Periodic Reporting for period 1 - EvoBias (The role of developmental biases on the tempo of phenotypic evolution)
Período documentado: 2023-09-01 hasta 2025-08-31
Understanding the mechanisms that generate phenotypic variation and their impact on the course of evolution is a central challenge in modern biology. The genetic architecture of development may limit or bias the phenotypic spectrum obtained after new mutations, deemed developmental bias. These phenotypic biases, in turn, may influence evolutionary trajectories. Despite the recognition of the potential importance of developmental biases in constraining and pacing the evolutionary process, there is scarce direct empirical evidence for this potential role.
Here, I proposed to tackle this key question using the ecologically-relevant behaviour of avoidance to pathogens in the genetically-tractable organism Caenorhabditis elegans. I would establish the neurobehavioural system of C. elegans as a powerful study system to dissect the mechanisms of developmental bias across levels of biological organisation, encompassing cell fate decisions during the development of the nervous system to behaviour. This project aimed to directly test the role of developmental biases on the tempo of phenotypic evolution, and how they affect pathogen avoidance, a trait that is highly relevant for the health of biological systems. The project was structured in two main scientific objectives:
Objective 1: Quantify and uncover mechanisms of developmental biases in behaviour.
I would investigate the mutational variation-covariation properties of locomotor and avoidance behaviour in the adult C. elegans (i.e. the behaviour mutational co-variance/ developmental biases).
Objective 2: Test the role of developmental biases on the tempo of phenotypic evolution.
I would directly test the impact of developmental biases on the tempo of phenotypic evolution in an ecologically relevant scenario using experimental evolution.
Here, I proposed to tackle this key question using the ecologically-relevant behaviour of avoidance to pathogens in the genetically-tractable organism Caenorhabditis elegans. I would establish the neurobehavioural system of C. elegans as a powerful study system to dissect the mechanisms of developmental bias across levels of biological organisation, encompassing cell fate decisions during the development of the nervous system to behaviour. This project aimed to directly test the role of developmental biases on the tempo of phenotypic evolution, and how they affect pathogen avoidance, a trait that is highly relevant for the health of biological systems. The project was structured in two main scientific objectives:
Objective 1: Quantify and uncover mechanisms of developmental biases in behaviour.
I would investigate the mutational variation-covariation properties of locomotor and avoidance behaviour in the adult C. elegans (i.e. the behaviour mutational co-variance/ developmental biases).
Objective 2: Test the role of developmental biases on the tempo of phenotypic evolution.
I would directly test the impact of developmental biases on the tempo of phenotypic evolution in an ecologically relevant scenario using experimental evolution.
The WP1 aimed at investigating the mutational variation-covariation properties (mutational variance) of locomotor and avoidance behaviour in the adult C. elegans hermaphrodite. I established the model organism C. elegans in the host team by replicating the best practices and protocols from my previous host group in Paris (Félix lab, IBENS, France). For this, I was in close contact to my previous colleagues (lab manager, technicians, postdocs) to implement reliable growing conditions for the nematodes (tasks that I had not been involved beforehand): ordering reagents, best practices for media preparation and plate pouring, and avoiding recurrent bacterial contaminants. Additionally, I received dozens of bacterial and hundreds of nematode strains from my previous team, collaborators in Europe and Caenorhabditis Genetics Center.
Task1.1 aimed at generating random mutant lines (RMLs) containing the NeuroPAL transgene. For this, I decided to increase the mutation rate when generating the RMLs with chemical mutagenesis. For this, I conducted the mutagenesis with conditional RNAi-induction knockdown of the mismatch repair gene msh-2 to increase mutational rate in more than one-fold, thus increasing the power of the phenotypic mutational variance analysis. To achieve this, I first optimised RNAi feeding protocols by varying the bacterial density, and concentration and delivery method of IPTG (gene expression inducer of double-stranded RNAi). Afterwards, I mutagenised hundreds of thousands of nematodes with conditional msh-2 knockdown, isogenised independent lines by at least 10 generations of selfing hermaphrodites and cryopreserved them. During the process of isogenisation, only about 15% of the lines survived the process due to the increased mutational rate (in comparison to circa 50% without the msh-2 knockdown). Consequently, I had to conduct several rounds of mutagenesis to achieve the aimed target of hundreds of RMLs. At the end of task 1.1 I successfully generated and cryopreserved 144 RMLs.
