Shear stress sensing in C. elegans

Abstract

Shear stress sensing represents a vital mode of mechanosensation.¹ Previous efforts have mainly focused on characterizing how various cell types-for example, vascular endothelial cells-sense shear stress arising from fluid flow within the animal body.^1,2 How animals sense shear stress derived from their external environment, however, is not well understood. Here, using C. elegans as a model, we show that external fluid flow triggers behavioral responses in C. elegans, facilitating their navigation of the environment during swimming. Such behavioral responses primarily result from shear stress generated by fluid flow. The sensory neurons AWC, ASH, and ASER are the major shear stress-sensitive neurons, among which AWC shows the most robust response to shear stress and is required for shear stress-induced behavior. Mechanistically, shear stress signals are transduced by G protein signaling in AWC, with cGMP as the second messenger, culminating in the opening of cGMP-sensitive cyclic nucleotide-gated (CNG) channels and neuronal excitation. These studies demonstrate that C. elegans senses and responds to shear stress and characterize the underlying neural and molecular mechanisms. Our work helps establish C. elegans as a genetic model for studying shear stress sensing.

Keywords: atherosclerosis; flow sensing; mechanical; mechanobiology; mechanosensory; mechanotransduction; shear force; touch.

Figure 1.. Shear stress-triggered behavioral responses and identification of AWC, ASH and ASER as the major shear-stress sensitive neurons.

(A) Snapshot images showing that fluid flow directed at the head of the animal promotes turning in swimming. Shear stress at the opening of the perfusion pipette: 755.4 dyn/cm2. As the liquid flowing out of the perfusion pipette was no longer in laminar flow when reaching the nose, it is difficult to estimate the shear stress level at the nose tip; but it should be substantially lower than that at the opening of the perfusion pipette. (B) Bar graph quantifying the turns per minute in swimming animals. Shear stress was applied for 3 min and found promote turning in swimming animals. n≥10. Error bars: SEM. ***p<0.0001, *p<0.05 (t-test). (C) Schematic depicting the microfluidic chip system used for calcium imaging. The animal was loaded in the animal channel, and its nose tip was exposed to fluid flow. (D) Heat map showing calcium responses to shear stress in all the known sensory neurons and AMsh glial cell in the head. Shown are calcium imaging traces from individual neurons/cells with each row representing one neuron/cell. Shear stress at the nose tip: 24.1 dyn/cm2. (E) Average calcium imaging traces of sensory neurons and AMsh glial cell in (D). Only AWC^on, AWC^off, ASH and ASER showed notable responses to shear stress. (F) Bar graph summarizing the data from AWC^on, AWC^off, ASH and ASER neurons in (C). n⩾10. Error bars: SEM. See also Figures S1 and S2A–D.

Figure 2.. AWC neurons are required for shear stress-induced behavior.

(A-B) Shear stress-induced behavioral responses are dependent on AWC neurons rather than ASH or ASER neurons. AWC neurons were ablated using a caspase transgene expressed in AWC. ASH neurons were inactivated using a transgene expressing tetanus toxin in ASH. che-1(ot75) mutant animals, in which ASE neurons failed to differentiate, were used to test the role of ASER. Shear stress at the opening of the perfusion pipette: 755.4 dyn/cm2. (A) Behavioral traces showing the number of turns per min over time. (B) Bar graph summarizing the data in (A). For each genotype, peak turning frequency under shear stress (1 min) was compared with the baseline (0 min). n≥10. Error bars: SEM. **p<0.005 (ANOVA with Tukey test), ns: not significant. (C-D) CuCl₂-induced behavioral responses are normal in AWC-ablated animals. (C) Behavioral traces showing the number of turns per min over time. (D) Bar graph summarizing the data in (C). For each genotype, peak turning frequency in response to CuCl₂ (1 min) was compared with the baseline (0 min). n≥10. Error bars: SEM. **p<0.005 (ANOVA with Tukey test), ns: not significant. (E-F) Shear stress-evoked behavioral responses depend on the CNG channel subunit TAX-4 but not TRP channels. (A) Shear stress failed to induce behavioral responses in tax-4(ok3771) mutant animals, while osm-9(ky10) ocr-2(ak47); trpa-1(ok999) mutant animals showed normal responses to shear stress. The tax-4 mutant phenotype was rescued by a transgene expressing wild-type tax-4 gene in AWC neurons. (F) Bar graph summarizing the data in (E). For each genotype, peak turning frequency under shear stress (1 min) was compared with the baseline (0 min). n≥10. Error bars: SEM. *p<0.05, **p<0.005 (ANOVA with Tukey test), ns: not significant. See also Figure S2E–G.

Figure 3.. Characterization of shear stress responses in AWC^on neuron.

(A) AWC is activated by shear stress in a dose-dependent manner. Shear stress produced by fluid flow at the flow speed of 100 μl/min was: 24.1 dyn/cm². The shear stress level is linear with the flow speed level, and its values at other flow speed levels were not listed. (B) Bar graph summarizing the data in (A). n≥10. Error bars: SEM. (C) Shear stress responses in AWC are normal in unc-13(e51) and unc-31(m68) mutants. Shear stress at the nose tip: 24.1 dyn/cm². (D) Bar graph summarizing the data in (C). n≥10. Error bars: SEM. p>0.05 (ANOVA with Dunnett test), ns: not significant. (E) AWC can be repetitively activated with modest desensitization. Shear stress: 24.1 dyn/cm². (F) Bar graph summarizing the data in (E). n≥10. Error bars: SEM. (G) Schematic of the electrophysiology assay. Bath solution was perfused towards the nose tip to generate shear stress. Pressure applied to the back of the perfusion pipette: 100 mmHg. (H) Sample traces of shear stress-induced current in AWC. The neuron can be activated repetitively. Voltage: −60mV. (I) Bar graph summarizing the data in (H). n≥8. Error bars: SEM. (J) IAA (isoamyl alcohol) induces an outward current in AWC, and removal of IAA then triggers an inward current. Voltage: −60mV. (K) Bar graph summarizing the data in (J). n≥6. Error bars: SEM. See also Figure S3.

Figure 4.. Shear stress transduction in AWC^on neuron is a G protein-mediated process.

(A) Shear stress fails to activate AWC in tax-4(p678) mutant animals. Shear stress applied to the nose: 24.1 dyn/cm². (B) Bar graph summarizing the data in (A). n≥10. Error bars: SEM. ***p<0.0001 (t-test). (C) Sample traces of shear stress-evoked current in AWC of wild-type (WT), tax-4 mutant, and WT and tax-4 mutant in the presence of the guanylate cyclase inhibitor LY83583 (100 μM) or cGMP (5 mM). LY83583 was diluted in bath solution. cGMP was diluted in pipette solution and dialyzed into AWC through the recording pipette. Voltage: −60mV. (D) Bar graph summarizing the data in (C). n≥8. Error bars: SEM. ***p<0.0001 (ANOVA with Dunnett test). (E) Shear stress-evoked current in AWC can be blocked by GDPβS. 1 mM GDPβS was dialyzed into AWC through the recording pipette. Voltage: −60 mV. (F) Bar graph summarizing the data in (E). n≥12. Error bars: SEM. ***p<0.0001 (t-test). (G) Schematic model of shear stress transduction pathway in AWC neurons. See also Figure S4.