MicroRNAs in the spotlight

By  Ian Dewar 13 November 2024 4 min read

MicroRNAs (miRNAs) were initially considered 'junk' genetic material when they were first discovered 30 years ago.

Now, the Nobel prize for Medicine has been awarded to the scientists who discovered them.

They can help detect infections based on the host response.

RNA is having a moment

Dr Ryan Farr specialises in microRNA research in our Host Response team based at the Australian Centre for Disease Preparedness in Geelong.

"Initially researchers thought microRNAs were junk genetic material that had just accumulated from random mutations or viral infections," Ryan says.

MicroRNAs were discovered by postdoctoral researchers Victor Ambros and Gary Ruvkun. We didn't know it at the time, but these tiny bits of genetic material are actually central to how we develop and function.

Hold the genome

For a long time, scientists thought that the complexity of a living organism was directly proportional to how many protein coding genes they had.

"Back in the 60s scientists predicted that human cells had 2 million genes," Ryan says.

The Human Genome Project revealed we actually have more like 20,000 genes. So how is it that a rice plant can have more than triple that number?

"It turned out that it wasn't how many genes you had that mattered. It was how those genes were used," Ryan says.

It came down to the fine tuning and regulation of gene expression. And this is where microRNAs come in.

MicroRNAs bind to protein-coding RNAs, tying or chopping them up to stop certain proteins being made. This fine-tuned regulation of gene expression allows complex organisms to develop their specialised organs and tissues.

"What really drew me into this area of research was this idea that there's no great correlation between the number of protein coding genes an organism has and how complex that organism is," Ryan says.

"But there is an association between the non-coding RNA an organism has and how complex it is."

The discovery of microRNAs kicked off the process of trying to understand what all the other classes of RNA do.

"Now we know there are lots of different classes of RNAs that do things other than create proteins," Ryan says.

Cell-ebrating microRNAs

MicroRNAs don’t just act within cells. They can be released from cells and move around the body. And they react rapidly to different stimulus.

"They're really important for development and function at a cellular and physiological level. They also give us insight into what's happening in the body," Ryan says.

The cell might be injured or it might die because it's been infected. Or it might be packaging up and sending out signals to help other cells regulate their gene expression as well.

MicroRNAs are also relatively stable and easy to detect and measure. This makes them promising biomarkers.

A revealing response

Our Host Response team is looking at how microRNAs change in response to infection. 

"Standard diagnostics look for either the pathogen itself or the antibodies made in response to the pathogen," Ryan says.

"Instead of going after the pathogen directly, we look at the host response," he says.

It's about tapping into the early immune response. So, what are the microRNAs doing? And can we pull out those signatures and use them to give us early warning about what's happening?

Artificial intelligence (AI) for microRNAs

We know of around 1,600 microRNAs in humans, and over 1,000 in animals. So, the question is which microRNAs to focus on.

This is where Machine Learning comes in. It trawls through all these molecules and figures out what combination of two, three or four molecules together gives us the best predictive power.

AI also sits underneath the diagnostic platform itself.

From infection to detection

Using all these microRNA clues our researchers are creating new diagnostic tests to find diseases earlier, even before symptoms appear.

For example, some diseases in livestock have a really long period, maybe years, where animals are infected but the disease can't be detected.

"We're now measuring the host response to different diseases of livestock," Ryan says.

"We're finding that the host response is very specific for each disease."

Cracking the code

Ryan wasn't a trained computer scientist when he started out. Instead, he was a wet lab scientist working in molecular biology. He slowly learned how to code and then jumped into machine learning and started applying it.

Subsequently, Ryan was awarded funding from the AI for Missions Program and used this to set up a model to turn raw data into predictive models. He and his team really wanted to develop tools using AI that was easily explainable. 

"We didn't want a closed black box that nobody could understand," he says.

"So, a lot of our work has been around cracking everything open and making sure we understand how these models are functioning."

Big future for tiny molecules

He now works at the interface between algorithms, machine learning and high-performance computing, and molecular biology.

"I love this area because each infection is slightly different," Ryan says.

Using microRNAs for host response disease detection is a new area, so it's not that well known.

"It's just super exciting to see microRNAs being harnessed in this way. And it's fantastic to see them now getting the recognition they deserve," Ryan says.

These tiny molecules have a big future ahead in disease diagnosis.