Olfactory combinatorial coding supports risk–reward decision-making in Caenorhabditis elegans
- Department of Molecular, Cellular, and Developmental Biology, The University of Toledo, United States
- Department of Neurosciences and Psychiatry, The University of Toledo, United States
Figures
Figure 1
1-oct chemotaxis in microfluidic arenas.
(A) Microfluidics arenas were configured with two zones of equal area, one containing a 1-oct solution and the other containing buffer alone (1-oct zone contains 0.01% xylene cyanol dye for visualization here; assays are run without dye). (B, C) Typical distribution of worms when first exposed to 1-oct (B), and after 15 min of 2.2 mM 1-oct exposure (C). (D) Time-course of 1-oct chemotaxis for the trial shown in (B, C). (E) Calculation of chemotaxis index, averaging worm counts every 2 minutes from 10 to 20 minutes after introduction of 1-oct solution.
Figure 2
Network recording of 1-oct (2.2 mM) response.
(A) Representative image of a NeuroPAL worm immobilized in a microfluidics device visualized using the mTagBFP2, CyOFP, and mNeptune2.5 reporters for neuronal identification (Yemini et al., 2021). (B, C) Pan-neuronal nGCaMP6s signal for the region indicated by the box in (A) during 1-oct stimulation, showing 1-oct offset (B) and onset (C). Specific time points in the recording for (B) and (C) are indicated by the black triangles in (D). Selected sensory neurons and interneurons are indicated by dashed white circles. (D) Heat map of activity profiles for 56 identifiable neurons. Raw (i.e., non-normalized) fluorescence values shown. Stimulation time-course shown on top row (white is buffer, orange is 1-oct). Green and orange vertical bars at left indicate first-layer and locomotory command neurons (+RIM), respectively.
Figure 3
Neuronal activity correlations.
(A–C) Neuronal activity patterns from worm shown in Figure 2. (A) Stochastic activity state transitions of locomotory command and associated interneurons, with AIY and AVB showing opposite activity patterns to AVA, AVD, AVE, RIM, AIB, and AIZ. (B) Sensory neuron activity patterns. (C) Selected neuron pairs highlighting the anti-correlated activity patterns of AWC and ASH (top two traces; averaged AWC/ASH data from 21 worms shown to highlight consistency of anti-correlation between these two neurons), and the correlated activity patterns of AWC with AIA (middle) and SAA (bottom). (D, E) Activity correlations for 12 selected neurons based on data from five worms. (D) Graph representation of activity correlations before stimulus (minutes 0–6, left) and during stimulus (minutes 6–12, right). Neurons pairs with the strongest positive correlations are closest to one another and connected by the darkest purple lines; neurons with the strongest negative correlations are furthest apart and connected by the darkest green lines. (E) Relationships from (D) presented as a correlation matrix using the same color code. Error bars are SEM.
Figure 4
Chemotactic outcome is influenced by the balance of attractive and repulsive sensory inputs.
(A–C) ASH ON responses (A, C) are concentration dependent, while AWC ON and OFF responses (B, C) are not, over a 10-fold concentration range from 0.22 to 2.2 mM. Representative single traces (A, B, left panels), averaged traces (A, B, right panels), and concentration-response curves (C) shown; ***p<0.001, ns p>0.05, paired ANOVA, n=5. (D) 1-oct chemotaxis transitions from positive (attractive) to negative (repulsive) with increasing concentration; intermediate concentrations (1.21 and 1.54 mM) result in little chemotaxis; *p<0.05 compared to 2.2 mM, ANOVA. (E) tax-4 mutants are repelled by 1-oct at all tested concentrations (*p<0.05; **p<0.01; ***p<0.005; ****p<0.0005; ns p>0.05, ANOVA). osm-9 mutants do not differ from wild type. (F, G) chemogenetic inactivation of ASH, AWC: ASH inactivation increases the chemotaxis index, suggesting reduced repulsive drive (F); AWC inactivation decreases the chemotaxis index, suggesting reduced attractive drive (G) *p<0.05 t-test. n=5 for panels (D–G). Error bars are SEM.
Figure 5
Locomotory reversals and speed modulation stimulated by 1-oct.
(A) Example of a single worm track traced by the automated tracking software. (B) Worms execute reversals probabilistically at the 1-oct-buffer interface dependent on TAX-4 and OSM-9 signaling cascades. (C) Reversal probabilities at 1-oct-buffer interface for wild-type, tax-4, and osm-9 (see panel B for summary and explanation of terms). Left panel: movement from buffer zone to 1-oct zone (‘entries’). Right panel: movement from 1-oct zone to buffer zone (‘exits’). (D) Reversal analysis of trials shown in Figure 4F and G. Bars color coded as in (C). Number of encounters scored shown in bars, from multiple independent trials (three in C, five in D). (E) Modulation of locomotory activity. Distance traveled in 2 minutes prior to, during, and after flooding the entire chip with 1-oct solution (59 tracks over three independent trials). (F) Simulation of chemotaxis outcome based on reversal probabilities (from C) and locomotory activity modulation (from E). Points shown are averages of 10 simulation trials with SEM error bars. ***, **** in (C, D) indicate p<0.001, 0.0001 Fisher’s exact test. *, ** in (E) indicate p<0.05, 0.01, ANOVA (Tukey’s multiple comparison test).
