Prevalence of nasopharyngeal bacteria during naturally occurring bovine respiratory disease in commercial stocker cattle

Study population

All procedures were performed with the approval of a commercial stocker producer on privately managed calves and the University of Tennessee Institutional Animal Care and Use Committee (Protocol # 2878-1221). Calves were received from an order buyer and maintained on pasture at a commercial stocker operation in Crossville, TN. Calves were kept on 15-acre field with mixed forage containing native warm and cool season grasses. All calves were supplemented daily with a mix of crushed corn, corn gluten, and straw, as well as a pelleted calf starter (14% crude protein) (Akter et al., 2022). Calves were managed according to the farm’s standard operating procedures. At conclusion of the study, because all calves were owned privately, calves remained at the stocker facility and maintained at the discretion of the producers.

Forty clinically healthy mixed-breed beef bull and steer calves (225 ± 25 kg) were selected for the study. Prior to arrival at the stocker facility, the order buyer administered broad-spectrum anthelmintics (12 mL Safe-Guard Suspension Drench; Merck Animal Health, Rahway, NJ, USA), antibiotics (7 mL Excede subcutaneously; Zoetis Inc., Parsippany-Troy Hills, NJ, USA), mineral injection (5 mL Multimin 90 sub-cutaneously; Multimin North America, Inc., Fort Collins, CO, USA), and vaccinations against BRD and clostridial pathogens (MpB Guard subcutaneously; American Animal Health Inc., Grand Prairie, TX, USA; Ultrabac 7 subcutaneously; Zoetis Inc., Parsippany-Troy Hills, NJ, USA), and Pyramid 5 + Presponse SQ subcutaneously (Boehringer Ingelheim Animal Health USA, Duluth, GA, USA). Ear notch samples were collected from all calves to test for persistent bovine viral diarrhea virus infections, with none testing positive. Intact bull calves were castrated (n = 27) and horns were removed as needed. No unexpected adverse events occurred during the study.

Clinical observation of calves

Calves were monitored once daily for BRD clinical symptoms by the producer. A trained clinician scored each animal and assigned BRD diagnosis on sampling day 0, 7, 14 and 21 (relative to entry to the stocker facility). Clinical severity score (CSS) was recorded as CSS 0 (normal; no signs of clinical illness), CSS 1 (slightly ill; gaunt or nasal/ocular discharge), CSS 2 (moderately ill; any of the two following clinical signs: gaunt nasal/ocular discharge, coughing, labored breathing, and physically walking behind other animals in the group), CSS 3 (severely ill; purulent nasal/ocular discharge, severe labored breathing), and CSS 4 (moribund; non-responsive to human approach) (Pillen et al., 2016).

If diagnosed with BRD calves were given 40 mg/kg body weight of florfenicol (Nuflor; Merck Animal Health, Rahway, NJ, USA) subcutaneously and 2.2 mg/kg body weight of flunixin meglumine (Benamine; Merck Animal Health, Rahway, NJ, USA) intravenously. If BRD persisted on the subsequent sampling date, calves were administered subcutaneous tulathromycin (Draxxin; Zoetis Inc., Parsippany-Troy Hills, NJ, USA) with a dose of 2.5 mg/kg body weight. Calves were assigned a retroactive treatment status based on the total number of antibiotic treatments received for BRD over the entire study period (NumTrt; 0×: never treated; 1×: treated one time; 2×: treated two times). Calves were assigned to category 0× (if they received no antibiotic treatments), 1× (if they received florfenicol and flunixin meglumine), and 2× (if they received tulathromycin on the consecutive sampling).

Lung consolidation was assessed by transthoracic ultrasound (TUS) for both left and right sides using a portable ultrasound. Isopropyl alcohol (70%) was used as a transducing agent and was applied between the 3^rd and 12^th intercostal space. A TUS score of 1 (normal; aerated lung with no consolidation), score of 2 (moderate consolidation, no more than two distinct areas of consolidation of <3 cm²; diffuse comet-tail artifacts and lobular lesions) or score of 3 (severe consolidation; consolidation area ≥3 cm² and/or greater than two distinct areas of consolidation of <3 cm²) was given to each calf (Akter et al., 2022). Calves with TUS score of 1 in both left and right lungs and no treatment for BRD on the sampling day were characterized as healthy. Calves with at least one lung scoring a TUS of 2 and no treatment were considered as subclinical, while those with a TUS of at least two in any of the lungs and received treatment for BRD were categorized as clinical.

Calves were categorized as BRD infected when they had CSS 1 with rectal temperature of >40 °C, or if they had a CSS >2, regardless of rectal temperature. TUS scoring was used as a measure of lung pathology and used in conjunction with CSS to diagnose BRD.

