 ± 2.895 0.024 ± 0.018 0.001 in the uterus 2 d prior to TAI in nonpregnant cows. Previously, Prevotella 2.819 ± 2.613 0.032 ± 0.019 0.031 Corynebacterium was determined to be highly abundant in Microbacterium 2.219 ± 2.185 0.009 ± 0.007 0.009 postpartum cows that had developed uterine infections and led Butyrivibrio 1.363 ± 0.433 0.789 ± 0.388 0.049 to negative effects on the ability to conceive another pregnancy Ralstonia 1.353 ± 0.780 0.041 ± 0.023 0.007 (Ruder et al., 1981). The present study found significantly lower Family 1.130 ± 0.482 0.341 ± 0.214 0.023 abundances of Corynebacterium in postpartum cows that were Comamonadaceae able to develop a pregnancy, supporting previous evidence 1All genera relative abundances that significantly differ between of Cornyebacterium's potential negative effects on fertility. In nonpregnant and pregnant cows at day −2 in the uterus are addition, results indicated Staphylococcus as the second most presented in Supplementary Table 5. abundant genus detected in the uterus of nonpregnant cows at day −2, which was also significantly less in the uterus of 2019). Throughout the estrous cycle, the diversity of bacterial pregnant cows at day −2. Staphylococcus is often found to be communities significantly decreases as the reproductive present in the uterus of postpartum cows that develop acute tract environment prepares for pregnancy (Ault et al., 2019). metritis and known as the most common pathogen causing The shift occurring in the diversity of the bacteria present mastitis (Vasudevan et al., 2003; Otero and Nader-Macías, leads to differences in the taxonomic composition of these 2006). Unexpectedly, Ureaplasma and Helcococcus genera had the communities, potentially affecting fertility. In the current study, greatest abundance detected in the uterus of pregnant cows at Ault et al. | 4311

day −2 but did not significantly differ between pregnant and Bavister, 2000). In the current study, vaginal pH was greater than nonpregnant cows. Previously, these bacteria have been shown uterine pH prior to breeding which may allow for a favorable to have infectious potential and are a factor in the development environment for sperm ascension into the uterus during natural of uterine metritis with the potential to lead to reproductive breeding as sperm is deposited into the vagina. Additionally, a issues such as infertility or abortion in cattle (Vasconcellos decrease in uterine pH at estrus has been shown to lead to a Cardoso et al., 2000). Similar results to the current study were favorable environment for sperm to reside prior to fertilization found by Jeon et al. (2015) with increased Ureaplasma correlated of the oocyte, contributing to increased pregnancy rates (Perry to a healthier uterine environment than those that developed and Perry, 2008a, 2008b). The relationship between the change metritis after calving. Further research needs to be conducted in pH and bacterial communities in the reproductive tract has to determine the specific species level differences and how they not previously been evaluated in bovine. Interestingly, our study relate to the development of uterine disease or presence in the found significant correlations of bacterial genera associated with healthy uterus. It is important to note that the current study increased relative vaginal pH, with other bacteria associated Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 cannot rule out effects of repeated flushing of the reproductive with decreased relative uterine pH. Although the current study tract prior to TAI. A recent study by Martins et al. (2018), reported does not report true pH measurements, rather the relative repeated flushing of the reproductive