Task 1.2 entailed to quantitatively characterise the locomotion behaviour of the RMLs to uncover locomotor biases. For this, I had to assemble from scratch a cost-effective advanced machine vision system that allows high-throughput automated multi-trait behavioural phenotyping of nematodes. This was achieved by contacting and initiating new collaborations with world-experts in C. elegans behaviour: Dr. Itamar Lev and Dipl.Ing. Lukas Hille from Zimmer lab (University of Vienna, Austria), and Prof. André Brown (MRC Laboratory of Medical Sciences, UK). Moreover, I sought the advice of a local colleague that has expertise in setting-up complex behavioural assays, Dr. Alexandre Leitão (senior postdoc of the Behavioural Neuroscience group at the Champalimaud Research Centre). After building a new state-of-the-art behavioural setup, I have optimised the recording conditions and analysis by varying the type of substrate, illumination and number of nematodes. We are currently performing a large-scale locomotory behaviour phenotyping in the same experimental design and in parallel to task 1.3. Hence, this task will be completed in the coming months.
For task 1.3 I had to select ecologically associated bacteria to C. elegans to perform the behavioural assays on pathogen avoidance. For this, I test the consistency of bacterial growth curves and pathogenicity in the C. elegans reference strain N2 and in the highly genetically divergent strain XZ1516. From these assays, we selected for further behavioural tests three bacteria that are highly pathogenic, intermediate, non-pathogenic, and the reference beneficial OP50. Next, I went to establish and optimise two new behavioural paradigms in pathogen avoidance in the host team: pre-contact avoidance, chemotaxis choice assay, and post-contact avoidance, lawn leaving assay. What’s more, while I was isogenising the RMLs of task 1.1 I noticed a broad developmental timing among the lines. Since the behavioural assays must be conducted at the same developmental stage and the assay plates with bacteria must be prepared on the day before, we conducted an experiment to describe the developmental timing of the RMLs. After having selected ecologically associated bacteria to C. elegans, established and optimised pathogen avoidance behavioural assays, and characterised developmental timing of the RMLs, we started the large-scale behaviour phenotyping of the hermaphrodite RMLs in a highly replicated and block design manner. We are halfway through the phenotyping, and we estimate to finish this 1.3 task in the next couple of months. Finally, in task 1.4 I was going to explore the mechanisms underlying behavioural biases. Since the completion of this task is dependent on the results of tasks 1.2 and 1.3 I will conduct and finalise it afterwards.
The WP 2 aimed at testing the role of developmental biases on the tempo of phenotypic evolution. To achieve this, I initially planned to use experimental evolution under ecologically relevant conditions to directly assess how developmental bias influences evolutionary rates. However, because WP2 is dependent on results currently being generated in WP1, work on WP2 has not yet commenced.
To mitigate this dependency and ensure progress toward testing the role of developmental bias on phenotypic evolutionary rates, I have completed a comparative phylogenetic analysis examining rates of phenotypic evolution across two nematode clades and their relationship to mutational variance.
Task1.1 aimed at generating random mutant lines (RMLs) containing the NeuroPAL transgene. For this, I decided to increase the mutation rate when generating the RMLs with chemical mutagenesis. For this, I conducted the mutagenesis with conditional RNAi-induction knockdown of the mismatch repair gene msh-2 to increase mutational rate in more than one-fold, thus increasing the power of the phenotypic mutational variance analysis. To achieve this, I first optimised RNAi feeding protocols by varying the bacterial density, and concentration and delivery method of IPTG (gene expression inducer of double-stranded RNAi). Afterwards, I mutagenised hundreds of thousands of nematodes with conditional msh-2 knockdown, isogenised independent lines by at least 10 generations of selfing hermaphrodites and cryopreserved them. During the process of isogenisation, only about 15% of the lines survived the process due to the increased mutational rate (in comparison to circa 50% without the msh-2 knockdown). Consequently, I had to conduct several rounds of mutagenesis to achieve the aimed target of hundreds of RMLs. At the end of task 1.1 I successfully generated and cryopreserved 144 RMLs.
