Feeding Saccharomyces cerevisiae fermentation postbiotic products alters immune function and the lung transcriptome of preweaned calves with an experimental viral-bacterial coinfection

Bovine respiratory disease (BRD) causes morbidity and mortality in cattle of all ages. Supplementing with postbiotic products from Saccharomyces cerevisiae fermentation (SCFP) has been reported to improve growth and provide metabolic support required for immune activation in calves. The objective of this study was to determine effects of SCFP supplementation on the transcriptional response to coinfection with bovine respiratory syncytial virus (BRSV) and Pasteurella multocida ( PM ) in the lung using RNA seq. Twenty-three calves were enrolled and assigned to 2 treatment groups: con-trol (CON, n = 12) or SCFP-treated (SCFP, n = 11, fed 1 g/d SmartCare® in milk and 5 g/d NutriTek on starter grain). Calves were infected with ~10 4 TCID50 BRSV, followed 6 d later by intratracheal inoculation with ~10 10 cfu of PM (strain P1062). Calves were euthanized on d 10 post viral infection. Blood cells were collected and assayed on d 0 and 10 after viral infection. Bronchoalveolar lavage (BAL) cells were collected and assayed on d 14 of the feeding period (preinfection) and d 10 after viral infection. Blood and BAL cells were assayed for proinflammatory cytokine production in response to stimulation with lipopolysaccharide (LPS) or a combination of Poly(I:C) and Imiquimod, and BAL cells were evaluated for phagocytic and reactive oxygen species (ROS) production capacity. Antemortem and postmortem BAL and lesion (LL) and non-lesioned (NLL) lung tissue samples collected at necropsy were subjected to RNA extraction and sequencing. Sequencing reads were aligned to the bovine reference genome (UMD3.1) and EdgeR used for differential gene expression (DEGs) analysis. Supplementation with SCFP did not affect


INTRODUCTION
Bovine respiratory disease (BRD) complex is an animal welfare concern and results in massive global economic losses for producers due to morbidity, mortality and decreased production in the feedlot and dairy industries (Blakebrough-Hall et al., 2020;Overton, 2020;Buczinski et al., 2021).Specifically, BRD is currently the second leading cause of mortality in preweaned heifers and the leading cause of mortality in weaned heifers (Overton, 2020).Bovine respiratory disease has a complex etiology that involves multiple pathogens and is often complicated by environmental and management factors such as heat, sanitation, diet, and comingling.Despite the use of preventive measures, including vaccination for multiple respiratory pathogens, Feeding Saccharomyces cerevisiae fermentation postbiotic products alters immune function and the lung transcriptome of preweaned calves with an experimental viral-bacterial coinfection the prevalence of the disease remains high.Treatment or metaphylaxis with antimicrobial drugs is widely used and often the most effective strategy.However, with growing public and human health concerns over use of medically important antibiotics in livestock and the increasing prevalence of antimicrobial resistance in BRD pathogens, there is an urgent need for alternative approaches to treat and prevent the disease.To this end, interest in the development of natural immunomodulators has increased dramatically in the livestock industry.
Saccharomyces cerevisiae fermentation products (SCFP) are postbiotic feed additives comprised of functional metabolites and bioactive compounds that are produced through a controlled fermentation process (Deters et al., 2018).Supplementation with SCFP products has been shown to improve the performance of calves and adult cattle in multiple production settings, and to have beneficial effects on immune function during times of stress and disease (Brewer et al., 2014;Li et al., 2016;Alugongo et al., 2017;Knoblock et al., 2019;Vélez et al., 2019;Vailati-Riboni et al., 2021;Guo et al., 2022;Klopp et al., 2022).For example, in dairy cows, feeding SCFP reduced serum haptoglobin during the periparturient period, thus limiting excessive immune system activation after calving (Knoblock et al., 2019).Cows experiencing a subacute ruminal acidosis (SARA) challenge had lower concentrations of serum acute phase proteins and less ruminal LPS when receiving SCFP treatments compared with controls, suggesting the postbiotic supplement stabilized the rumen microbiota and reduced inflammation associated with SARA events (Li et al., 2016;Guo et al., 2022).Supplementation with SCFP also improved the outcome of a Salmonella enterica serotype typhimurium challenge in preweaned calves (Brewer et al., 2014;Harris et al., 2017), and Streptococcus uberis in dairy cows (Vailati-Riboni et al., 2021) by reducing clinical disease including rectal temperatures and somatic cell count in the latter.In a recent study with Holstein bull calves, feeding SCFP reduced the incidence of BRD post weaning and the number of antibiotic treatment events, while improving body weight and average daily gain post weaning (Klopp et al., 2022).
We have previously reported that neonatal calves supplemented with SCFP developed less severe clinical disease and lung pathology compared with control calves following an experimental bovine respiratory syncytial virus (BRSV) infection (Mahmoud et al., 2020).We observed that circulating immune cells from the blood of SCFP-treated calves had an increased capacity to secrete the proinflammatory cytokines tumor necrosis factor α (TNF-α), interleukin-1 β (IL-1β) and interleukin-6 (IL-6) in response to toll like receptor (TLR) stimulation.However, we observed an opposing effect on immune populations isolated from the lung, as cells collected from the bronchoalveolar lavage (BAL) of SCFP treated calves produced less proinflammatory cytokines in response to TLR stimulation (Mahmoud et al., 2020).Thus, the immunomodulatory effects of SCFP supplementation on both peripheral blood and lung responses may have a beneficial effect on the outcome of BRD.In our recent follow up study using a viral-bacterial coinfection model with BRSV and Pasteurella multocida (PM), we again observed that neonatal calves receiving SCFP supplementation experienced a milder disease (McDonald et al., 2021).Calves receiving SCFP had lower thoracic ultrasonography (TUS) scores, indicating less lung consolidation; less neutrophil infiltration into the airways compared with controls and tended to have less lung consolidation at necropsy.Further, although there were no differences in viral or bacterial burden in the lungs, the SCFP calves experienced metabolic changes that correlated with improved growth and development (McDonald et al., 2021).The objectives of the current study were to further explore the mode of action by which SCFP supplementation may impact immunity and disease resistance in the respiratory tract.To that end, we evaluated innate immune functions, including phagocytosis, reactive oxygen species (ROS) and proinflammatory cytokine production by immune cells from the blood and lungs of control and SCFP treated calves experiencing an experimental BRSV/PM infection.We also analyzed the transcriptome of airway immune cells collected in the BAL before and after infection, as well as lesioned (LL) and non-lesioned lung (NLL) tissue samples collected on d 10 after experimental BRSV/ PM infection (McDonald et al., 2021).We hypothesized that SCFP supplementation modulated lung mucosal immune responses leading to controlled inflammation and increased resistance to BRD.

