Serendipity in Science

Like Comment

The paper in Nature Microbiology is here.

This project has been a commentary on the serendipity of the scientific process and the stories you stumble across whilst pursuing other questions. I have always been interested in the interactions between the pathogen, host microbiota and the host immune response in the one location where all three must encounter each other - the mucosa. After all, from a mucosal pathogen’s point of view, they must first encounter the host microbiota before they encounter (and activate) the host immune response.  I came to the Iwasaki lab to explore the role of the vaginal microbiota in mucosal viral infections. Exciting work from the Pfeiffer, Karst and Golovkina labs have detailed ways in which enteric viruses take advantage of the host intestinal microbiota to enhance infection. So when I first observed that pretreatment of mice with an intravaginal antibiotic cocktail protected them from genital herpes infection, I thought it fit well with the larger idea of viruses leveraging host commensals to better infect the host.

Hunting the elusive commensal

We quickly figured out that aminoglycosides were the key antibiotic in the cocktail mediating the antiviral effect. We attacked the problem from both the host and the commensal end trying to determine the cellular and intracellular requirements for aminoglycoside-based protection, while simultaneously trying to identify the host commensal population targeted by aminoglycosides. On the host side, microarray analysis showed us that aminoglycoside treatment induced robust upregulation of a broad array of transcripts in the interferon-signaling pathway (interferon-stimulated genes or ISGs). This ISG induction was critical to aminoglycoside antiviral activity as members of the aminoglycoside family that didn’t induce ISGs also failed to protect mice against viral infections.

However, after a frustrating year trying to identify mouse vaginal commensals using PCR and culture techniques, it finally occurred to us to test if commensals were involved at all. With our collaborators, we treated germ-free mice with aminoglycosides and found that they were just as protected against HSV infection. This was a direct effect of the aminoglycoside on the host! 

Aminoglycosides directly induce host immune response

Aminoglycosides kill bacteria by binding ribosomal RNA and inducing faulty protein synthesis. These compounds have been shown to bind human mitochondrial ribosomal RNA which can result in specialized tissue toxicity (ototoxicity and nephrotoxicity). Within the vaginal mucosa, application of aminoglycoside does not seem to cause any discernible toxicity - for example, a pathologist could not differentiate between treated and untreated samples. We do see activation of RNA sensor TLR3 however, and the TLR3-TRIF-IRF3/7 pathway is required for aminoglycoside-ISG induction and subsequent antiviral activity. Intravaginal aminoglycoside treatment increased recruitment of multiple cell types to the vaginal mucosa and we identified the specific dendritic cell subtype XCR1+ DCs (which uniquely express TLR3) as responsible for the increased ISG production.

Our data suggest a model wherein aminoglycoside-bound RNA binds TLR3 and induces ISG expression. However, the nature of the host RNA bound by the aminoglycoside remains an open question. Mitochondrial rRNA seems a likely candidate but we have no data to support that hypothesis yet. We do show data that aminoglycosides can enhance the immunogenicity of synthetic TLR3 ligands suggesting that they may play a similar role with endogenous RNA ligands.

A word of warning

We must emphasize that we are not advocating the use of aminoglycoside antibiotics as antiviral compounds, especially since their effect on the commensal bacteria is an important caveat to their usefulness as antivirals. The rise of antibiotic-resistant infections due to overuse and misuse of antibiotics remains a critical concern. Rather, we report an interesting side-effect of commonly used antibiotic compounds and suggest that further investigation of these mechanisms of protection will help us design better pan-antivirals for mucosal infections. 

Illustrating science

Continuing with the theme of serendipity, we were lucky to have a talented illustrator within our science family. One lab meeting, Akiko surprised us with the video above that perfectly illustrated the essence of our findings in under two minutes. This video was made by her daughter Emi, whose drawings I had often admired on the walls of Akiko’s office. Discussing the illustrations with Emi clarified the questions we had left to answer and resulted in the lovely chalkboard drawings that form the banner for this post. 

This post was written by Smita Gopinath and illustrated by Emi Iwasaki-Medzhitova.

Smita Gopinath

Postdoctoral Associate, Howard Hughes Medical Institute, Yale School of Medicine