Groundbreaking Research: Imaging Mycorrhizal Fungi Network Growth in Real-time. A Five Year Study.

Scientists from AMOLF, Vrije Universiteit and SPUN built a robot in order to image  mycorrhizal fungi network growth in real-time. This allowed them to build precise maps of the networks, and track the nutrient-carrying flows inside.

‍

After five years of research they published their initial findings in NATURE.

The paper is about how these fungi build and operate their supply chains for underground nutrient exchange/trade.

‍

Executive  Summary

This research is about mapping mycorrhizal fungal network growth in real time. How the network grows, and how the nutrients, including carbon flow inside are regulated by the network itself.

NOVEL RESEARCH: SHAPE OF THE COMPLETE NETWORK

To be able to link the flows to the architecture is completely new. Recording and imaging the shape of the complete network map has also never been done before.

OVERVIEW:

The research program we're building on mycorrhizal fungi is about trying to understand how these very unusual microbes move and exchange nutrients with plants underground. They are present in soils worldwide across all terrestrial environments, so it’s important to understand how they function as something like a connective tissue for ecosystems. There is compelling evidence that these fungi can change the way they move/exchange nutrients depending on their plant partners as well as the resource environment. But we know very little about how they do that. We want to figure out how they manage to control their behavior, despite having no brain, et cetera.

INFORMATION PROCESSING NETWORK

Compared to something like a bacterial cell, these fungi are very complex. And what's particularly complex about them is the topology of these networks. In effect, what we are doing is making a map of an urban network. We think of the fungal network as a road network, which allows us to ask questions about distribution and transport.

DEPARTURE POINT and CHALLENGE

Eventually, we came to the realization that people just haven't done the simplest thing, which is just to let them grow and watch what they do. And when we realized that's what we wanted to do, there was one major challenge, which was to grow them in the lab.

NO HYPOTHESIS - EXPECTATIONS - PIONEERING RESEARCH

Nobody's really managed to look at how these fungi build their networks on this scale. People have taken pictures of the networks at different times, but nobody has managed to make a movie. This has been one of those truly delightful moments in science. Sometimes when you're diving into something completely new, it's actually more productive to not have a hypothesis.

WHY THE ROBOT WAS BUILT

What the robot really does is to crawl around above all the mycorrhizal networks that are growing in petri dishes. The robot that we use in the Nature paper can accomplish that with forty different petri dishes. Forty different networks at the same time. And it's running now twenty-four / seven.

‍

WHAT THE ROBOT ALLOWS US TO DO

Scientists labelled and monitored a half million new nodes and mapped the network expansion in real-time.

PIONEERING DESIGN PRINCIPLES

In the "real world" mycorrhizal fungi networks live in soil, which is a particularly difficult environment for observation. This explains why to really do imaging in a systematic way just hasn't been done very much. So we really needed to build this robot in order to see this network grow. That would allow us to image, map, and create models for studying the network.
‍

SHAPE OF THE FUNGAL NETWORK - SHAPE AND ACTIVITY: TRAFFIC INFORMATION

The robot we built hovers over the petri dishes, taking lots of zoomed in images. To construct maps from those images, we created a data analysis pipeline. It stitches together all those images and extracts the fungal network’s topology and morphology, which makes up the map. But that is just the structure. If you really want to understand how this network is operating - how it's moving nutrients underground - then you need to be able to zoom in at greater detail. Basically, we needed to look inside every tube of the network and study what we have come to think of as "traffic information".

NOVEL ASPECT OF RESEARCH - ROBOTIC IMAGING

To be able to image in detail what's going on inside these networks together with the complete network map is what's really new here. This was only possible because we built this imaging robot - without it, we just wouldn’t be able to collect the necessary data fast enough. You can think of what the robot does as something like satellite imagery - it takes relatively low magnification snapshots as it moves over the fungal terrain. To see the flows inside, we zoom in by switching to higher magnification at specific points of interest within the network map. This part is more like helicopters that hover over cities to gather traffic information. In this way, the robot allows us to really study how the dynamic traffic patterns are related to the structure of the network as a whole.

VISUALS IMAGES - SURPRISES

One of the things that got me really excited to work on mycorrhizal fungi is that when you zoom in and actually look inside these pipes, you see fluid flowing in both directions  - at the same time! And that is just very counterintuitive. Certainly to a physicist, but I think really for anyone (when is the last time you saw a river flowing in two directions at once?). According to what we know about fluid flows in narrow vessels, the usual laws just tell you that the direction of fluid movement is determined just by the pressure difference across the pipe. So you don't get things flowing in both directions at the same time, and that’s why we were very surprised to see that in these fungi. I was immediately captivated, because you could just see, by eye, that these organisms have invented something remarkable - a strange kind of physics - in order to make this possible. Sometimes when you see something, you just immediately know it's special. This was definitely one of those cases.

BI-DIRECTIONAL FLOW - CARBON - EXPECTED AND UNEXPECTED FINDINGS

And it makes sense that the fluid should flow in both directions. Just think about what these organisms need to do. They get all of their carbon from plants. But they don't get it for free, they have to "pay" for it with other nutrient resources (primarily phosphorus and nitrogen) that they forage for by growing their networks outward from the root and into the soil. To fuel that growth, they constantly have to move the carbon away from the root. At the same time, they also need to bring back the foraged soil nutrients to pay for more carbon. Hence the two-way traffic. Things need to be moving both ways all of the time.