Animal study design and sample collection
This study was approved by the Iowa State University Institutional Animal Care and Use Committee (protocol 20-133) and the Institutional Biosafety .Details of the experimental design, infection and clinical disease parameters have been previously published (McDonald et al., 2021).The study was performed as a randomized complete block design consisting of 1 32-d period.A power analysis was performed to estimate sample size.The effect size of SCFP supplementation was considered medium (d = 0.55).With an α = 0.05, and power = 0.7, the pro-jected samples size to measure a significant effect of SCFP supplementation on gross lung pathology scores (in challenge studies, Mann-Whitney test) was n = 12.To account for animal losses, n = 14 calves were enrolled per group.One calf failed the health assessment at enrollment.Therefore, 27 Holstein x Angus cross mixed sex calves (1 to 2-d old) were blocked by body weight and randomly assigned to a treatment group: 1) control (CON): base milk replacer and calf starter or 2) SCFP supplemented (SCFP): base milk replacer supplemented with 1 g/d SmartCare and calf starter supplemented with 5 g/d NutriTek.Feeding protocols followed the outline indicated in (Mahmoud et al., 2020;McDonald et al., 2021).One SCFP calf was removed from the study during the feeding period.Calves were divided into groups and challenged with ~10 4 Tissue Culture Infectious Dose (TCID) 50 /mL of BRSV strain 375 via aerosol inoculation on d 20, 21 or 22 of the trial.The virus inoculum was suspended in 5mL and delivered by forced-air nebulizer to a mask covering the nose and mouth of the calf (EquiResp, Blanchard, OK).The inoculation procedure takes approximately 10 min per animal.On d 4 post viral infection, calves were infected via intratracheal inoculation with ~10 10 colony forming units of PM strain P1062 type A:3 in 60 mL of sterile saline (McDonald et al., 2021).Peripheral blood was collected on d 6, 7, and 10 post viral infection for leukocyte respiratory burst assays (described below).Peripheral blood mononuclear cells (PBMC) were also isolated and cryopreserved for future analysis as previously described (Mahmoud et al., 2020).
At the end of the study, on d 10 after BRSV infection, animals were humanely euthanized by intravenous administration of barbiturates.The lungs were collected and observed for pathological differences before collecting post-mortem BAL sample and lesioned and non-lesioned lung tissue samples.Lung samples were used for viral quantification, gene expression, bacterial recovery, and sequencing analysis, while BAL samples were used for immunology assays, gene expression and sequencing analysis.Postmortem BAL fluid was collected as previously described (Mahmoud et al., 2020;McDonald et al., 2021) and processed in the same manner as antemortem fluid.Pathogen burden, pathological evaluations, and tissue gene expression data can be found in the complement study's publication (McDonald et al., 2021).

Leukocyte respiratory burst and phagocytosis assays
A portion of fresh, whole blood from all sample time points were loaded with 0.1 µM dihydroergotamine (DHR)-123 (20 min in a 37°C water bath) to monitor leucocyte respiratory burst activity via DHR cleavage.Loaded samples were incubated for 20 min in a 37°C water-bath and stimulated with 0.2 µM Phorbol-12-myristate-13-acetate (PMA) (Sigma-Aldrich, St. Louis, MO, USA) as in (Mahmoud et al., 2020).Leukocyte phagocytic activity in the whole blood samples was determined using the commercial Phagocytosis Assay Kit (pHrodo Green AM, ThermoFisher Scientific, Carlsbad, CA, USA) according to the manufacturer's instruction and previously published modifications (Mahmoud et al., 2020).Fresh, antemortem and postmortem BAL samples were processed using the same protocols.After stopping the reactions by placing the samples on ice, the samples were surface stained with primary mouse anti-bovine CD11c (clone BAQ153A) and CD14 (clone CAM36A) to identify macrophages, and mouse antibovine granulocytes (clone CH138A).Cells were then fixed with BD FACS lysis buffer, washed and resuspended in FACS buffer for flow cytometry analysis as previously described (Mahmoud et al., 2020).Samples were analyzed using a BD FACS Canto flow cytometer and analyzed using Flowjo software (version 10.8.1).

