Aerosol vaccination with Bacille Calmette-Guerin induces a trained innate immune phenotype in calves

Abstract

Mycobacterium bovis Bacillus Calmette-Guérin (BCG) is a live attenuated vaccine for use against tuberculosis (TB); however, it is known to reduce childhood mortality from infections other than TB. The unspecific protection induced by BCG vaccination has been associated with the induction of memory-like traits of the innate immune system identified as 'trained' immunity. In humans and mouse models, in vitro and in vivo BCG training leads to enhanced production of monocyte-derived proinflammatory cytokines in response to secondary unrelated bacterial and fungal pathogens. While BCG has been studied extensively for its ability to induce innate training in humans and mouse models, BCG's nonspecific protective effects have not been defined in agricultural species. Here, we show that in vitro BCG training induces a functional change in bovine monocytes, characterized by increased transcription of proinflammatory cytokines upon restimulation with the toll-like receptor agonists. Importantly, in vivo, aerosol BCG vaccination in young calves also induced a 'trained' phenotype in circulating peripheral blood mononuclear cells (PBMCs), that lead to a significantly enhanced TLR-induced proinflammatory cytokine response and changes in cellular metabolism compared to PBMCs from unvaccinated control calves. Similar to the long-term training effects of BCG reported in humans, our results suggest that in young calves, the effects of BCG induced innate training can last for at least 3 months in circulating immune populations. Interestingly, however, aerosol BCG vaccination did not 'train' the innate immune response at the mucosal level, as alveolar macrophages from aerosol BCG vaccinated calves did not mount an enhanced inflammatory response to secondary stimulation, compared to cells isolated from control calves. Together, our results suggest that, like mice and humans, the innate immune system of calves can be 'trained'; and that BCG vaccination could be used as an immunomodulatory strategy to reduce disease burden in juvenile food animals before the adaptive immune system has fully matured.

Fig 1. In vitro BCG infection of bovine monocytes enhances the expression of proinflammatory cytokines after stimulation with LPS or Pam3CSK4.

(A) Diagram of the in vitro training model used in bovine isolated monocytes. (B) IL-6 and TNFα expression after restimulation for 4 hours. Monocytes isolated from peripheral blood of six healthy heifers were plated at 1x10⁶ cells/mL in a 96-well plate. Cells were infected in vitro with BCG at a MOI of 5:1 for 24 hours, washed once with warm cRPMI, incubated for 6 days, and stimulated with LPS, Poly(I:C) or Pam3CSK4 for 4 hours. Uninfected cells (cRPMI) served as negative controls. Changes in gene expression were assessed by RT-PCR. Data represent means ± SEM (n = 6). *p<0.05.

Fig 2. BCG vaccination increased nonspecific production of monocyte-derived proinflammatory cytokines.

(A) Diagram showing the BCG vaccination schedule in calves. Calves were vaccinated with 1x10⁸ colony forming units (CFU) BCG Danish via aerosol. Peripheral blood was collected four weeks post vaccination from calves in both groups. (B) PBMCs isolated from calves after four weeks post vaccination, were stimulated in vitro with LPS or Pam3CSK4 for 4 or 72 hours to measure cytokine expression. Proinflammatory cytokine gene expression was assessed by RT-PCR (B), and protein expression was assessed by ELISA on the cell supernatants (C). Data represent means ± SEM. *p<0.05, **p<0.01.

Fig 3. BCG vaccination increases glucose consumption and lactate production in bovine trained monocytes.

Peripheral blood was collected four weeks postvaccination from calves in both groups. PBMCs from unvaccinated (grey bars) and BCG vaccinated (black bars) animals were stimulated in vitro for 72 hours with LPS or Pam3CSK4, and then glucose consumption (A) and lactate concentration (B) were measured in the cell culture media. Data represent means ± SEM. Not significant (NS), *p<0.05, **p<0.01, *** p<0.001.

Fig 4. BCG aerosol immunization does not alter the frequency or surface marker expression of bovine circulating monocytes.

(A) Flow cytometric analysis of circulating CD14+ CD11b+ monocytes isolated 4 weeks after BCG immunization. Average frequencies of CD14+ or CD11b+ cells were analyzed, as were surface expression levels (MFI) of CD11b on cells isolated from both groups of calves (n = 7/group). (B) Monocytes were isolated as described in Materials and Methods section and analyzed by RT-PCR for the mRNA expression of TLR2 and TLR4, after 4 weeks of BCG vaccination. Data represent means ± SEM. No significant (NS) differences were observed between treatment groups.

Fig 5. BCG vaccination induces long-term ‘trained’ phenotype in bovine PBMCs.

PBMCs were isolated at twelve weeks post vaccination and were stimulated in vitro with LPS or Pam3CSK4 for 4 or 72 hours to measure cytokine expression. Proinflammatory cytokine gene expression was assessed by RT-PCR (A) and protein expression was assessed by ELISA in the supernatants (B). Data represent means ± SEM. No significant (NS), *p<0.05, **p<0.01, *** p<0.001.

Fig 6. BCG aerosol vaccination does not alter the function of alveolar macrophages.

Calves were vaccinated with 1x10⁸ colony forming units (CFU) BCG Danish via aerosol. Peripheral blood was collected four weeks post vaccination from calves in both groups. PBMCs were isolated from calves at four weeks post vaccination and were stimulated in vitro with LPS or Pam3CSK4 for 4 or 72 hours to measure cytokine expression. Proinflammatory cytokine gene expression was assessed by RT-PCR (A), and protein expression was assessed by ELISA in the supernatants (B). Data represent means ± SEM. No significant (NS), *p<0.05.