Identifying novel antifungal proteins from ‘killer yeasts’ – an interview with Paul Rowley


With rising resistance to antifungals, there is a need for new approaches and new drugs to treat fungal infections. In this interview, Paul Rowley (University of Idaho, ID, USA) speaks about his research identifying novel antifungal proteins from Saccharomyces cerevisiae yeasts and whether these could be translated into the medications of the future.

First can you just introduce yourself and tell us a bit about your background?

I was born in the UK and grew up in the town of Burton-on-Trent, which has been brewing beer since the 11th century. Three generations of my family have worked as brewers, which fueled my initial interest in the biology of the brew kettle during high school. Knowing that in the 1800s Burton-on-Trent hosted Fellows of the Royal Society as innovators in biology following the work of Pasteur, motivated me to study microbiology and virology as a first-generation college student at the University of Warwick (UK). I did my PhD at the University of Aberdeen in Scotland (UK). Then I went over to the USA on what I thought would be a short postdoctoral opportunity at the University of Texas at Austin (TX, USA) and that actually turned into a 7-year position in two different labs followed by a very short postdoctoral position at Colorado University, Boulder (CO, USA). While I was relocating to Boulder, I was offered an Assistant Professorship at the University of Idaho, where I am currently a tenure-track Assistant Professor.

And could you give me a bit of background about the research you are presenting here at ASM Microbe (20–24 June 2019, CA, USA)?

The research I am presenting here focuses on antifungals produced by common brewer’s yeasts that target Candida glabrata, but the project didn’t initially start that way! Since the early 1960s Brewer’s yeast (Saccharomyces cerevisiae) has been known to produce antifungal toxins, however, it is widely reported in the literature that these toxins are strain-specific and species-specific. So, if you exposed different strains of fungi to the same toxin, you will find varying levels of susceptibility and resistance and that’s not the greatest property for an antifungal drug because you ideally want a drug that would be broad in its spectrum, at least across an entire species.

From the S. cerevisiae data, we identified almost 400 toxin-producing yeasts within the Saccharomyces genus and as expected, there is a mosaicism of susceptibility to these toxins within the same species. I was interested in the mosaicism because Saccharomyces yeasts are very genetically similar. It is plausible that these differences in susceptibility are being driven by intense competition within the environment between different strains of yeasts; for example, if an apple falls from a tree and gets eaten by an insect, the insect most likely carries yeasts with it and the yeast takes advantage of the sugars present within that apple, leading to fermentation. Then the yeasts start producing these toxins essentially to ward off competitors, so we have a dynamic, competitive environment.

When I was writing grants to look into this phenomenon most of the feedback was stating that these toxins could never be therapeutics because of the known strain- and species-specificity, which is a valid concern. But my students were pushing me to try our toxin-producing yeasts against some pathogens, and eventually I relented. To start with there wasn’t much to report – we grew these toxin producers in the presence of Candida albicans, which one or two yeasts inhibited just a little bit, and Candida auris, which again saw a little bit of inhibition. Then we tried the toxins against Candida glabrata and the initial results with the wild-type strain suggested the fungi was susceptible to about 40% of the killer yeasts that we threw at it. This was interesting as all the other Candida yeasts had around 1–2% susceptibility. That spurred us to contact the CDC (GA, USA) to obtain strains of C. glabrata that had resistance to common therapeutics, including azoles and echinocandins. To summarize, we have now tested over 100 different strains of C. glabrata and we have yet to find one that is resistant to our killer yeasts – this broad spectrum of antifungal activity has never been observed for any toxin-producing Saccharomyces yeast against any particular fungal species. That’s what our work is based on and we have a provisional patent now in place for these toxins.

Could you tell us a bit more about the toxins and their mechanism of action?

The toxins in themselves are not new but we did have to confirm the type of toxins that were active against C. glabrata. Interestingly, these toxins aren’t actually encoded by the Saccharomyces yeast, they are encoded by a double-stranded RNA, and this RNA requires a double-stranded RNA virus to support its replication. Killer yeasts are infected with these dsRNAs most of the time; but without the virus the yeast can’t maintain the toxin-encoding dsRNA and can’t produce the toxin, so this is a multi-layered symbiosis. We developed a sequencing technique, published on this in the early part of this year, directly sequencing these double-stranded RNAs to gain insights into the viruses and these antifungal toxins. We have identified the toxins that are most inhibitory to Candida glabrata are canonical toxins that are thought to punch holes in cellular membranes. The novelty in our work is the fact that it was assumed that these toxins don’t obliterate an entire species but our data with C. glabrata suggests otherwise.

Could you comment a bit more on the need for new antifungals?

Talking specifically about Candida glabrata, it is the second leading cause of invasive candidosis and anywhere between 20–30% of Candida glabrata isolates have been reported to be resistant to frontline antimicrobial therapeutics. This is a drug-resistant hospital-acquired pathogen that can have quite a high mortality rate in individuals who lack a fully functional immune system, which is a big clinical concern. If a strain is resistant to azoles and echinocandins we don’t really have anything else that you can use to effectively treat infections caused by these yeasts, so we need new antifungals.

This image shows Yeasts inhibiting Candida glabrata

Used with permission of Dr Paul Rowley

Currently, we are interested in exploring vulvovaginal candidosis, which is another major problem that Candida glabrata contributes to. If a patient experiences recurrent bouts of vulvovaginal candidosis that is refractory to azole treatment then that may be indicative of Candida glabrata infection. The toxins we’ve isolated are natural products and if you think about where they are coming from – rotting fruit environments etc. – they have evolved a very tight pH optimum at about 4.5–4.6. This isn’t ideal if you want to put it in the bloodstream, however, the vaginal mucosa has a pH of approximately 4.5. So, could we use natural products in this scenario where you want to specifically target Candida glabrata. As we learn more about the potential clinical application of these antifungal toxins, we hope to engage the biotech industry to bring these potential therapeutics to market.

So, is that the next step for your research?

Exactly. What we don’t really understand is that there is variability in the ability of different strains to inhibit the growth of pathogenic yeasts, for example, we have several strains that produce very similar toxins but have differing abilities to inhibit the growth of certain yeasts. We are not 100% sure whether these subtle difference between toxins are changing their antifungal activities. A paper I published earlier this year shows that a few mutations can have quite a drastic effect on the efficacy of an antifungal toxin.

Another aspect is trying to test the cytotoxicity of these toxins to human cells, which is something that’s not widely considered because we are constantly exposed to Saccharomyces yeasts in many foods and products; coffee production, cacao production, brewing, winemaking, baking, all use yeasts that produce these toxins. In milligram amounts, if that was the therapeutic range, maybe there is some cytotoxicity and that’s never been explored in the literature as far as I am aware.

Finally, any other comments?

All of the work in my laboratory was performed by undergraduate researchers at the University of Idaho and my amazing technician. I am very grateful to my teams’ hard work and their perseverance and without them we would not have been able to make any of these discoveries. The work was financially supported by grants from the National Institutes of Health (NIGMS #P20GM104420 and INBRE Grant #P20GM103408), the National Science Foundation (MCB #1818368) and intramural grants that support undergraduate research at the University of Idaho (particularly the Office of Undergraduate Research).

You might also like:


Leave A Comment