Thawing and in vitro cell stimulation
To evaluate the innate cellular response, PBMC and BAL cells were removed from the liquid nitrogen and thawed.Cells were washed using 9 mL of warm cRPMI.After confirming viability, PBMC and BAL cells (5 × 10 5 cells/mL, 100 µL/well) were added to a 96-well round-bottom plate.Cell cultures were stimulated using cRPMI (negative control), 1 µg/mL LPS, or a mixture of 50 µg/mL Poly(I:C) with 10 µg/mL imiquimod (all from Sigma-Aldrich, St. Louis, MO, USA).Cultures were incubated at 37°C, 5% CO 2 for 4 h (TNF-α) or 48 h .Supernatants for all stimulations

Enzyme-linked immunosorbent assays
Commercial bovine ELISA kits for IL-1β, IL-6 (Ther-moFisher Scientific, Carlsbad, CA, USA) and TNF-α (R&D Systems, Minneapolis, USA) were performed according to manufacturer's instructions.Samples were measured in duplicate and read using a Thermo Fisher Multiskan FC Microplate Photometer.

Sample collection and RNA extraction for RNA seq library preparation and sequencing
Approximately 0.5 g of tissue were collected from 2 to 3 locations of affected (lesioned, LL) and unaffected (non-lesioned, NLL) lung at necropsy on d 10 post BRSV infection from CON (n = 12) and SCFP (n = 11) and stored in RNAlater (Invitrogen, Carlsbad, CA, USA).BAL was collected antemortem on d 14 of the feeding period and postmortem at necropsy from control (n = 12) and SCFP supplemented calves (n = 11).BAL samples were stored in RNAlater (Invitrogen, Carlsbad, CA, USA) to preserve integrity and expression profiles of the samples.
Total RNA was extracted from lung tissues using Trizol Reagent (Invitrogen, Carlsbad, CA, USA), followed by clean-up and DNase digestion on a Qiagen RNeasy column (Qiagen, Hilden, Germany) as in (McGill et al., 2019).RNA extractions were performed using MagMAX mirVana Total RNA Isolation Kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.Quality control of the BAL and lung RNA samples was evaluated using the RNA 6000 Nano LabChip kit (Agilent Technologies Ireland Ltd.,Dublin, Ireland) with Agilent 2100 Bioanalyzer system (Agilent, Ireland Ltd., Dublin, Ireland) to confirm size, concentration, and integrity.All samples had an RNA integrity number (RIN) number above 8.RNA was transported frozen at −80°C on dry ice to Iowa State University Sequencing Core Facility for RNA seq library preparation using the NEBNext® Ultra II RNA Library Prep Kit-Illumina (Illumina, San Diego, CA, USA) and high-throughput sequencing (100 bp pairedend) on a single S1 flow cell on Illumina NovaSeq 6000 (Illumina, San Diego, CA, USA).

Reads Alignment and Differential Expression Analysis of RNA Seq
Two sequence reads (FASTQ format) were provided by Iowa State University Sequencing Core Facility, one for each sequencing lane.Counts were merged be-tween lanes by averaging counts between the 2 files.Sequence reads were analyzed for quality metrics using FastQC (version 0.11.7) and MultiQC (version 1.5) and all reads passed the basic quality statistics (i.e., read length, sequence length distribution, per base sequency quality, overrpresented sequences, etc.).The sequence reads were aligned to the bovine reference genome (UMD3.1)using HISAT2 program.Differential gene expression was performed with the program edgeR version 3.32.1 (Dai et al., 2014) using a general linearized model with control and treatment as factors.The Benjamini and Hochberg correction was used to estimate the false discovery rate (FDR).Sequences were normalized using the reads per kilobase per million mapped read (RPKM) method.Hierarchical clustering analysis was performed using the program MClust version 5.4.9 (Scrucca et al., 2016) for genes that were found to be significantly different (FDR <0.05) using the average of normalized counts across samples with the same treatment.Functional enrichment analysis was performed with PANTHER Database version 11.1 released on 2016-10-24 (http: / / pantherdb .org/ ) classification system utilizing Bos taurus as the reference organism using the Benjamini-Hochberg False Discovery Rate for multiple test correction (Mi et al., 2017).

Validation of transcriptome sequencing using real time quantitative PCR
Eleven DEGs were selected from the sequencing data and verified using RT-qPCR.The genes are AIRE, CCL1, CCR8, CD8α, CD79β, FCRLA, ICOS, IFNγ, MZB1, TIGIT, and XCR1.Primers used for these genes can be found at (McGill (2023) https: / / figshare .com/articles/ dataset/ _/ 24036855).Primers were designed using NCBI PrimerBlast software.The original anteand postmortem BAL RNA preparations were used for cDNA synthesis as previously described (McGill et al., 2018;McDonald et al., 2021).Reactions were conducted in duplicate and performed using a QuantStudio 5 qPCR Machine with the following cycling conditions: 2 min at 50°C, 10 min at 95°C, 40 cycles of 15 s at 95°C and 1 min at 60°C, and a dissociation step (15 s at 95°C, 1 min at 60°C, 15 s at 95°C, 15 s at 60°C).Relative gene expression was calculated using the 2 -ΔΔT method, using RPS9 as the housekeeping gene.The fold change was calculated following normalization.