‍

Another really fascinating thing we noticed is that the flows through the network are extremely dynamic. It's one thing for fluid to flow in both directions at the same time in the same pipe. What we also noticed pretty quickly is that if you observe different parts of the network at different times, the flow behaviors are very diverse, and also change over time. Not only do they flow in both directions at the same time, those streams can also suddenly switch direction, and start flowing the other way.

‍

NETWORK ARCHITECTURE - AUTOMATED COMPUTE PIPELINE FOR EXTRACTION

A comparable challenge to building the robot has been to automate the processing of data, which the robot is constantly producing en masse. We needed to develop a computer pipeline that extracts the network architecture without human intervention. That allowed our team to study the resulting maps to decide which specific network locations to zoom into. 

RESEARCH QUESTION & DESIGN - ACCOMPLISHMENTS

In terms of trade, the network the mycorrhizal fungi grow outside of the plant root is like a supply chain. The question we ask in this paper is, how does the fungus build that supply chain? And once it's built, how does it operate it? The robot allowed us to image the complete network of the supply chain. It's one thing to just be able to see one part of a network, but when you have the map of the whole thing, you can start to ask different kinds of questions.

DECISION-MAKING - CHANGING PARTNERS - COGNITION?

At the scale of the whole network, these fungi are known to change their nutrient trade behaviors depending on the specific plant partners and environmental conditions. So in some real sense, you could say the network is making a decision to do X or Y. But if you think about how organisms make decisions, usually, the ones that we tend to think of are ones with brains. Those organisms make decisions using a nervous system. The fungus doesn’t have that at it’s disposal. We know very little about the mechanisms for information processing / decision making in a fungus.

MULTI-DISCIPLINARY TEAM

The research paper was produced by a multidisciplinary team made up of four institutions and twenty-eight authors. The main classes of people were physicists, biologists, engineers, and computer scientists. The two heroes that led the paper are the first authors Loreto Oyarte Galvez, who built the robot together with an excellent team of AMOLF engineers, and Corentin Bisot, who developed the compute pipeline and also much of the theory/modeling. The project started as a collaboration between my own group at AMOLF, Toby Kier's group at Vrije Universiteit, and the group of Howard Stone at Princeton. I am a physicist who works on living things, a biophysicist. Toby is an evolutionary biologist and ecologist, and Howard Stone is an expert in fluid mechanics. The team also benefited from participation of computer scientist Christophe Godin at ENS Lyon in France, an expert in plant morphogenesis.

‍

THREE MAIN FINDINGS

1. The first thing we learned is that the way the fungi build these networks seem to emphasize long term gains over short term benefit. This is not what most microbes tend to do, when they find a good spot, they just grow as much as they can until they run out of resources. In contrast, these fungi, after attaching to a plant root, spread out like a wave at a very low density. What they seem to be prioritize is exploration for other plants. Perhaps thoes that might offer them a better trade deal. They appear to be looking for new opportunities for resource extraction and trade, rather than immediate growth. So the fungi are solving some kind of  foraging problem, but a peculiar one. It’s as if they’re foraging for trade opportunities, rather than for food per se. 

2. The second finding is that the networks basically always form two-way traffic systems inside their tubes. I mentioned that's something we saw when we first started studying the images. Every single filament in the network is a tube. And in nearly every tube where we saw movement, there were flows in both directions. It's very interesting, and kind of makes sense. If you think about human road traffic, two-way roads allow for more efficient routing than one way systems. But they are also more susceptible to congestion. But for fungal traffic, the flows consist of fluids, not cars.  Fluids don’t really suffer from congestion in the way that car traffic does. So two-way fluid flows are really an ingenious solution for efficient nutrient traffic through these networks.

3. The third finding is that the shape of the fungal network is determined by just a few very simple and elegant rules. When you look at enough of these networks, you begin to wonder, how do the fungi manage to control the shape of their networks? It turns out the rules they follow require only local information. To make a decision about what to do with building here, you just need information about what’s happening nearby, not about what’s happening across the entire network. In other words, the process doesn't really need a centralized control system or a brain in order to control the shape of the network. One good example is that if a growing tip bumps into another growing tip, another edge of the network, the two just fuse.

THE MOST SURPRISING DISCOVERIES AND OUTCOME ACCOMPLISHED

We didn't expect to find this extreme self-regulation of growth in favor of exploration and trade. And when we found it, we didn’t expect that we could identify the rules of branching and fusion that explain that self-regulation. Nor was it obvious from the beginning that we would really be able to map the full network at every moment in time as it grew. So that was a really exciting moment, when we realized that it was actually going to work. These data are making us realize how different these networks are from typical microbial colonies, which usually try to fill the available space by growth. AM fungi seem to grow only as densely as they need to connect their trade partners and resource extraction sites. It’s a challenge to mathematically understand these structures. We’re excited that we were able to develop a really simple mathematical model that describes the process of how these networks are generated dynamically, in the wake of wave-like patterns of growth. In a very real sense, the dynamics of those waves also describe how plant-derived carbon moves through soil environments.

‍