Figure 6
Entrainment of neuronal activity by 1-oct stimulation.
(A, B) Validation of the entrainment index measurement using ASH (A) and AWC (B), showing very strong positive and very strong negative entrainment by 1-oct application, as expected. Gray dots indicate each randomized sum, while red and blue dots indicate the stimulus-specific sums of ASH and AWC, respectively (Average of 21 independent trials for ASH, AWC; see ‘Materials and methods’). (C–F) Superimposed ASH and AVA activity patterns after stimulation by 0.22, 0.66, and 2.2 mM 1-oct in wild type (C, D) and tax-4 (E, F). Single worm traces (C, E) and averaged traces are shown (D, F, n ranges from 6 to 21 worms). (G) AIB traces (averaged); purple is unprocessed AIB signal, green is AIB minus AVA signal. (H) Entrainment indices for all recordings in (C) through (G), with gray dots indicating randomized EIs and the bold dots representing the stimulus-specific EIs for AVA (black), unprocessed AIB (purple), and AVA-subtracted AIB (green). p values in (A), (B), and (H) represent the probability that a randomized EI would be as great or greater than the stimulus-selected sum, with a significance cut-off of p<0.05. Error bars are SEM.
Figure 7
Balanced repulsion and attraction support context-dependent modulation of sensory responses.
(A, B) Combinatorial coding-based mechanism for modulating sensory responses. (A) At lower 1-oct concentrations, the attractive afferent pathway is relatively strong, resulting in attraction. (B) Addition of IAA reduces AWC activity, weakening the attractive pathway relative to the repulsive pathway, and reversing the 1-oct valence. (C) Addition of 0.92 mM IAA (10-4 dilution) reverses the valence of 1-oct (0.66 mM) from attractive to repulsive (number of independent trials shown in bars). (D) IAA biases reversal behavior (at the 1-oct/buffer interface) toward repulsion. Reversals upon entering the 1-oct zone are increased (left side/red), and reversals upon exiting the 1-oct zone are decreased (right side/green). Number of encounters scored shown in bars, from multiple independent trials (six for -IAA, four for +IAA). (E, F) Effect of IAA on locomotory activity. (E) 1-oct (0.66 mM) in the absence of IAA does not cause increased speed. (F) 1-oct (0.66 mM) in the presence of IAA causes markedly increased speed. Number of tracks analyzed shown in bars, from multiple independent trials (three in E, four in F) (G) Single (upper) and averaged (lower) recordings of ASH and AVA before and after addition of IAA and 1-oct (0.66 mM) as shown, demonstrating increased AVA entrainment to stimulus (compare to Figure 6D/middle). Blue bar indicates addition of IAA (in both control buffer and 1-oct solutions). (H) AVA entrainment index of averaged traces confirms that the presence of IAA causes AVA to become significantly entrained with the stimulus. ** in (C) indicates p<0.01, t-test; *** and **** in (D) indicate p<0.001, 0.0001, respectively, Fisher’s exact test; *, *** in (F) indicate p<0.05 and 0.001, respectively, ANOVA and Tukey’s multiple comparison test. Error bars are SEM.
Figure 8
Combinatorial coding of 1-oct as an attractant and a repellent, simultaneously, facilitates sensory motor coupling and context-dependent modulation in C. elegans.
(A) 1-oct simultaneously activates sensory neurons in the periphery that mediate both repulsion (ASH and others) and attraction (AWC and others), such that repulsive and attractive afferent pathways become active concurrently. These inputs are then integrated centrally to control at least three key locomotory parameters (exit reversals, entry reversals, and locomotory speed), and the activity state of the reverse command neurons. (B) An analogy based on refraction on light through two prisms highlights the potential versatility of olfactory combinatorial coding in C. elegans. An individual odorant activates many chemosensory neurons to produce multiple distinct afferent inputs propagating through the downstream circuitry simultaneously, analogous to the first prism separating white light into its component colors (top panel, left). The motor integration circuitry formulates a locomotory response by integrating the active afferent pathways, analogous to the second prism recombining the incoming light (top panel, right). Because chemosensory neurons have different odorant affinities, the specific outputs of the sensory array and motor integration circuit may be concentration dependent. In this study, ASH (coded in red) is a low affinity sensor, while AWC (coded in blue) is a high affinity sensor. Reducing the 1-oct concentration biases the output of the sensory array toward AWC (middle panel, left), and the motor integration circuitry uses this information to produce an attractive locomotory response (middle panel, right). This coding strategy produces a simple framework for modulation. The behavioral outcome may be influenced by simply shifting the balance of the afferent pathways, much like an optical filter placed in a light path can selectively eliminate specific wavelengths (lower panel) to change the color of the transmitted light. In this study, IAA was added to the diluted 1-oct solution, inhibiting AWC, such that the afferent pathways were re-balanced toward the ASH end of the spectrum, and repulsion was restored. This ‘filtering’ effect could, in principle, take place peripherally or centrally, and could subserve modulation due to the presence of other odorant molecules, monoamine/neuropeptide modulators serving as interoceptive signals, and effectors of experience-dependent plasticity.
