Development and Testing of a Five-Subunit Biofilm Vaccine for the Prevention of Pulmonary Tuberculosis

Abstract

Mycobacterium tuberculosis (MTB) is the causative agent of tuberculosis, a disease that kills 1.4 million people and infects 8.7 million people worldwide every year. This bacterium has traditionally been studied using in vitroshaking cultures including detergents, which force the bacteria into an artificial style of planktonic growth. We utilized a static model of MTB growth that allows the bacteria to naturally grow as a biofilm pellicle at the air-liquid interface and utilized this model to identify antigens common to both shaking and biofilm cultures through a process of two-dimensional gel electrophoresis combined with western blotting. Having identified proteins that were both immunogenic and produced at three biofilm time points (three, five, and seven weeks) as well as in our control shaking culture, we proceeded to recombinantly express five in their full length forms. Three proteins were purified in their native forms and two proteins first denatured during the purification process and then refolded. We vaccinated female BALB/c and C3HeB/FeJ mice using 25 micrograms of each of the five antigens combined with either cyclic-di-GMP or DDA/MPL as an adjuvant over the course of three vaccine trials. While all vaccine formulations elicited a strong immune humoral response with high IgG1 and IgG2a titers against the vaccine antigens, there were no significant reductions in the number of CFU/g tissue in spleens or lungs of vaccinated animals compared to unvaccinated controls. Overall lung burden in BALB/c mice was significantly reduced in BCG vaccinated animals (p < 0.05) while there was a trend for bacterial burden reduction in animals vaccinated with the antigen subunits and DDA/MPL adjuvant (p = 0.055). In addition, the difference in the immune responses of the difference strains of mice was apparent when the T-helper responses of BALB/c and C3HEB/FeJ mice were evaluated. In the case of testing the DDA/MPL vaccine, BALB/c mice produced a strong antigen-specific IFN-gamma; response from vaccinated mouse splenocytes compared to the C3HeB/FeJ mice that produced significantly more IL-10 relative to IFN-gamma;. This indicated the fundamental differences in mouse strain responses to vaccination with BALB/c mice skewing towards a Th1 response relative to C3HeB/FeJ mice that had a more Th2/Treg skewed immune response. In addition, there were fundamental differences in the properties of the adjuvant-dependent immune responses. While DDA/MPL adjuvant showed a trend towards challenge protection, we found that cyclic-di-GMP-treated animals did significantly worse than control animals with an accompanying elevation in lung IFN-gamma; levels. These results call into question the use of cyclic-di-GMP adjuvant for tuberculosis vaccines, since it may be counterproductive to a protective immune response to MTB. Furthermore, the recombinant proteins 35kdag, CeoB, Mkl, and TB31.7 warrant further investigation as potential diagnostic markers for active or latent tuberculosis.