Statistics
The larger experiment was analyzed as a randomized complete block design.Individual calves were considered the experimental unit.For data in which assessments were performed on multiple days, results were analyzed by a linear mixed-effects model, fit using Restricted Maximum Likelihood with the Geisser-Greenhouse correction, followed by Sidak's test for multiple comparisons.The model used the fixed effects of time, treatment (CON vs. SCFP), and treatment × time interaction.Calf was considered a random effect.Data are reported as least squares means ± SEM.Differences of P ≤ 0.05 were considered significant.Graphs and statistical analyses were performed using GraphPad Prism v9.0.2 (GraphPad Software, Inc.).

Effects of SCFP supplementation on innate immune function in neonatal calves
Respiratory burst activity and phagocytosis are key effector functions of innate immune cells; therefore, we first evaluated the impact of SCFP supplementation on these processes in cells from both blood and lungs.We observed no effects of SCFP supplementation on respiratory burst activity by peripheral blood neutrophils (P = 0.63) or monocytes (P = 0.74) before infection or on d 10 after BRSV infection (neutrophils, P = 0.89; monocytes, P = 0.98) BAL samples were also collected and analyzed for respiratory burst activity before infection ( d 14 of the feeding period) and on d 10 after BRSV infection.We observed no treatment effects (P > 0.05) on the respiratory burst activity of either lung neutrophils or macrophages.We observed no differences in phagocytic activity by neutrophils or macrophages either before (P = 0.83, P = 0.22, respectively) or after (P = 0.75, P = 0.99 respectively) viral challenge.
We have previously observed that SCFP supplementation impacts the capacity of circulating and lung immune cells to produce inflammatory cytokines in response to TLR stimulation (Mahmoud et al., 2020).To corroborate our previous findings, PBMC were isolated on d 0, immediately before BRSV infection.Cells from the BAL were isolated before infection, on d 14 of the feeding period, or at necropsy on d 10 post infection.Both PBMC and BAL cells were stimulated with either LPS or a combination of Poly(I:C) and Imiquimod, then cell supernatants were evaluated by ELISA for IL-1β, IL-6 secretion after 48 h or TNF-α secretion after 4 h.In response to LPS stimulation, we observed no treatment effect on IL-1β or IL-6 secretion by PBMC from SCFP calves compared with cells from CON calves immediately before BRSV infection (P = 0.09, P = 0.69 respectively).We also observed no treatment effects for IL-1β or IL-6 responses to Poly(I:C)/ imiquimod stimulation by PBMC (P = 0.26 and P = 0.33, respectively).(Table 1).
In contrast to circulating cells, BAL cells from SCFP secreted less TLR-induced inflammatory cytokines compared with CON.BAL cells isolated from SCFP calves before BRSV infection produced less IL-6 in response to Poly(I:C)/imiquimod stimulation (P = 0.023), although we did not observe treatment effects on Poly(I:C)/imiquimod-induced IL-6 secretion after infection (P = 0.38).We observed no treatment effects for LPS-induced cytokine secretion before BRSV infection.However, after infection, BAL cells from SCFP calves produced less TNF-α in response to both Poly(I:C)/imiquimod (P = 0.008) and LPS (P = 0.06) compared with cells from control calves.Other cytokine responses were not different (P > 0.05) between CON and SCFP (Table 1).

Differential gene expression and functional annotation in lung tissues from SCFP and control calves with a viral-bacterial coinfection
Given the beneficial effects of SCFP supplementation on BRD (Mahmoud et al., 2020;McDonald et al., 2021;Klopp et al., 2022), and the observed effects on innate immune function in the lung, we next chose to investigate how SCFP supplementation altered local gene transcriptome profiles, with the objective of identifying novel pathways and mechanisms of SCFP treatment effects.We conducted RNA seq analysis on NLL and LL tissue samples collected at necropsy from CON (n = 12) and SCFP (n = 11) calves (Figure 1A).In the NLL, a total of 376 genes were identified as differentially expressed between control and SCFP treated calves (false discovery rate [FDR] of <0.05).In LL tissue, 270 differentially expressed genes (DEGs) were identified.As seen in the Venn diagram in Figure 1B, 73 of these genes were shared between the 2 tissue types (Figure 1B, https: / / figshare .com/articles/ dataset/ _/ 24036852).A heat map was generated for 537 DEGs identified in both NLL and LL samples and organized by supervised hierarchical clustering using the Log 2-fold change between SCFP and CON, which was calculated from the average of normalized counts across samples with the same treatment (Figure 1C).Four clusters were identified and labeled as 1, 2, 3 and 4 (Figure 1C) and functional enrichments for Gene Ontology (GO) terms for biological, cellular, and molecular processes were assessed for each cluster (Table 2).In each cluster a selection of the top enriched pathways among the predicted target genes are shown.In Cluster 1, LL samples from SCFP had distinctly higher expression of genes enriched in "plasminogen activating cascade" (FDR 8.49e-05) and "blood coagulation" including F9, F10, FGA, FGG, FGB, PLG, SERPINA1, and SERPINF2.In cluster 2, NLL samples from SCFP calves had lower expression of Maina et al.: Postbiotics impact lung transcriptome during BRD genes enriched in "microtubule-based movement" (FDR 6.72e-11) such as CFAP65, CCDC180/181/113, IQCG-IQ, SPAG6.Cluster 2 samples also had decreased expression of genes involved in immune responses, such as CISH, CFI, IL17REL, CSF2, CX3CL1, IL7R, IL1A, and lower expression of GO terms for "humoral immune responses" (FDR 3.13e-02) and "inflammatory response" (FDR 3.43e-02) in SCFP supplemented calves.In Cluster 3 both LL and NLL samples from SCFP had lower expression of GO terms such as "leukocyte migration involved in inflammatory responses" (FDR 1.18E-04), which includes genes such as CCL8, DEFB10, TNFAIP6, ISG20, CXCL2, CCL24, BCL2A1, TNFRSF6B, CXCL5, CXCL8.Genes involved in "granulocyte chemotaxis" (FDR 1.42E-09), "regulation of response to stimulus" (FDR 7.50E-06) and "B cell receptor signaling pathway" (FDR 3.39E-04) were also lower in both LL and NLL tissue from SCFP supplemented calves, but the effect was more pronounced in the NLL samples.Cluster 4 was the least congruous cluster.To facilitate the interpretation of cluster 4, this subset of genes was re-clustered, and a heat map was generated, and each cluster assessed for functional enrichment (Figure 1C).The LL samples had lower expression of genes involved in "extracellular matrix" (FDR 2.87E-02) in Cluster 4.1 and higher expression of genes involved in "interleukin-27-mediated signaling pathway" (FDR 2.62E-02) and "negative regulation of viral genome replication" (FDR 4.29E-05) in Cluster 4.2 in SCFP calves compared with control.Also, in cluster 4.3 NLL samples experience higher expression of genes enriched in "ABC-type transporter activity (FDR 2.61E-05)" in SCFP compared with CON.In cluster 4.4 both LL and NLL samples from SCFP calves had lower expression of genes enriched in "negative regulation of single-species biofilm formation in or on host organism processes" (FDR 3.94E-03) and "defense response to bacterium" (FDR 2.57E-03) compared with control (Figure 1D).Overall, non-lesioned lung samples showed lower expression of genes involved in inflammation and immune related processes while lesion lung samples showed higher expression of plasminogen and blood coagulation in SCFP compared with CON lungs.

