"Amicus meus, inimicus inimici mei". "The enemy of my enemy is my friend".
As an agro-ecology Bachelor's student, I develop high awareness for the many dangers (agricultural) plants face. Pathogens are widespread, and it is just a matter of time until one will strike your field. Three years of study taught me that preventing diseases in crops is almost unnatural. It is so unlikely to prevent diseases in a field year after year, that the term 'disease management' is often used to describe the goal 1.
Pesticides. Crop rotation. Biological pest control. Pruning. Prediction of risk by climate monitoring. These practices, among others, are required to maintain our global food supply. Given this educational background I have, I was surprised to find out that wild plants can actually live in harmony with their pathogens! But I will get to that in a bit.
Contrary to plants of agricultural ecosystems, wild plants rarely present a serious disease breakout, mainly due to the high diversity of the ecosystem they live in 2. In short, high-diversity means that not all plants can be attacked or even colonized by the same pathogen, thus at one point there will be a barrier. Whereas this scenario portrays how disease outbreaks are stopped from an epidemiological point of view, it doesn't tell much about the effect of pathogens on individual wild plants.
When I arrived at the Max Planck Institute for Developmental Biology (now Max Planck Institute for Biology Tübingen) for my doctoral research, my new colleagues had just finished a big-scale study of Pseudomonas isolated from wild Arabidopsis thaliana, concluding that one lineage dominated local plant populations near Tübingen in Southwest Germany 3. When tested in the lab, this lineage was highly pathogenic, despite being isolated from seemingly healthy plants. This riddle puzzled our team. It means that these wild plants can capacitate a pathogenic lineage while maintaining their health. What causes this?
Given that the Pseudomonas collection yielded more than 1500 strains, and some are non-pathogenic, we speculated that other Pseudomonas strains can suppress the pathogenic lineage, preventing disease. We based this speculation on the wide literature about protective Pseudomonas, as well as on knowledge about competitive behaviours of phylogenetically-related bacteria (bacteria sharing similar metabolic pathways often compete for the same resources). Another less scientific but practical reason to use other Pseudomonas is because we already had them... Of course, there could be other causes too.
Examining co-infections of individual commensal strains with the pathogenic lineage is one way to study the protective ability of non-pathogenic Pseudomonas. Nonetheless, we decided to thicken the plot, aiming to imitate a more realistic scenario; In the wild, most plants were simultaneously colonized by a few strains. Some are pathogenic, some are not. Therefore, we decided to infect plants with communities of commensal strains, pathogenic strains and a mixture of both. Since many of the strains we chose for the study were highly similar, they could not be differentiated according to their 16S rDNA sequences - the current mainstream practice. We hence developed a barcoding system, enabling us to differentiate closely-related isolates.
The idea of moving away from simple and highly controlled experiments towards complex and more realistic ones became increasingly attractive to us, considering that our phenomenon of interest was observed in the field (and as is common in ecology and biology, simple phenomena may reflect highly complex realities). Therefore, we decided to test the bacterial effect on several Arabidopsis thaliana genotypes that were sampled from the same region as the bacteria, looking for host genotype-specific effects. Also we decided to grow plants in soil rather than in plates (with agar-based media). Lastly, we chose to conduct experiments in non-sterile rooms, and to infect plants with the synthetic communities using an airbrush, imitating the movement of bacteria via wind and rain. As we found out, these decisions were not as trivial as they first seemed when discussed. The bad news - we ended up with very large experiments that required many more than two hands. The good news - working together is fun. We got eleven of our lab members volunteering to help, working in shifts. One of these was none other than the head of the lab Detlef Weigel, enjoying getting back to his hands-on days (picture attached). What was first an intimidating size of experiments converted into quality time. The grand finale was pizza and beers, taken on the rooftop, celebrating our hopefully-not-too-bad work. (Did I mention that we replicated every experiment? And that there was this one time in which I confused the treatments? Well, there was a lot of fun time together, and accordingly pizza toppings and beers). Luckily, our efforts paid-off not just socially, but scientifically too.
Our results portray in detail how a consortium of commensal Pseudomonas protect the plant via both direct and host-mediated measures. Importantly, we found that commensal Pseudomonas elicited a plant immune response, leading to a specific suppression of only the pathogenic Pseudomonas lineage. This is not a trivial outcome, as all infecting bacteria were from the same genus, and despite their high relatedness the plant immune response specifically affected one pathogenic line. Moreover, the commensal consortium comprised a diverse set of commensal species, though none were suppressed by the plant, further demonstrating how taxon-specific the host-mediated suppression was.
We found two types of outcomes: general and genotype-specific, both from the host and bacterial perspective. For example, we were surprised to find out that the commensal consortium failed to protect one specific host genotype. We pinpointed one pathogenic strain out of the pathogenic consortium as the causal agent. Removal of this strain from the pathogenic consortium re-enabled the protective ability of commensals. These dynamics demonstrate that the consequences of host-microbe-microbe interactions are highly dependent on fine grain genetic differences. Another important aspect of our work is the characterization of a collective protective effect. Rather than pinpointing specific individuals as the protectors, we found a synergistic effect of the collective.
1. He, Dun-Chun, Jia-Sui Zhan, and Lian-Hui Xie. 2016. “Problems, Challenges and Future of Plant Disease Management: From an Ecological Point of View.” Journal of Integrative Agriculture 15 (4): 705–15.
2. Karasov, Talia L., Juliana Almario, Claudia Friedemann, Wei Ding, Michael Giolai, Darren Heavens, Sonja Kersten, et al. 2018. “Arabidopsis Thaliana and Pseudomonas Pathogens Exhibit Stable Associations over Evolutionary Timescales.” Cell Host & Microbe 24 (1): 168–79.e4.
3. Karasov, Talia L., Gautam Shirsekar, Rebecca Schwab, and Detlef Weigel. 2020. “What Natural Variation Can Teach Us about Resistance Durability.” Current Opinion in Plant Biology 56 (August): 89–98.
Poster image: 'the dance of good and evil', Curtis Verdun, oil painting on canvas.