Background: Killed bacteria are used in functional assays to examine cell-mediated immunity. Several methods are available to deactivate bacteria, but it is uncertain which process produces an inactivated organism that most closely resembles a living organism. We aimed to identify which method could kill Streptococcus pyogenes, whilst maintaining cellular surface structures and its ability to activate immune pathways.

Methods: We used ethanol, heat, ultra-violet light (UV), paraformaldehyde and γ-irradiation to kill four S. pyogenes strains (M1, M12, M53, M75). The resulting preparations, and live controls, were used to stimulate human whole blood. Cytokine release (GM-CSF, IFN-γ, TNF-α, IL-2, IL-4, IL-6, IL-8, IL-10) was measured by multiplex bead immunoassay, and scanning electron microscopy (SEM) was used to examine cell structures and integrity. Inter-individual immunity to the four strains was measured by bactericidal assay and correlated with stimulation results.

Results: Heat, ethanol, and γ-irradiation completely killed S. pyogenes, whilst paraformaldehyde and UV did not. Cytokine production varied between strains, donors and killing methods, with a trend towards greater cytokine production from γ-irradiated bacteria. Variance observed between strains killed by ethanol, heat and γ-irradiation was comparable. While killed and live bacteria appeared similar by SEM, UV led to relatively denuded bacteria and cell division was affected by ethanol and heat.

Conclusion: Bacteria killed by different methods give variable results in immunoassays, highlighting the need for consistent, standardised reagents. Our findings support the use of γ-irradiated S. pyogenes in immunoassays, which reliably elicited high levels of cytokines, without obvious destruction of bacterial cell-surface structures.