Differential gene expression and functional annotation in BAL samples from SCFP and control calves with a viral-bacterial coinfection
To explore the impact of SCFP supplementation on disease outcome of a viral-bacterial coinfection we conducted RNA seq analysis of BAL collected antemortem (before infection) and postmortem (d 10 post viral infection) from the CON and SCFP supplemented calves.We performed a linear model analysis accounting for the baseline transcriptional state before infection to identify genes that showed differential expression when responding to infection due to SCFP supplementation.We identified 254 genes (FDR <0.05) that were dif-  ferentially expressed between control and SCFP treated animals.Six clusters were identified and labeled as 1, 2, 3, 4, 5 and 6 (Figure 2B) and functional enrichment for Gene Ontology (GO) terms for biological, cellular, and molecular processes were assessed for each cluster (Table 3).The lists showing the genes expressed are provided at https: / / figshare .com/articles/ dataset/ _/ 24036852 .Clustering analysis of Cluster 1 and 2 showed that after infection SCFP treated calves had lower transcriptional signatures of processes involved in "glutathione catabolic process"(FDR 2.95E-02), "leukocyte chemotaxis involved in inflammatory response," (FDR 1.39E-03), "interleukin-2-mediated signaling pathway" (FDR 3.83E-02), "B cell receptor signaling pathway" (FDR 1.23E-10) and "phagocytosis, recognition" (FDR 5.31E-06) compared with CON calves.These pathways involved genes such as IL2RB, IL21R, CD4, IL1R2, CD2, ICOS, CD8A, IFNγ, LAMP3, ITK, CCL20 and CTLA4.Similarly, we observed high expression of genes involved in "leukocyte chemotaxis and inflammatory response" (FDR 3.96E-02) and "cytokine receptor binding" (FDR 4.78E-02) (cluster 3 and 4, respectively) including PRF1, GZMA, C3AR1, IL12B, CCRL2, MZB1, CXCL2, LTB, TNIP3, CFB, and DEFB10, after infection in the BAL samples.Interestingly, Cluster 5 had opposite response to infection between control and SCFP.We observed higher expression of genes such as CFAP74, CROCC2, CCDC, which are processes involved in "microtubule-based movement" (FDR 1.26E-11) and "epithelial cilium movement involved in extracellular fluid movement" (FDR 1.81E-03) in CON calves compared with SCFP calves.There was no statistical enrichment of GO terms found for Cluster 6.Overall, calves treated with SCFP had lower transcriptional responses to the coinfection compared with control calves (Clusters 1-4), and different transcriptional patterns when responding to infection between control and treatment samples in Cluster 5 (Figure 2B.Cluster 5, red for lower transcription).

