Jeffrey Moffat and David Erickson, Microbiology and Molecular Biology

Introduction

Chemokines are tiny proteins that play a valuable role in defending our bodies against bacteria and other pathogens. Chemokines are best known for their role in attracting immune cells to areas of infection. Some chemokines, however, have demonstrated the ability to not only recognize pathogens and/or recruit white blood cells, but also to kill various classes of bacteria and fungi¹. One of these antimicrobial chemokines is CCL28.

Although CCL28 has been shown to bind to and kill many different types of microbes, some bacteria are able to resist its antimicrobial effects. One resistant bacteria species is Yersinia pseudotuberculosis, a close relative of the bubonic plague-causing Yersinia pestis. The aim of this project was to find out how Yersinia pseudotuberculosis is able to resist the antimicrobial activities of CCL28 and thereby elucidate the mechanism by which CCL28 kills bacteria in general.

Other researchers, under the direction of Dr. Eric Wilson and Dr. David Erickson, previously found that truncating Yersinia pseudotuberculosis’ lipopolysaccharides (LPS – structures that consist of a long sugar chain bound to a lipid extending from the cell membrane) made the bacteria considerably more susceptible to CCL28 binding. Truncating the LPS structure in this manner exposes a pair of negatively charged phosphate groups proximal to the cell membrane, which are believed to attract a positively charged portion of CCL28². When bacterial strains that expressed truncated LPS structures were treated with an agent that removed all of the bacterial surface proteins and then incubated with CCL28, however, binding of CCL28 was almost completely inhibited. This indicated that, while truncating the LPS structure does increase CCL28 binding, the actual target to which CCL28 could potentially be a protein.

Methodology

In order to discover the protein target for CCL28 binding, we randomly mutated strains of Yersinia pseudotuberculosis that already possessed a truncated LPS structure to screen for genetic mutations which inhibit CCL28 binding. These screens were carried out using magnetic immunoprecipitation and flow cytometry to quantify CCL28 binding. We then used DNA sequencing to identify protein targets required for CCL28 binding. These potential targets include membrane proteins that are predicted to be physically associated with LPS. One isolated mutant was particularly interesting, because a gene known as pmrI was knocked out. PmrI is part of a larger group of connected genes, the pmr operon, which encode several proteins. One of these proteins, pmrK, associated closely with the LPS structure, making it a prime potential target for CCL28 binding.

In order to identify whether the presence of the pmrK protein has an effect on CCL28 binding, we amplified the pmrK gene and inserted it into a plasmid, which we grew up in E. coli and isolated. In order to test whether the pmrK protein is important for CCL28 binding, we planned to insert the plasmid carrying the pmrK gene into a Yersinia pseudotuberculosis strain in which LPS was truncated and the pmr operon was knocked out. If expressing pmrK in these bacteria were to appreciably restore CCL28 binding, our hypothesis that the pmrK protein is a co-target for CCL28 binding would be supported.

Results

In the course of carrying out the experiments outlined above, however, we ultimately decided to take a different approach to ascertain CCL28’s pathogen binding mechanism. Our binding assay and sequencing data indicated that several bacterial proteins potentially contribute to CCL28 binding, but no single mutant produced as extreme of a phenotype as we saw when bacteria were cleared of all surface proteins. We were successful in isolating the pmrK gene, inserting it into a plasmid, and amplifying it, but we did not go on to perform a transformation with the pmrK containing plasmid and pmrI-knockout bacteria.

Discussion

Before continuing with any experiments utilizing the pmrI knockout strain, we have been, and will continue, treating Yersinia pseudotuberculosis strains resistant to other antimicrobial peptides, such as polymixin, with CCL28. This will help us to discover any similarities or differences in binding between CCL28 and peptides with known antimicrobial mechanisms. We would also like to devise a way to isolate all surface proteins and then to use co-immunoprecipitation, or something similar, to find which proteins bind with CCL28.

Conclusion

Very little research has been done on CCL28 and its antimicrobial properties at this point. Uncovering its targets for binding to the surface of Yersinia pseudotuberculosis and other bacteria, however, could help in developing treatments for a wide range of infections. Furthermore, continuing this project could aid in the future implementation of CCL28, or its components, in a clinical setting. On a more basic level, this research could also provide valuable insight into the processes by which the immune system generally fights infection and deals with resilient bacteria.

¹Yung S.C., Murphy P.M. Antimicrobial chemokines. Front Immunol. 2012;3:276.
²B. Liu, E. Wilson. 2010. The Antimicrobial Effect of CCL28 is Dependent on C-terminal Positively Charged Amino Acids. Eur. J. Immunol. 40(1): 186–196.