PBMC transcriptomic responses to primary and secondary vaccination differ due to divergent lean growth and antibody titers in a pig model

Signaling differences in peripheral blood mononuclear cells of high and low vaccine responders prior to, and following, vaccination in piglets

Individual variability in responses to vaccination can result in vaccinated subjects failing to develop a protective immune response. Vaccine non-responders can remain susceptible to infection and may compromise efforts to achieve herd immunity. Biomarkers of vaccine unresponsiveness could aid vaccine research and development as well as strategically improve vaccine administration programs. We previously vaccinated piglets (n = 117) against a commercial Mycoplasma hyopneumoniae vaccine (RespiSure-One) and observed in low vaccine responder piglets, as defined by serum IgG antibody titers, differential phosphorylation of peptides involved in pro-inflammatory cytokine signaling within peripheral blood mononuclear cells (PBMCs) prior to vaccination, elevated plasma interferon-gamma concentrations, and lower birth weight compared to high vaccine responder piglets. In the current study, we use kinome analysis to investigate signaling events within PBMCs collected from the same high and low vaccine responders at 2 and 6 days post-vaccination. Furthermore, we evaluate the use of inflammatory plasma cytokines, birthweight, and signaling events as biomarkers of vaccine unresponsiveness in a validation cohort of high and low vaccine responders. Differential phosphorylation events (FDR < 0.05) within PBMCs are established between high and low responders at the time of vaccination and at six days post-vaccination. A subset of these phosphorylation events were determined to be consistently differentially phosphorylated (p < 0.05) in the validation cohort of high and low vaccine responders. In contrast, there were no differences in birth weight (p > 0.5) and plasma IFNγ concentrations at the time of vaccination (p > 0.6) between high and low responders within the validation cohort. The results in this study suggest, at least within this study population, phosphorylation biomarkers are more robust predictors of vaccine responsiveness than other physiological markers.

Plasma Cytokines and Birth Weight as Biomarkers of Vaccine-Induced Humoral Responses in Piglets

Failure to mount an effective immune response to vaccination leaves individuals at risk for infection and can compromise herd immunity. Vaccine unresponsiveness can range from poor responses “low responders” to a failure to seroconvert “non-responders.” Biomarkers of vaccine unresponsiveness, particularly those measured at the time of vaccination, could facilitate more strategic vaccination programs. We previously reported that pro-inflammatory cytokine signaling within peripheral blood mononuclear cells, elevated plasma interferon-gamma (IFNγ), and low birth weight correlated with vaccine-induced serum IgG titers in piglets that were below the threshold of detectable seroconversion (vaccine non-responders). These observations suggested that plasma IFNγ concentration and birth weight might serve as pre-vaccination biomarkers of vaccine unresponsiveness. To test this hypothesis, piglets (n = 67) from a different production facility were vaccinated with the same commercial Mycoplasma hyopneumoniae bacterin (RespiSure-One) to determine if there was a consistent and significant association between vaccine-induced serum IgG titers and either plasma cytokine concentrations or birth weight. All piglets seroconverted following vaccination with significantly less variability in vaccine-induced serum IgG titers than observed in the previous vaccine trial. Piglets exhibited highly variable birth weights and plasma cytokine concentrations prior to vaccination, but there were no significant associations (p > 0.05) between these variables and vaccine-induced serum IgG titers. There were significant (p < 0.001) differences in plasma IFNγ concentrations among individual litters (n = 6), and plasma IFNγ concentrations decreased in all pigs from birth to 63-days of age. One of the six litters (n = 11 piglets) exhibited significantly elevated plasma IFNγ concentrations during the first 3 weeks of life (p < 0.001) and at the time of vaccination (p < 0.01). This litter, however, had similar vaccine-induced serum IgG titers when compared to the other piglets in this study. Collectively the two studies indicate that while plasma cytokines and birth weight can be associated with vaccine non-responsiveness, their temporal and individual variation, as well as the complexity of the vaccine responsiveness phenotype, make them inconsistent biomarkers for predicting the less extreme phenotype of vaccine low responders.