Differential gene expression and functional annotation in BAL samples from SCFP and control calves before experimental coinfection
Both lung tissues (Figure 1) and BAL samples (Figure 2) from calves supplemented with SCFP have a transcriptional response that indicates that animals experienced a less severe bacterial-viral challenge.To provide additional insights on the SCFP product's mode of action we explored transcriptional changes of BAL samples before infection and a volcano plot of the identified 678 DEGs was generated (Figure 3B) and the target genes annotated (https: / / figshare .com/articles/ dataset/ _/ 24036852).Most of these changes involved an increase in transcription (positive log 2-Fold Change) when calves were supplemented with SCFP (Figure 3B).Functional enrichment of the upregulated DEGs such as CD19, CD79, FCRLA, MZB1, ICOS, CD2, CD8a, CTL4, CCL20, MHC II, IL12RB1, IRF4, IL17A, CCDC13, TNF superfamily member and C-X-C motif chemokine receptor (https: / / figshare .com/articles/ dataset/ _/ 24036852), were associated with 8 biological pathways corresponding to processes related to immune processes that included regulation of "T cell differentiation," and "immune system development" among others (Figure 3C).To validate these results 11 DEGs (AIRE, CCL1, CCR8, CD8α, CD79β, FCRLA, ICOS, IFNγ, MZB1, TIGIT, and XCR1) from preinfection and necropsy BAL samples were selected and tested by qRT-PCR to test reproducibility and repeatability of the differentially expressed genes in RNA seq data.RPS9 gene was used as the housekeeping gene and all primers used in this study are listed at https: / / figshare .com/articles/ dataset/ _/ 24036855.All 11 DEGs identified by RNA sequencing followed the same expression pattern when confirmed by qRT-PCR method.