tract prior to embryo change in pH, the decreased species number and phylogenetic transfer, led to decreased pregnancy rates. Further, there is the diversity with the change in abundances of bacteria occurring possibility that the observed bacterial communities may be a in the uterus leading to TAI may contribute to the change in pH. result of bacteria introduced from previous pregnancies since all Previous studies in humans determined estrogen stimulates animals in the current study were multiparous; however, these epithelial cells of the reproductive tract to produce glycogen, a experiments were designed to minimize these effects. fuel source for Lactobacillus, which dominates the reproductive The indicated presence of pathogenic bacteria in microbiota (Boskey et al., 1999; Lamont et al., 2011; Ravel et al., the reproductive tract may have negative effects on the 2011; Brotman et al., 2014). Boskey et al. (1999) indicated the development and maintenance of pregnancy; however, other glycogen produced is sufficient forLactobacillus to produce a high present bacteria may be contributing to a healthy or optimal concentration of lactic acid to maintain a low pH environment, reproductive tract environment. The unassigned genera from suggesting the microbiota are the major contributor to the pH the family Ruminococcaceae was determined to have the of the reproductive tract to potentially protect against pathogen greatest abundance in the reproductive tract in the current colonization (Eschenbach et al., 1989; Boskey et al., 1999). Future study. Although the specific species present has yet to be studies may be able to use bacteria known to lower pH, such classified, the Ruminococcaceae family of bacteria is known as Lactobacillus, in probiotic form to manipulate the microbiome to be highly abundant in the gastrointestinal tract of healthy of the bovine reproductive tract and evaluate the effect on humans and cattle (Flint et al., 2012; Rajilić-Stojanović and improving pregnancy establishment. de Vos, 2014; Mao et al., 2015). These bacteria contribute to In addition to pH, hormone concentrations have been suggested carbohydrate degradation resulting in the production of short- to affect the presence of some bacterial species (Sandrini et al., chain fatty acids (SCFA) such as butyrate (Forbes et al., 2016; 2015; Org et al., 2016). Otero et al. (1999) found greater abundances Zheng et al., 2017). Butyrate has been previously determined of bacteria in bovine bacterial cultures in the presence of high to prevent local inflammation by increasing the population of estrogen, or low progesterone, suggesting bacterial community regulator T cells (Tregs; Smith et al., 2013; Zhang et al., 2016). As abundances shift through different phases of the estrous cycle. the regulation of inflammation in the reproductive tract by Tregs Similarly, our results indicated the shift of Firmicutes in the has been demonstrated to be important for the establishment vagina was correlated to progesterone concentration, with of pregnancy (Shima et al., 2010) and to prevent rejection of decreased progesterone indicating an increase in Firmicutes. In the fetus, the high abundance of bacteria from the family contrast, an increase in Proteobacteria relative abundance in the Ruminococcaceae may be contributing fermentation products vagina was correlated to an increase in progesterone, suggesting to regulate immune cell populations in the reproductive tract the response to hormonal concentrations may differ by species. to maintain the proper inflammatory environment (Aluvihare No differences in progesterone concentration were detected in et al., 2004). Additionally, other bacteria present in lower the