Automation, Insight, and the Human Element

The only good microbe is a dead microbe.

David Kirk¹ and Willem Meindert de Vos^3
1 D Kirk Communcations, ³ University of Helsinki

“The only good microbe is a dead microbe.” That thinking, even just a few decades ago, dominated clinical microbiology. But today, microbes are recognised as central players in human health – some even becoming real-world therapies. Now, the research challenge is understanding how these microbes function within their communities.

In this article, we speak with Professor Willem M. de Vos, Emeritus Professor at the University of Helsinki and former Emeritus Professor and Chair of Microbiology at Wageningen University. His work on the human microbiome led to the discovery and characterisation of the keystone human gut bacterium Akkermansia muciniphila. He has also co-founded several companies, including The Akkermansia Company (recently acquired by Danone) and Alba Health, which focuses on gut health and nutrition in children.

From Enemies to Good Bacteria 

For most of the 20th Century, bacteria were the enemy – the best microbe was a dead one. But over the last 30 years, we’ve come to appreciate their role in human health.

Parents’ microbes colonise their infants from birth and play a major role in protecting against pathogens, priming our immune system and metabolism from an early age. 

“Basically, we’re born sterile and colonised by our mother’s microbes. We have also shown we’re colonised a bit by the father recently, but it’s mainly the mother.”

Our microbiome also plays a protective role throughout our lives, preventing pathogenic bacteria from colonising our gut. However, when our microbiome becomes disrupted – such as after a course of antibiotics – opportunistic, antibiotic-resistant bacteria like Clostridium difficile can take hold.

Fortunately, bugs are even better than drugs.

De Vos highlights a study at Amsterdam Medical Centre he was involved in that used a faecal microbiological transplant (FMT) – a mixture of gut microbes sourced from a healthy gut – as an early success of microbiome therapy.

“Patients with recurrent Clostridium difficile infections treated with FMT walked out of intensive care – there’s no other therapy that’s worked on the same level.”

As simple as microbiome transplanting sounds, every microbiome is different and much remains unknown about what a “healthy microbiome” looks like.

For us to tap into its full potential and understand how it relates to human health, we need to fully understand what it’s made of and how its different species interact.

The Early Days of Microbiome Research

Studying the microbiome isn’t easy. The gut is a niche environment – a community of microbes feeding on a wide variety of metabolites in a largely anaerobic atmosphere. Some species depend on others for survival, so isolating and characterising strains in the lab is a challenging process.

In the past, colony picking, anaerobic culturing, and painstaking sequencing dominated the workflow. De Vos recalls how even simple tasks took extraordinary effort:

“If you focused, you could maybe pick 2,000 colonies an hour. Today it’s a completely different ball game.”

With Sanger sequencing the only option available, De Vos developed a human intestinal tract microarray chip to identify gut microbe composition based on ribosomal RNA fragments. High-throughput screening became possible with this technology, but proved costly.

“But every chip was about €200, so I needed an investment of hundreds of thousands of euros, which wasn’t easy to do as a standalone early-career researcher.”

Fortunately, De Vos’ persistence paid off, and his first EU-funded project on culturing gut microbes enabled the discovery of Akkermansia.

In recent years, however, high-throughput next-generation sequencing, proteomics, and metabolomics technologies circumvented the need for painstaking colony picking, generating a wealth of data on what our microbiomes contain.

Today’s Challenge: Discovering Functionality

High-throughput sequencing, proteomics, and metabolomics have opened unprecedented windows into the gut ecosystem. 

But, de Vos warns, having data isn’t the same as having answers:

“Big data are nice, but the problem is that we don’t know what most of the genes do – almost 30% remain unknown… You can’t predict physiology based on genomes alone.”

Large datasets can reveal compositions and correlations, but when it comes to developing therapies, these associations aren’t enough. Despite vast biobanks of bacteria, cultivating thousands of species of anaerobes and scaling functional studies is painstaking work.

The result is a field brimming with information but still short on answers. 

“What we’ve uncovered so far is just the tip of the iceberg. We know, more or less, what the top hundred bacteria are doing in our gut. And that’s already an enormous gain compared to 20 years ago.”

To truly unlock therapeutic use, we need to grasp the cause-and-effect relationships between microbial communities and their interaction with the human body.

Automation’s Role in Research

Automation and high-throughput technologies are now inseparable from modern microbiome research. Not only do they replace lengthy and laborious processes of the past, but they also open new avenues of inquiry.

De Vos points out that progress has been rapid. 

“It has started with DNA, which is so easy to work with – RNA too nowadays. When I started there was no DNA sequencing apart from the classical Maxam–Gilbert technique and later Sanger… Now you can outsource it and get your DNA sequences back. That has changed the landscape completely.”

Physical automation in the lab is also reshaping what’s possible. Colony-picking robots can now be adapted for anaerobic chambers, automated liquid handlers accelerate large-scale assays, and new sampling capsules allow researchers to capture gut contents directly.

Other tools accelerating microbiome research include flow cytometry and mass spectrometry. Miniaturised flow cytometry devices allow more accurate viability testing and enhanced cell cultivation from mixed populations, while mass spectrometry and MALDI-ToF allow more detailed examination of microbes and their components.

“You can use mass spectrometry for identifying proteins and small molecules. Some people also use it for identifying microbes, getting a signature part of the ribosomal RNA. MALDI is a bit faster because you can get your result quite quickly – it’s revolutionised analysis.”

Automation doesn’t solve every problem, but it consistently helps researchers scale their work and move closer to causality.

Now, the research bottleneck is less about data collection and increasingly about data interpretation.

Conclusion: Creativity Still Matters

As much as automation has transformed the microbiome field, machines don’t replace creativity.

“I think you have to use your brains and ask good scientific questions. The most important thing is to find angles other people haven’t and be creative in finding your niche.”

While automation tools aren’t a magic bullet for research problems, they offer scientists one critical thing they don’t get when they’re sweating in an anaerobic cabinet, labelling plates and focusing on hand-picking thousands of colonies:

The time to thoroughly examine the data and ask good scientific questions.

Creativity matters and, as de Vos’ career demonstrates, it is precisely those creative leaps – from looking at our microbiome as friends rather than foes – that change the way we think about life itself.

Understanding the microbiome means isolating and characterising difficult microbes on a grand scale. Doing all of this by hand is a long and tiring task, risking errors and unnecessary repetition – even for the best of us pipette-wielders.

Fortunately, our PIXL colony picking instruments can tackle plenty of these – including anaerobic environments, integrating with FACS cell sorting, and plating to MALDI-TOF targets.

Now, you can focus on interrogating data and asking amazing research questions! 

Willem Meindert de Vos | Emeritus Professor,
University of Helsinki