DISCUSSION
The use of immunomodulatory supplements such as SCFP is one practice that may be useful to manage the risk of BRD in cattle populations.We have previously reported that neonatal calves supplemented with SCFP develop less severe clinical disease and lung pathology compared with control calves following experimental BRSV infection (Mahmoud et al., 2020) and following an experimental viral-bacterial coinfection (McDonald et al., 2021).The current study focused on the mechanisms by which SCFP supplementation modifies immune function in calves with BRD.We observed less pro-inflammatory cytokine production by immune cells sampled from the airways of SCFP supplemented calves, which corroborated our previous findings (Mahmoud et al., 2020).This result suggests that SCFP treatment may have a positive downstream effect on the inflammatory milieu in the lungs, thus resulting in the reductions in lung pathology that we have consistently observed in calves with experimental respiratory infection.We went on to evaluate the local lung transcriptional responses associated with SCFP supplementation in preweaned calves challenged with respiratory pathogens.We have proposed a model summarizing our findings from the transcriptional analyses and suggesting ways in which SCFP supplementation may be impacting the host immune system in the lungs (Figure 4).We observed that SCFP supplementation altered the expression of many genes involved in immunity and lung repair.Specifically, in response to infection from the LL tissue samples we identified genes involved in blood coagulation, plasminogen activating cascade and serine proteases inhibitors, such as F9, F10, FGA, FGG, FGB, PLG, SERPINA1, that were differentially increased in SCFP calves.The plasminogen activating system is an enzymatic cascade that triggers the degradation of fibrin by catalyzing the conversion of plasminogen into plasmin via 2 serine proteinases.This pathway is involved in many pathological processes including wound repair and infection (Berri et al., 2013).A recent study with hospitalized COVID-19 patients demonstrated that aerosol treatment with plasminogen was effective in reducing the severity of COVID-associated lung lesions and improving hypoxemia, suggesting that plasminogen may play a beneficial role in lung repair (Wu et al., 2020).In rodents, blocking activation of the plasminogen activating system promotes lung fibrosis during inflammation, while increased activation of the plasminogen activating system protects the lung from fibrin deposition and promotes productive lung repair (Eitzman et al., 1996;Hattori et al., 2000).The SERPINA1 gene encodes for α-1 antitrypsin, a serine protease inhibitor that protects the lungs from neutrophil elastase, which can damage lung tissue if not properly controlled (Kelly-Robinson et al., 2021).Neu-trophils are a primary component of the inflammatory response to BRD, and their presence correlates with increased lung pathology (Slocombe et al., 1985;Radi et al., 2002;Agnes et al., 2013).Increased SERPINA1 expression in the lesioned lung tissue of calves receiving SCFP compared with control calves may reduce the negative effects of neutrophil activity in the lungs during BRD.Together, the combined activities of the plasminogen activity assay and SERPIN family of genes suggests a more active and effective lung repair system in calves receiving SCFP supplementation.Our results align with another recent study demonstrating that postbiotic supplementation is associated with improved outcomes at the local site of infection.Vailati-Riboni et al. ( 2021) recently conducted a transcriptomic analysis of mammary gland tissue in response to a Streptococcus uberis mastitis challenge in mid-lactation dairy cows supplemented with SCFP.The mammary tissue in SCFP supplemented cows upregulated many genes and pathways involved in resolution of inflammation.For example, SCFP fed cows upregulated pathways related to 'glutathione metabolism', a master anti-inflammatory molecule found in many tissues (Ghezzi, 2011;Silvagno et al., 2020).Cows receiving SCFP also upregulated pathways related to "complement and coagulation," including several serpin-related genes, much like our own results, which are likely important for resolving  inflammation and protecting mammary tissues from immune-mediated damage (Vailati-Riboni et al., 2021).
There are some indicators that SCFP calves may have more appropriate antiviral responses.Samples of LL lung tissue from SCFP fed calves had higher expression of pathways related to "negative regulation of viral genome replication," which includes genes related to type I interferon responses such as OAS, IFI6, ISG15 and ISG20.Genes in the "IL-27 mediated signaling pathway" were also elevated in LL samples of SCFP fed animals.IL-27 is involved in the immune response against viruses indirectly by increasing production of interferons which have various antiviral effects (Amsden et al., 2022).Although we did not observe significant differences in viral or bacterial burden, overall quantities of virus shed in nasopharyngeal swabs from SCFP fed calves tended to be lower than control calves (Mc-Donald et al., 2021).Thus, in addition to modulating inflammatory responses, SCFP supplementation may also influence antiviral immunity.
In NLL samples, which are grossly healthy lung samples from infected calves, we observed decreased expression of genes involved in inflammation and humoral immune responses in samples from SCFP treated calves compared with controls.During infection, viruses are known to mediate changes in the spatial organization and dynamics of microtubules to drive subcellular viral trafficking toward the nucleus of a target cell (Leopold and Pfister, 2006;Naghavi and Walsh, 2017;Simpson and Yamauchi, 2020).In the current study, we observed decreased expression of genes involved in microtubule movement in SCFP treated calves compared with CON.This decrease in the expression of microtubulesbased movement genes in the NLL tissue could be an indication of reduced viral transport and in turn less inflammation in calves treated with SCFP.(Figure 1, Cluster 2,3).Each row represents expression differences between after and before infection of a given gene.Green represents higher expression after infection and red represents lower expression after infection, with fold change according to the legend.Columns represent samples from CON or SCFP.Data were organized by hierarchical clustering.
Several studies have attempted to understand hostpathogen interactions at the transcriptional level in cattle experiencing BRSV infection or BRD.Previous studies evaluating the immune status of lungs in calves challenged with BRSV identified expression of biological pathways related to stress-associated cellular damage for HSPs such as DNAJB1 (Lebedev et al., 2021) and genes involved in oxidative phosphorylation, mitochondrial dysfunction, and hypoxia (Behura et al., 2017).We did not observe these pathways activated in the BRSV/PM co-infected calves in this trial, regardless of SCFP treatment status.Of note, Behura et al. (2017) collected tissue samples during the peak of BRSV infection, on d 7 after infection, while our tissues were collected slightly later.Another possible explanation could be our use of the combined viral-bacteria coinfection model compared with BRSV infection alone (Behura et al., 2017;Lebedev et al., 2021).However, there were similarities between our results and others.Lebedev et al. (2021) noted upregulation of pathways related to "complement and coagulation cascades," which included several members of the complement cascade such as C1QA, C1QB and CR2.These were expressed at lower levels in lesioned tissues compared with non-lesion sites.In our own animals, we observed upregulation of C8, C9, and CFI, and CFB, all members of the complement cascade, although these genes were less expressed by SCFP calves compared with CON.
Analysis of the transcriptional changes in the lungs before infection enabled further insights into the mecha-nism of action by which SCFP treatment.We observed increased transcription in several genes related to both innate and adaptive immune responses, such as chemokine and cytokine signaling, complement and TLR cascades, leukocyte activation, and cell adhesion.Interestingly, several identified genes are expressed by B cells, including CD19, CD79 and MZB1 (marginal zone B and B1 cell specific protein).Innate-like B cells, such as marginal zone (MZ) and B1 B cells constitutively express MZB1.It plays an important role in integrinmediated cell adhesion (Flach et al., 2010), thus is critical for rapid recruitment, and in secretion of polyreactive IgM (van Anken et al., 2009).MZB1 has also been shown to regulate early IgA secretion, which is critical for defense of mucosal sites such as the lung and gut (Xiong et al., 2019).We have not evaluated IgM or IgA responses in calves receiving SCFP supplementation.However, this gene signature suggests a potential for early and robust antibody responses in mucosal sites of animals receiving SCFP supplementation.We also observed an increase in genes associated with T cell and NK cell signatures in the lungs of SCFP fed calves, such as ICOS, CD8A, CD2, PRF1 and CD4.CD8A is expressed by CD8 T cells, while CD4 is expressed by T helper cells.CD2 is a surface molecule on both T cells and NK cells, that plays a role in their activation.PRF1 encodes for perforin, an important cytotoxic molecule used by both NK cells and T cells to kill virally infected target cells.The elevation of these genes suggests an increased presence of T and NK cell populations in the   airways before the experimental infection.We have not evaluated the frequency of either T cells or NK cells in the lungs of SCFP fed calves before infection; however, we previously noted greater numbers of CD4 T cells and gamma delta T cells in the lungs of SCFP fed calves compared with controls on d 10 after infection (Mahmoud et al., 2020).Having greater proportions of B cells, particularly innate-like B cells, as well as T cells and NK cells positioned in the airways may enable a more rapid cellular and humoral immune response in the lung and could be one mechanism by which SCFPfed animals are more effective at controlling respiratory infection.
The main limitations of our work are the absence of lung tissue samples from uninfected SCFP and control calves.This would have been a valuable comparison for the lung RNaseq analyses.A longitudinal analyses of lung and BAL transcriptomes over time would also provide further insight into the mechanisms of disease, and of SCFP supplementation, on barrier functions and immunity in the lungs of preweaned calves.
Overall, we and others have established that SCFP supplementation benefits health and performance in cattle.Based upon the results of our transcriptional analysis, feeding of SCFP postbiotics appears to have both systemic and local immunomodulatory effects which enhance immune function, while promoting expression of more effective tissue resolution and repair pathways in barrier sites such as the lung.This work provides insight into the potential mode of action by which SCFP products support calf health.), increased expression of cytokine and chemokine genes (i.e., CCL1, CCL20, IL17A) and Complement and TLR signaling cascades.Together, these gene signatures suggest that the respiratory tract of SCFP fed calves is positioned to rapidly respond and repel invading pathogens.Analysis of the BAL and lung tissue transcriptome at necropsy (right panel) revealed that SCFP-fed calves expressed more genes involved with lung tissue repair such as SERPINA2 and SERPINF2, and genes in the plasminogen activating cascade, compared with CON.SCFP fed calves also expressed more genes related to the Type I IFN, or antiviral, response (OAS, IFI6, ISG15 and ISG20), but had overall reduced proinflammatory responses in the lungs, suggesting more effective control of the viral infection, and a more regulated inflammatory response.Figure was created using BioRender.com.
Maina et al.:  Postbiotics impact lung transcriptome during BRD were collected at respective times and stored at −80°C until analysis.
Maina et al.: Postbiotics impact lung transcriptome during BRD Maina et al.:  Postbiotics impact lung transcriptome during BRD Table1.Effects of TLR agonist-induced inflammatory cytokine production by innate immune cells from blood (d 0 and d 10 post viral infection) and BAL (d 14 of the feeding period and d 10 post viral infection) of calves supplemented with or without SCFP for 32 d and coinfected with bovine respiratory syncytial virus (BRSV) and Pasteurella multocida (PM).Data is represented as the least means ± SEM and differences of P ≤ 0.05 were considered significant Cytokine (ng/ml) Figure 1.SCFP calves experienced lower expression of inflammation genes in NLL samples and high expression of genes involved in wound repair in LL samples.(A) Analysis scheme.Lesion (LL) and Non-lesion lung (NLL) tissue gene expression was compared between CON (n = 12) and SCFP (n = 11) calves.(B) Venn diagram showing the number of DEGs in different treatments (FDR <0.05).Shown are 537 DEGs identified in both tissue samples, 52.9% were differentially expressed in NLL compared with 34.4% in the LL samples and only 73 (13%) genes were shared between the 2 tissues.(C) A heat map was generated for the DEGs.Each row represents expression differences between SCFP and CON sample of a given gene, calculated as Log2 fold change of RPKM counts.Green represents higher expression in SCFP samples compared with CON, and red represents lower expression in SCFP compared with CON, with fold change according to the legend.(D) Cluster 4 re-clustering.DEGs between treatment (SCFP) and control for LL and NLL samples.Shown are 132 DEGs previously assigned to cluster 4. Data were organized by hierarchical clustering.
Maina et al.: Postbiotics impact lung transcriptome during BRD Table 2: Gene ontology (GO) analysis.Functional annotation of the DEGs in each cluster of the enriched pathways from nonlesion lung tissue (NLL) and lesioned lung tissue (LL) of calves fed SCFP, or not, for 32 d and infected with bovine respiratory syncytial virus (BRSV) and Pasteurella multocida (PM) Figure 2. Supplementation with SCFP in neonatal calves caused less transcription of inflammation related genes in BAL following BRSV/PM coinfection challenge.(A) Analysis scheme.BAL collected antemortem (d 14 of the feeding period) and postmortem (d 10 post viral infection) was compared between CON (n = 12) and SCFP supplemented (n = 11) calves.(B) Heat Map showing expression differences to the infection response between treatments.Shown are 254 differentially expressed genes identified by the linear model (FDR <0.05).Each row represents expression differences between after and before infection of a given gene.Green represents higher expression after infection and red represents lower expression after infection, with fold change according to the legend.Columns represent samples from CON or SCFP.Data were organized by hierarchical clustering.