Task 1.2 entailed to quantitatively characterise the locomotion behaviour of the RMLs to uncover locomotor biases. For this, I had to assemble from scratch a cost-effective advanced machine vision system that allows high-throughput automated multi-trait behavioural phenotyping of nematodes. This was achieved by contacting and initiating new collaborations with world-experts in C. elegans behaviour: Dr. Itamar Lev and Dipl.Ing. Lukas Hille from Zimmer lab (University of Vienna, Austria), and Prof. André Brown (MRC Laboratory of Medical Sciences, UK). Moreover, I sought the advice of a local colleague that has expertise in setting-up complex behavioural assays, Dr. Alexandre Leitão (senior postdoc of the Behavioural Neuroscience group at the Champalimaud Research Centre). After building a new state-of-the-art behavioural setup, I have optimised the recording conditions and analysis by varying the type of substrate, illumination and number of nematodes. We are currently performing a large-scale locomotory behaviour phenotyping in the same experimental design and in parallel to task 1.3. Hence, this task will be completed in the coming months.
For task 1.3 I had to select ecologically associated bacteria to C. elegans to perform the behavioural assays on pathogen avoidance. For this, I test the consistency of bacterial growth curves and pathogenicity in the C. elegans reference strain N2 and in the highly genetically divergent strain XZ1516. From these assays, we selected for further behavioural tests three bacteria that are highly pathogenic, intermediate, non-pathogenic, and the reference beneficial OP50. Next, I went to establish and optimise two new behavioural paradigms in pathogen avoidance in the host team: pre-contact avoidance, chemotaxis choice assay, and post-contact avoidance, lawn leaving assay. What’s more, while I was isogenising the RMLs of task 1.1 I noticed a broad developmental timing among the lines. Since the behavioural assays must be conducted at the same developmental stage and the assay plates with bacteria must be prepared on the day before, we conducted an experiment to describe the developmental timing of the RMLs. After having selected ecologically associated bacteria to C. elegans, established and optimised pathogen avoidance behavioural assays, and characterised developmental timing of the RMLs, we started the large-scale behaviour phenotyping of the hermaphrodite RMLs in a highly replicated and block design manner. We are halfway through the phenotyping, and we estimate to finish this 1.3 task in the next couple of months. Finally, in task 1.4 I was going to explore the mechanisms underlying behavioural biases. Since the completion of this task is dependent on the results of tasks 1.2 and 1.3 I will conduct and finalise it afterwards.
The WP 2 aimed at testing the role of developmental biases on the tempo of phenotypic evolution. To achieve this, I initially planned to use experimental evolution under ecologically relevant conditions to directly assess how developmental bias influences evolutionary rates. However, because WP2 is dependent on results currently being generated in WP1, work on WP2 has not yet commenced.
To mitigate this dependency and ensure progress toward testing the role of developmental bias on phenotypic evolutionary rates, I have completed a comparative phylogenetic analysis examining rates of phenotypic evolution across two nematode clades and their relationship to mutational variance.
My main scientific contribution during this project was to show that the evolution of developmental biases is aligned and explains the divergent patterns of phenotypic rates. This has important implications for the role of development in shaping the tempo of phenotypic evolution. These results were published in the high impact journal PNAS, with me as the first and co-corresponding author.
Additionally, we test an open hypothesis on the field behaviour ecology: that avoidance to pathogens and resistance are costly survival strategies and should therefore covary negatively. Together with two Master students, I tested the correlation between bacterial pathogenicity and host pathogen avoidance across twelve divergent wild strain of C. elegans exposed to nine ecologically associated bacteria. We obtained the unexpected and interesting result of no correlation between avoidance and resistance, challenging theoretical models of host-pathogen evolution. A manuscript reporting these findings is currently in preparation.
Additionally, we test an open hypothesis on the field behaviour ecology: that avoidance to pathogens and resistance are costly survival strategies and should therefore covary negatively. Together with two Master students, I tested the correlation between bacterial pathogenicity and host pathogen avoidance across twelve divergent wild strain of C. elegans exposed to nine ecologically associated bacteria. We obtained the unexpected and interesting result of no correlation between avoidance and resistance, challenging theoretical models of host-pathogen evolution. A manuscript reporting these findings is currently in preparation.