Characterization of whole blood transcriptome and early-life fecal microbiota in high and low responder pigs before, and after vaccination for Mycoplasma hyopneumoniae

We investigated gene expression patterns in whole blood and fecal microbiota profile as potential predictors of immune response to vaccination, using healthy M. hyopneumoniae infection free piglets (n = 120). Eighty piglets received a dose of prophylactic antibiotics during the first two days of life, whereas the remaining 40 did not. Blood samples for RNA-Seq analysis were collected on experimental Day 0 (D0; 28 days of age) just prior to vaccination, D2, and D6 post-vaccination. A booster vaccine was given at D24. Fecal samples for microbial 16SrRNA sequencing were collected at 7 days of age, and at D0 and D35 post-vaccination. Pigs were ranked based on the levels of M. hyopneumoniae-specific antibodies in serum samples collected at D35, and groups of 'high' (HR) and 'low' (LR) responder pigs (n = 15 each) were selected. Prophylactic antibiotics did not influence antibody titer levels and differential expression analysis did not reveal differences between HR and LR at any time-point (FDR > 0.05); however, based on functional annotation with Ingenuity Pathway Analysis, D2 post-vaccination, HR pigs were enriched for biological terms relating to increased activation of immune cells. In contrast, the immune activation decreased in HR, 6 days post-vaccination. No significant differences were observed prior to vaccination (D0). Two days post-vaccination, multivariate analysis revealed that ADAM8, PROSER3, B4GALNT1, MAP7D1, SPP1, HTRA4, and ENO3 genes were the most promising potential biomarkers. At D0, OTUs annotated to Prevotella, CF21, Bacteroidales and S24-7 were more abundant in HR, whereas Fibrobacter, Paraprevotella, Anaerovibrio, [Prevotella], YRC22, and Helicobacter positively correlated with the antibody titer as well as MYL1, SPP1, and ENO3 genes. Our study integrates gene differential expression and gut microbiota to predict vaccine response in pigs. The results indicate that post-vaccination gene-expression and early-life gut microbiota profile could potentially predict vaccine response in pigs, and inform a direction for future research.

Genetic variation in humoral response to an Escherichia coli O157:H7 vaccine in beef cattle

Individuals often respond differently to the same vaccine; some of this variation may be caused by genetic differences among animals. Our objective was to estimate heritability and identify genomic regions associated with humoral response to an Escherichia coli O157:H7 vaccine in beef cattle. Crossbred beef cattle (n = 651) were vaccinated with a commercially available E. coli O157:H7 vaccine. Serum was collected at time of initial vaccination (d 0), booster (d 21), and d 56 after initial vaccination. Total antibodies specific to siderophore receptor and porin proteins in the vaccine were quantified by enzyme-linked immunosorbent assay. Genomic DNA was isolated from whole blood and genotyped with the bovine GeneSeek Genomic Profiler-High Density 78K or 26K Single Nucleotide Polymorphism BeadChip and imputed to 777,000 SNP genotypes. Heritability was estimated by restricted maximum likelihood (REML) using both 1) pedigree and 2) genomic relationships among individuals. Fixed effects were contemporary group, calf age, sex, principal components from SNP genotype data, and pedigree-derived heterozygosity effects. Additive and dominance effects of SNPs were estimated individually while accounting for contemporary group, sex, and the top 20 principal components calculated from the genomic relationship matrix. Heritability of initial response to vaccination (d 21 –d 0) was 0.10 ± 0.175 using pedigree relationships and 0.14 ± 0.149 using genomic relationships, but neither estimate was statistically different from zero. Heritability of booster (d 56 –d 21) and overall (d 56 –d 0) responses were low and not statistically significant from zero. There were no clusters of linked SNP associated with vaccine response, but eight regionally isolated SNPs were significantly associated with initial or overall response to vaccination. Regional genetic variation for initial response to an E. coli O157:H7 vaccine was observed, although overall heritability of this response was not statistically significant from zero.