Figure 3 .
Figure 3. Analysis of SCFP effects before infection to discern mode of action.(A) Analysis scheme.BAL collected antemortem (d 14 of the feeding period) was compared between CON (n = 12) and SCFP supplemented (n = 11) calves, red box.(B) Volcano plot for differential abundant genes before infection showing the comparison of SCFP vs. CON with significant differential abundance (FDR <0.05).Positive Log 2-Fold Change indicate genes that have higher abundance in SCFP compared with CON, and negative changes indicate lower abundance in SCFP compared with CON.q > 0.001.(C) Functional enrichments for Gene Ontology (GO) terms for biological, cellular, and molecular processes of the DEGs.q-value (FDR-adjusted p-value).

Figure 4 .
Figure 4. Proposed model summarizing the effects of SCFP supplementation on the lung transcriptome of preweaned calves before and after a viral-bacterial coinfection.Comparison of the BAL transcriptome before infection (left panel) between SCFP and CON calves revealed elevated expression of genes encoded by immune cells such as NK cells, CD8 T cells and innate-like B cells (i.e., CD8A, CD2, PRF1, CD79, CD19, MZB1), increased expression of cytokine and chemokine genes (i.e., CCL1, CCL20, IL17A) and Complement and TLR signaling cascades.Together, these gene signatures suggest that the respiratory tract of SCFP fed calves is positioned to rapidly respond and repel invading pathogens.Analysis of the BAL and lung tissue transcriptome at necropsy (right panel) revealed that SCFP-fed calves expressed more genes involved with lung tissue repair such as SERPINA2 and SERPINF2, and genes in the plasminogen activating cascade, compared with CON.SCFP fed calves also expressed more genes related to the Type I IFN, or antiviral, response (OAS, IFI6, ISG15 and ISG20), but had overall reduced proinflammatory responses in the lungs, suggesting more effective control of the viral infection, and a more regulated inflammatory response.Figure was created using BioRender.com.

Table 3 .
Gene ontology (GO) analysis.GO functional enrichment of the differentially expressed genes between BAL samples collected antemortem (preinfection, d 14 of the feeding period) and postmortem (d 10 after infection) from CON and SCFP fed calves infected with bovine respiratory syncytial virus (BRSV) and 1FDR, false discovery rate <0.05.