current study between pregnant and nonpregnant cows at abundances may also have positive effects on the reproductive day −21, day −2, or day 0 indicating a physiological response to the tract environment. Although Lactobacillus abundance is low in protocol prior to breeding and no effect on the bacteria differences the reproductive tract of cattle, as confirmed by the current observed at each day. The shifts in hormonal concentrations study, these bacteria may be beneficial to the reproductive tract throughout the estrous cycle, however, may influence the change environment. Genís et al. (2018) used strains of Lactobacillus as a in bacterial abundances over time. Further research is necessary probiotic administered into the vagina and uterus of dairy cows to determine the mechanisms contributing to the fluctuations prior to calving. They found cows treated with Lactobacillus had in the reproductive tract microbiome and circulating or local a lower incidence of metritis, suggesting lactic acid-producing hormone concentrations. bacteria may provide benefits to the reproductive tract health In conclusion, the taxonomic composition in the uterus and and fertility (Genís et al., 2018). vagina shifts in the relative abundance of bacteria throughout The overall environment of the uterus plays a critical role in the TAI synchronization protocol. The bacterial diversity was proper sperm transport for successful fertilization. The present previously determined to change over time in the uterus study indicated a decrease in uterine pH leading up to TAI in (Ault et al., 2019), resulting in differences in phyla and genera cows that became pregnant, but an increase in pH in those who relative abundances. Although various factors may affect the failed to conceive. These results support evidence that a lower development and maintenance of pregnancy, our study suggests uterine pH at the time of sperm deposition may be beneficial to differences in bacterial abundances prior to breeding may affect fertilization. Studies in cattle have shown that sperm motility is the reproductive tract environment potentially affecting fertility increased in higher pH environments and inhibited by a lower outcomes. Future research should determine potential causes pH environment, leading to an increased life span (Jones and of altered bacterial abundances shifting toward a pathogenic 4312 | Journal of Animal Science, 2019, Vol. 97, No. 10

microbiome preventing the development and maintenance R. C. Bicalho, S. Lima, et al. 2015. Uterine microbiota of pregnancy, or toward a healthy microbiome supporting a progression from calving until establishment of metritis in successful pregnancy. Additional research may evaluate the use dairy cows. Appl. Environ. Microbiol. 81:6324–6332. doi:10.1128/ of probiotics to potentially maintain a healthy microbiome in the AEM.01753-15 Jones, J. M., and B. D. Bavister. 2000. Acidification of intracellular reproductive tract to reduce fertility issues and help producers pH in bovine spermatozoa suppresses motility and extends improve their herd's reproductive efficiency. viable life. J. Androl. 21:616–624. doi:10.1002/j.1939-4640.2000. tb02128.x Supplementary Data Laguardia-Nascimento, M., K. M. Branco, M. R. Gasparini, S. Giannattasio-Ferraz, L. R. Leite, F. M. Araujo, A. C. Salim, Supplementary data are available at Journal of Animal Science J. R. Nicoli, G. C. de Oliveira, and E. F. Barbosa-Stancioli. 2015. online. Vaginal microbiome characterization of Nellore cattle using metagenomic analysis. PLoS One 10:e0143294. doi:10.1371/ Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 Literature Cited journal.pone.0143294 Lamont, R. F., J. D. Sobel, R. A. Akins, S. S. Hassan, Aluvihare, V. R., M. Kallikourdis, and A. G. Betz. 2004. Regulatory T. Chaiworapongsa, J. P. Kusanovic, and R. Romero. 