Nasopharyngeal swab collection

Nasopharyngeal (NP) samples were collected using one swab per nare, on days 0, 7, 14, and 21. Nasopharyngeal (NP) swabs were collected using a 33-inch double guarded swab (Santa Cruz Animal Health, Dallas, TX, USA). Immediately prior to NP swab collection, external nares were wiped with an alcohol swab. The nasopharyngeal swab was inserted into the ventral meatus of the nose and rotated against the pharyngeal wall for 20–30 s. Swabs were removed and cotton tips were broken off into 5 mL conical tubes containing Brain Heart Infusion (BHI) and glycerol (1:1 of BHI and glycerol). Samples were transported on ice to the laboratory and processed on the sample day as sampling., Samples were vortexed, samples from left and right nares were pooled, and the liquid portion was stored at −80 °C for future analysis. Because swabbing was minimally invasive and causes little distress, no anesthetics were administered.

DNA extraction and sequencing

For DNA extraction, a QIAamp^® PowerFecal^® DNA kit (Qiagen, Hilden, Germany) was utilized with modifications. Samples were thawed, centrifuged at 13,000× g for 5 min, and the supernatant was discarded. The subsequent pellet was resuspended in 180 µL lysis buffer (Hart et al., 2015), and incubated for 1 h (37 °C). After which, 25 µL proteinase K and buffer AL were added, followed by incubation for 30 min (56 °C). The supernatant was transferred to lysis tubes (Zymo Research Corp, Tustin, CA, USA), and cell lysis was performed utilizing a TissueLyser II system (Qiagen, Hilden, Germany) for 3 min at 21 Hz. After which, 200 µL of 100% ethanol was added to the lysate, mixed by vortexing, and then transferred to an DNeasy spin column placed within a 2 mL collection tube (Qiagen, Hilden, Germany). The lysate was centrifuged at 6,000× g for 1 min (37 °C), and flow-through was discarded. After replacing the collection tube, 500 µL of buffer AW1 was added, and centrifuged for 1 min at 6,000× g for 1 min. After discarding flow-through and replacing the collection tube, 500 µL of buffer AW2 was added and centrifuged at 20,000× g for 3 min to dry the DNeasy membrane. The DNA was eluted by the addition of 200 µL of buffer AW directly to the DNeasy membrane, followed by centrifugation at 6,000× g for 1 min. DNA sample concentration and purity was measured using a spectrophotometer (DeNovix, Wilmington, DE, USA).

Extracted DNA was sent to the University of Tennessee Genomics Core laboratory facility for library preparation and amplicon sequencing. Library preparation and amplicon sequencing targeted the V4 hypervariable region of the 16S rRNA gene with modified adapters used for Illumina MiSeq sequencing (Henniger et al., 2022). Broad-spectrum forward (515Fb; Parada, Needham & Fuhrman, 2016) and reverse (806Rb; Apprill et al., 2015) primers were used and sequencing was conducted in a two-step polymerase chain reaction (PCR). Conditions of PCR were performed using the following thermal cycler profile: 10-min hold at 95 °C, 35 cycles of 95 °C for 30 s, 55 °C for 30 s, and 75 °C for 30 s, followed by 72 °C for 10 min. Purification of the PCR product was completed using 20 µL of AMPure XP beads (Beckman Coulter, Breca, CA, USA), with indexing of each sample utilizing a unique combination of forward and reverse Nextera XT v2 indexes (Illumina, San Diego, CA, USA). Library sequencing was completed using Illumina MiSeq (2 × 300 bp) and the MiSeq Reagent Kit v3 (600-cycle; Illumina, San Diego, CA, USA).

Analysis of gene sequences

DADA2 v1.14 (Callahan et al., 2016) was used to process raw sequences of the 16S rRNA gene in R v1.4.1717 (R Development Core Team, 2010). Unless stated otherwise, DADA2 was used to process and analyze the gene sequences. Low quality sequences were identified, trimming was performed using Phred quality scores (trimmed if <30), and 250 and 200 base lengths were truncated in the forward and reverse reads, respectively. Error rate was determined using approximately 106 million forward and reverse reads. Forward and reverse reads were merged, and chimeric sequences were removed. Using a 97% similarity threshold value, all high-quality sequences were assigned to a specific operational taxonomic unit (OTU). Each OTU was analyzed against the SILVA 138.1 SSU taxonomic database (Quast et al., 2013) by naïve Bayesian RDP classifier method (Wang et al., 2007). Shannon diversity and Chao1 diversity index were measured to identify alpha diversity using Phyloseq v1.30.0 (McMurdie & Holmes, 2013) and vegan v2.5-7 (Oksanen et al., 2013) packages in R (R Development Core Team, 2010). Beta diversity analyses were performed using the Bray-Curtis dissimilarity in vegan. Principal coordinate analysis (PCoA) utilized the Bray-Curtis dissimilarity to determine beta diversity based on the TUS status at different sampling days.

