Neuro-genetic plasticity of Caenorhabditis elegans behavioral thermal tolerance

Abstract

Background: Animal responses to thermal stimuli involve intricate contributions of genetics, neurobiology and physiology, with temperature variation providing a pervasive environmental factor for natural selection. Thermal behavior thus exemplifies a dynamic trait that requires non-trivial phenotypic summaries to appropriately capture the trait in response to a changing environment. To characterize the deterministic and plastic components of thermal responses, we developed a novel micro-droplet assay of nematode behavior that permits information-dense summaries of dynamic behavioral phenotypes as reaction norms in response to increasing temperature (thermal tolerance curves, TTC).

Results: We found that C. elegans TTCs shift predictably with rearing conditions and developmental stage, with significant differences between distinct wildtype genetic backgrounds. Moreover, after screening TTCs for 58 C. elegans genetic mutant strains, we determined that genes affecting thermosensation, including cmk-1 and tax-4, potentially play important roles in the behavioral control of locomotion at high temperature, implicating neural decision-making in TTC shape rather than just generalized physiological limits. However, expression of the transient receptor potential ion channel TRPA-1 in the nervous system is not sufficient to rescue rearing-dependent plasticity in TTCs conferred by normal expression of this gene, indicating instead a role for intestinal signaling involving TRPA-1 in the adaptive plasticity of thermal performance.

Conclusions: These results implicate nervous system and non-nervous system contributions to behavior, in addition to basic cellular physiology, as key mediators of evolutionary responses to selection from temperature variation in nature.

Keywords: C. elegans behavior; Computational ethology; Thermal ecology.

Fig. 1

Thermal tolerance curve for wildtype C. elegans. a C. elegans thermal tolerance curve. Locomotion index (LI1) with temperature increments from 21 to 41 °C. Wildtype N2 adult hermaphrodite animals were reared at 23 °C. 54 worms and ≥ 35 animals retained in calculations at each temperature step. b C. elegans locomotion index profiles in response to acute high temperature exposure from baseline 23 °C (black arrow dotted line, 80 s exposure). Error bars indicate ± SEM. 54 individuals tested per series; ≥ 34 retained in calculations at each step. c Locomotion index over time at constant hold temperature after incrementing up from an initial 25 °C (upper half of panel). Hold temperature shown on the bottom half of the panel, with hold temperature of 40 °C comparable to the standard TTC assay; 18 worms tested per series, ≥ 9 worms included in LI calculations at each step

Fig. 2

Effect of rearing temperature and developmental stage on C. elegans TTC. a Rearing conditions at both high and low thermal extremes shift wildtype N2 C. elegans thermal performance curves relative to rearing at intermediate thermal conditions. 36–114 animals tested per thermal rearing condition, ≥ 16 individuals included in calculations at each assay step. b TTCs for adult and larval developmental stages. 9–27 worms tested per treatment, ≥ 5 worms included in calculations at each assay step. Error bars indicate ± SEM

Fig. 3

Effect of mutations on C. elegans TTCs. a Thermal reaction norms of swimming behavior for 58 C. elegans genetic mutant strains (gray and red lines) and two wildtype strains (black line = N2, dashed line = CB4856). All worms reared at 23 °C; 17–36 individuals included in calculations at each assay step for each strain (Table 1). TTCs for subsets of mutant strains with sensory disruptions from (a) are shown for strains with similar or higher paralysis threshold temperatures than wild-type (b) and strains with lower paralysis temperatures than wild-type (c)

Fig. 4

Comparison of wildtype and trpa-1 mutants over different rearing temperatures. a The sensitivity of the locomotion index TTC to rearing temperature observed for wildtype N2 C. elegans (dashed lines) is reduced for trpa-1 mutant animals (dark lines), which show similar TTCs regardless of rearing conditions. Error bars indicate ± SEM; 54 animals tested per treatment, ≥ 44 individuals included in calculations at each assay step. b Comparison of inflection point estimates for locomotion index TTC curves for wildtype N2, trpa-1 and ttx-1 mutants, and trpa-1 expression-rescue strains (see Table 2 and Additional file 1: Fig. S3). Similar inflection point values for different rearing conditions of a given genotype indicate low TTC plasticity; distinct values for different rearing conditions indicate high TTC plasticity. Genomic and gut expression-rescue constructs for trpa-1 also rescue TTC plasticity, unlike the neural expression-rescue construct. Values for wildtype N2 and trpa-1(ok999) correspond to TTCs shown in a. Error bars indicate 95% CI from a three-parameter logistic function fit to locomotion index (LI1) values for each strain and rearing temperature; 45–54 animals tested per treatment, ≥ 31 individuals included in calculations at each assay step