2011. The T cells mediate maternal tolerance to the fetus. Nat. Immunol. vaginal microbiome: new information about genital tract 5:266–271. doi:10.1038/ni1037 flora using molecular based techniques. BJOG 118:533–549. Ault, T. B., B. A. Clemmons, S. T. Reese, F. G. Dantas, G. A. Franco, doi:10.1111/j.1471-0528.2010.02840.x T. P. L. Smith, J. L. Edwards, P. R. Myer, and K. G. Pohler. 2019. Mao, S., M. Zhang, J. Liu, and W. Zhu. 2015. Characterising the Uterine and vaginal bacterial community diversity prior to bacterial microbiota across the gastrointestinal tracts of artificial insemination between pregnant and non-pregnant dairy cattle: membership and potential function. Sci. Rep. postpartum cows. J. Anim. Sci. doi:10.1093/jas/skz210 5:16116. doi:10.1038/srep16116 Bellows, D. S., S. L. Ott, and R. A. Bellows. 2002. Review: cost of Martins, T., G. Pugliesi, M. Sponchiado, A. M. Gonella-Diaza, reproductive diseases and conditions in cattle. Prof. Anim. O. A. Ojeda-Rojas, F. D. Rodriguez, R. S. Ramos, A. C. Basso, Scient. 18:26–32. doi:10.15232/S1080-7446(15)31480-7 and M. Binelli. 2018. Perturbations in the uterine luminal fluid Boskey, E. R., K. M. Telsch, K. J. Whaley, T. R. Moench, and composition are detrimental to pregnancy establishment R. A. Cone. 1999. Acid production by vaginal flora in vitro is in cattle. J. Anim. Sci. Biotechnol. 9:70. doi:10.1186/ consistent with the rate and extent of vaginal acidification. s40104-018-0285-6 Infect. Immun. 67:5170–5175. Meikle, A., M. Kulcsar, Y. Chilliard, H. Febel, C. Delavaud, Brotman, R. M., J. Ravel, P. M. Bavoil, P. E. Gravitt, and K. G. Ghanem. D. Cavestany, and P. Chilibroste. 2004. Effects of parity and 2014. Microbiome, sex hormones, and immune responses in body condition at parturition on endocrine and reproductive the reproductive tract: challenges for vaccine development parameters of the cow. Reproduction 127:727–737. doi:10.1530/ against sexually transmitted infections. Vaccine 32:1543–1552. rep.1.00080 doi:10.1016/j.vaccine.2013.10.010 Moreno, I., F. M. Codoñer, F. Vilella, D. Valbuena, J. F. Martinez- Caporaso, J. G., J. Kuczynski, J. Stombaugh, K. Bittinger, Blanch, J. Jimenez-Almazán, R. Alonso, P. Alamá, J. Remohí, F. D. Bushman, E. K. Costello, N. Fierer, A. G. Peña, A. Pellicer, et al. 2016. Evidence that the endometrial microbiota J. K. Goodrich, J. I. Gordon, et al. 2010. QIIME allows analysis of has an effect on implantation success or failure. Am. J. Obstet. high-throughput community sequencing data. Nat. Methods Gynecol. 215:684–703. doi:10.1016/j.ajog.2016.09.075 7:335–336. doi:10.1038/nmeth.f.303 Org, E., M. Mehrabian, B. W. Parks, P. Shipkova, X. Liu, T. A. Drake, Clemmons, B. A., S. T. Reese, F. G. Dantas, G. A. Franco, T. P. L. Smith, and A. J. Lusis. 2016. Sex differences and hormonal effects on O. I. Adeyosoye, K. G. Pohler, and P. R. Myer. 2017. Vaginal and gut microbiota composition in mice. Gut Microbes 7:313–322. uterine bacterial communities in postpartum lactating cows. doi:10.1080/19490976.2016.1203502 Front. Microbiol. 8:1047. doi:10.3389/fmicb.2017.01047 Otero, M. C., and M. E. Nader-Macías. 2006. Inhibition of Edgar, R. C. 2010. Search and clustering orders of magnitude Staphylococcus aureus by H2O2-producing Lactobacillus gasseri faster than BLAST. Bioinformatics 26:2460–2461. doi:10.1093/ isolated from the vaginal tract of cattle. Anim. Reprod. Sci. bioinformatics/btq461 96:35–46. doi:10.1016/j.anireprosci.2005.11.004 Eschenbach, D. A., P. R. Davick, B. L. Williams, S. J. Klebanoff, Otero, C., C. Silva de Ruiz, R. Ibañez, O. R. Wilde, K. Young-Smith, C. M. Critchlow, and K. K. Holmes. 