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Strain, strain background (Escherichia coli) | OP50 | Caenorhabditis Genetics Center (CGC) | OP50 | RRID:WB-STRAIN:WBStrain00041970 |
| Strain, strain background (Caenorhabditis elegans) | Wild type | Caenorhabditis Genetics Center (CGC) | N2 | RRID:WB-STRAIN:WBStrain00000001 |
| Strain, strain background (C. elegans) | otIs669 V, otIs672 [rab-3::NLS::GCaMP6s+arrd-4:NLS:::GCaMP6s] | Caenorhabditis Genetics Center (CGC) | OH15500 | Yemini et al., 2021 |
| Strain, strain background (C. elegans) | otIs669 V, wtfIs145 [rab-3p::his-24::GCaMP6s::unc-54 3' UTR +pBX] | Caenorhabditis Genetics Center (CGC) | AML320 | |
| Strain, strain background (C. elegans) | tax-4 (ok3771) III | Caenorhabditis Genetics Center (CGC) | VC3113 | RRID:WB-Strain:WBStrain00037678 |
| Strain, strain background (C. elegans) | tax-4 (ok3771) III, otIs669 V, wtfIs145 [rab-3p::his-24::GCaMP6s::unc-54 3' UTR +pBX] | This study | FY1041 | See ‘Worm Maintenance and Strains’; Available from C. elegans Genetics Center (CGC) |
| Strain, strain background (C. elegans) | njEx1547[ceh-36p:: HisCl1::SL2::GFP, ges-1p::NLS:: TagRFP] | Ikeda et al., 2020 | IK3591 | Ikeda et al., 2020 |
| Strain, strain background (C. elegans) | kyEx5104 [sra-6p:: HisCl1::SL2:: mCherry, 50 ng/μL+elt-2::mCherry, 1 ng/μL] | Pokala et al., 2014 | CX15206 | Pokala et al., 2014 |
| Chemical compound, drug | Sylgard 184 | Electron Microscopy Sciences | Cat# 24236-10 | |
| Chemical compound, drug | Tetramisole hydrochloride | Sigma | CAS# 16595-80-5 | |
| Software, algorithm | Adobe Premier Pro | Adobe Inc. | https://www.adobe.com/ | |
| Software, algorithm | Multi-Worm Tracker | Itskovits et al., 2017 | https://github.com/itskov/MultiAnimalTrackerSuite | Itskovits, 2017 |
| Software, algorithm | Mathematica | Wolfram Research | https://www.wolfram.com/ | |
| Software, algorithm | MATLAB | MathWorks | https://www.mathworks.com | |
| Software, algorithm | R | Posit PBC | RStudio v2022.07.1 | |
| Software, algorithm | Fiji/ImageJ | Schindelin et al., 2012 | https://imagej.net/software/fiji/downloads | Schindelin et al., 2012 |
| Software, algorithm | Graphpad | GraphPad Software | Prism 10.2.3 | |
| Software, algorithm | MoCo plugin | Dubbs et al., 2016 | https://github.com/NTCColumbia/moco | NTCColumbia, 2016 |
| Software, algorithm | WormLab tracking software | MBF Biosciences | https://www.mbfbioscience.com | |
| Other | Silicon microfluidics masters | Lockery and D. Albrecht | See section ‘1-octanol repulsion in microfluidic arenas’ | |
| Other | Olfactory chips | Microkosmos | https://www.ukosmos.com/ | See section ‘Network-wide response to repulsive olfactory stimulation’ and Chronis et al., 2007 |
Additional files
-
MDAR checklist
- https://cdn.elifesciences.org/articles/104828/elife-104828-mdarchecklist1-v1.docx
-
Source code 1
A computer simulation predicting the effects of speed and reversal probability on the partitioning of moving particles into two adjacent zones.
- https://cdn.elifesciences.org/articles/104828/elife-104828-code1-v1.zip
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Olfactory combinatorial coding supports risk–reward decision-making in Caenorhabditis elegans
eLife 14:RP104828.
https://doi.org/10.7554/eLife.104828.4