Statistical analyses

The GLIMMIX procedure in SAS 9.4 (SAS Institute Inc., Cary, NC, USA) was used to test differences in relative abundances occurring due to NumTrt and day. Temporal variation in alpha diversity matrices or relative abundances of the top phyla or genera within cattle never receiving treatment was measured by using the mixed model analysis of variance (ANOVA), where the fixed effect of sampling day and the random effect of individual calf ID with the random residual of the day were included. Additionally, variation in the alpha diversity on treatment days 14 and 21 were measured using the GLIMMIX procedure in SAS 9.4. The effect of Numtrt on alpha diversity matrices or relative abundances of top phyla or genera were analyzed using the mixed model ANOVA with the fixed effect of NumTrt, and the random effect of individual calf ID, with the random residual of the day. In a separate model, sampling day was considered the fixed effect, whereas individual calf ID was the random effect, with the random residual of the day included as an autoregressive repeated term. To measure differences in the bacterial community based on TUS status, permutational multivariate analysis of variance (PERMANOVA) was computed using the vegan package in R with 999 permutations (R Development Core Team, 2010). Differences in alpha diversity matrices or relative abundances of top phyla or genera between calves that were either healthy or received treatment on a given day were also evaluated. The fixed effect of treatment by day was evaluated by ANOVA. Since these analyses were subset by day, no random effects were required. For all the analyses, differences were considered significant when P ≤ 0.05.

Differences in alpha and beta diversity measures

Although bacterial composition differed based on TUS status, a low R² value indicates poor clustering (Fig. 4). Among the subset of clinically healthy calves, no differences in temporal variation from the Chao1 metric for the OTUs were observed (P > 0.05) (Table S1). Likewise, there was no temporal variation in the Shannon diversity index in the subset of healthy calves observed, except at day 14 (P < 0.05). Neither alpha diversity measures (Chao1 and Shannon diversity) differed based on treatment status at days 14 and 21 (P > 0.05). Based on beta diversity analyses, the composition of bacteria differed based on TUS staus (P = 0.001; R² = 0.04).

Temporal changes of bacterial phyla and genera among healthy calves

Temporal differences in the relative abundance were observed in phyla Firmicutes (P = 0.03), Actinobacteriota (P < 0.0001), Bacteroidota (P = 0.001), and Verrucomicrobiota (P = 0.001) in healthy calves (Table S2). No differences were observed in the relative abundance for Proteobacteria (P = 0.13). Significant differences were observed in the relative abundance based on day for Mycoplasma (P = 0.01); however, no differences were observed for Histophilus, Pasteurella, Bacillus, and Lactobacillus (P > 0.05) in healthy calves (Table S3). In the clinically healthy calves, Histophilus was not identified within 1% abundance at arrival; however, on day 7, one calf had a population of Histophilus (Table S3). Based on the antibiotic treatment status, calves who received no treatment had a higher abundance of Mycoplasma (31.9%), Histophilus (10.56%), Geobacillus (12.15%), Bacillus (10.02%), Saccharococccus (7.27%), Lactobacillus (4.64%), Clostridium sensu stricto 7 (3.85%), Moraxella (3.57%), and Pasteurella (3.04%) (Table S3).

Relative abundance of bacterial phyla and genera based on treatment status

The effect of the number of antibiotic treatments given (NumTrt), day, and the two-way interaction on relative abundance for top bacterial phyla highlighted differences in Actinobacteriota (P = 0.02) based on NumTrt (Table S4). Calves with BRD receiving 1× antibiotic treatment had greater abundance of Actinobacteriota. Although calves receiving no treatment expressed numerically higher Actinobacteriota abundance, they were not statistically different from the calves receiving 1× or 2× antibiotic treatments (Table S4). Based on NumTrt, relative abundance of Firmicutes, Proteobacteria, Bacteroidota, and Verrucomicrobiota did not differ (P > 0.05) (Table S4). A day effect occurred with relative abundance of Proteobacteria (P = 0.03), Actinobacteriota (P < 0.0001), Bacteroidota (P = 0.002), and Verrucomicrobiota (P = 0.0001) (Table S4). At day 7, clinically healthy calves had significantly reduced Mycoplasma than 1× and 2× antibiotic treated calves; however, no differences were observed among NumTrt status for other days (Fig. 4). Effect of NumTrt were observed for Pasteurella (P = 0.04) and Bacillus (P = 0.03) (Table S5). Based on NumTrt, Pasteurella were significantly higher in the clinically healthy calves, whereas the 2× antibiotic treated group had significantly lower Bacillus (Table S5). Significant effect of day was observed for Histophilus (P < 0.0001), Lactobacillus (P = 0.001), and Bacillus (P = 0.007) (Table S5).