1989. A. A. P. de Ruiz Holgado, and M. E. Nader-Macias. 1999. Prevalence of hydrogen peroxide-producing Lactobacillus Lactobacilli and Enterococci isolated from the bovine vagina species in normal women and women with bacterial vaginosis. during the estrous cycle. Vet. Micro. 5:305–307. doi:10.1006/ J. Clin. Microbiol. 27:251–256. anae.1999.0245 Flint, H. J., K. P. Scott, S. H. Duncan, P. Louis, and E. Forano. 2012. Perry, G. A., and B. L. Perry. 2008a. Effect of preovulatory Microbial degradation of complex carbohydrates in the gut. concentrations of estradiol and initiation of standing estrus Gut Microbes 3:289–306. doi:10.4161/gmic.19897 on uterine pH in beef cows. Domest. Anim. Endocrinol. 34:333– Földi, J., M. Kulcsár, A. Pécsi, B. Huyghe, C. de Sa, J. A. C. M. Lohuis, 338. doi:10.1016/j.domaniend.2007.09.003 P. Cox, and G. Huszenicza. 2006. Bacterial complications of Perry, G. A., and B. L. Perry. 2008b. Effects of standing estrus and postpartum uterine involution in cattle. Anim. Reprod. Sci. supplemental estradiol on changes in uterine pH during 96:265–281. doi:10.1016/j.anireprosci.2006.08.006 a fixed-time artificial insemination protocol.J. Anim. Sci. Forbes, J. D., G. Van Domselaar, and C. N. Bernstein. 2016. The 86:2928–2935. doi:10.2527/jas.2008-1181 gut microbiota in immune-mediated inflammatory diseases. Pohler, K. G., M. H. C. Pereira, F. R. Lopes, J. C. Lawrence, Front. Microbiol. 7:1081. doi:10.3389/fmicb.2016.01081 D. H. Keisler, M. F. Smith, J. L. M. Vasconcelos, and J. A. Green. Genís, S., R. L. A. Cerri, À. Bach, B. F. Silper, M. Baylão, J. Denis- 2016. Circulating concentrations of bovine pregnancy- Robichaud, and A. Arís. 2018. Pre-calving intravaginal associated glycoproteins and late embryonic mortality in administration of lactic acid bacteria reduces metritis lactating dairy herds. J. Dairy Sci. 99:1584–1594. doi:10.3168/ prevalence and regulates blood neutrophil gene expression jds.2015-10192 after calving in dairy cattle. Front. Vet. Sci. 5:135. doi:10.3389/ Potter, T. J., J. Guitian, J. Fishwick, P. J. Gordon, and I. M. Sheldon. fvets.2018.00135 2010. Risk factors for clinical endometritis in postpartum Jeon, S. J., A. Vieira-Neto, M. Gobikrushanth, R. Daetz, dairy cattle. Theriogenology 74:127–134. doi:10.1016/j. R. D. Mingoti, A. C. Parize, S. L. de Freitas, A. N. da Costa, theriogenology.2010.01.023 Ault et al. | 4313

Rajilić-Stojanović, M., and W. M. de Vos. 2014. The first 1000 Shima, T., Y. Sasaki, M. Itoh, A. Nakashima, N. Ishii, K. Sugamura, cultured species of the human gastrointestinal microbiota. and S. Saito. 2010. Regulatory T cells are necessary for FEMS Microbiol. Rev. 38:996–1047. doi:10.1111/1574-6976.12075 implantation and maintenance of early pregnancy but not Ravel, J., P. Gajer, Z. Abdo, G. M. Schneider, S. S. K. Koenig, late pregnancy in allogeneic mice. J. Reprod. Immunol. 85:121– S. L. McCulle, S. Karlebach, R. Gorle, J. Russell, C. O. Tacket, 129. doi:10.1016/j.jri.2010.02.006 et al. 2011. Vaginal microbiome of reproductive-age women. Smith, P. M., M. R. Howitt, N. Panikov, M. Michaud, C. A. Gallini, Proc. Natl. Acad. Sci. USA 108(Suppl. 1):4680. doi:10.1073/ M. Bohlooly-Y, J. N. Glickman, and W. S. Garrett. 2013. The pnas.1002611107 microbial metabolites, short-chain fatty acids, regulate Richards, M. W., J. C. Spitzer, and M. B. Warner. 1986. Effect of colonic Treg cell homeostasis. Science 341:569–573. varying levels of postpartum nutrition and body condition doi:10.1126/science.1241165 at calving on subsequent reproductive performance in beef Swartz, J. D., M. Lachman, K. Westveer, T. O'Neill, T. Geary, cattle. J. Anim. Sci. 62:300–306. doi:10.2527/jas1986.622300x R. W. Kott, J. G. Berardinelli, P. G. Hatfield, J. M. Thomson, Rönnqvist, D. J., B. Forsgren-Brusk, and E. Grahn-Hkansson. 2006. A. Roberts, et al. 2014. Characterization of the vaginal Downloaded from https://academic.oup.com/jas/article-abstract/97/10/4305/5525064 by Columbia University user on 29 April 2020 Lactobacilli in the female genital tract in relation to other microbiota of ewes and cows reveals a unique microbiota genital microbes and vaginal pH. Acta Obstet. Gynecol. Scand. with low levels of Lactobacilli and near-neutral pH. Front. Vet. 85:726–735. doi:10.1080/00016340600578357 Sci. 1:19. doi:10.3389/fvets.2014.00019 Rodney, R. M., P. Celi, W. Scott, K. Breinhild, J. E. P. Santos, Vasconcellos Cardoso, M., A. Blanchard, S. Ferris, R. Verlengia, and I. J. Lean. 2018. Effects of nutrition on the fertility of J. Timenetsky, and R. A. Florio Da Cunha. 2000. Detection of lactating dairy cattle. J. Dairy Sci. 101:5115–5133. doi:10.3168/ ureaplasma diversum in cattle using a newly developed PCR- jds.2017-14064 based detection assay. Vet. Microbiol. 72:241–250. doi:10.1016/ Ruder, C. A., R. G. Sasser, R. J. Williams, J. K. Ely, R. C. Bull, and S0378-1135(99)00203-5 J. E. Butler. 1981. Uterine infections in the postpartum cow. Vasudevan, P., M. K. Nair, T. Annamalai, and Theriogenology 15:573–580. doi:10.1016/0093-691X(81)90060-1 K. S. Venkitanarayanan. 2003. Phenotypic and genotypic Sandrini, S., M. Aldriwesh, M. Alruways, and P. Freestone. 2015. characterization of bovine mastitis isolates of Staphylococcus Microbial endocrinology: host-bacteria communication aureus for biofilm formation. Vet. Microbiol. 92:179–185. within the gut microbiome. J. Endocrinol. 225:R21–R34. doi:10.1016/S0378-1135(02)00360-7 doi:10.1530/JOE-14-0615 Weiss, S., Z. Z. Xu, S. Peddada, A. Amir, K. Bittinger, A. Gonzalez, Schloss, P. D., S. L. Westcott, T. Ryabin, J. R. Hall, M. Hartmann, C. Lozupone, J. R. Zaneveld, Y. Vázquez-Baeza, A. Birmingham, E. B. Hollister, R. A. Lesniewski, B. B. Oakley, D. H. Parks, et al. 2017. Normalization and microbial differential C. J. Robinson, et al. 2009. Introducing Mothur: open-source, abundance strategies depend upon data characteristics. platform-independent, community-supported software for Microbiome 5:27. doi:10.1186/s40168-017-0237-y describing and comparing microbial communities. Appl. Wolfenson, D., Z. Roth, and R. Meidan. 2000. Impaired Environ. Microbiol. 75:7537–7541. doi:10.1128/AEM.01541-09 reproduction in heat-stressed cattle: basic and applied Sheldon, I. M., G. S. Lewis, S. LeBlanc, and R. O. Gilbert. 2006. aspects. Anim. Reprod. Sci. 60–61:535–547. doi:10.1016/ Defining postpartum uterine disease in cattle.Theriogenology S0378-4320(00)00102-0 65:1516–1530. doi:10.1016/j.theriogenology.2005.08.021 Zhang, M., Q. Zhou, R. G. Dorfman, X. Huang, T. Fan, H. Zhang, Sheldon, I. M., S. B. Price, J. Cronin, R. O. Gilbert, and J. Zhang, and C. Yu. 2016. Butyrate inhibits interleukin-17 and J. E. Gadsby. 2009. Mechanisms of infertility associated with generates Tregs to ameliorate colorectal colitis in rats. BMC clinical and subclinical endometritis in high producing Gastroenterol. 16:84. doi:10.1186/s12876-016-0500-x dairy cattle. Reprod. Domest. Anim. 44(Suppl. 3):1–9. Zheng, H., M. Chen, Y. Li, Y. Wang, L. Wei, Z. Liao, M. Wang, F. Ma, doi:10.1111/j.1439-0531.2009.01465.x Q. Liao, and Z. Xie. 2017. Modulation of gut microbiome Sheldon, I. M., E. J. Williams, A. N. Miller, D. M. Nash, and composition and function in experimental colitis treated S. Herath. 2008. Uterine diseases in cattle after parturition. with Sulfasalazine. Front. Microbiol. 8:1703. doi:10.3389/ Vet. J. 176:115–121. doi:10.1016/j.tvjl.2007.12.031 fmicb.2017.01703