Specific expression of channelrhodopsin-2 in single neurons of Caenorhabditis elegans

Abstract

Optogenetic approaches using light-activated proteins like Channelrhodopsin-2 (ChR2) enable investigating the function of populations of neurons in live Caenorhabditis elegans (and other) animals, as ChR2 expression can be targeted to these cells using specific promoters. Sub-populations of these neurons, or even single cells, can be further addressed by restricting the illumination to the cell of interest. However, this is technically demanding, particularly in free moving animals. Thus, it would be helpful if expression of ChR2 could be restricted to single neurons or neuron pairs, as even wide-field illumination would photostimulate only this particular cell. To this end we adopted the use of Cre or FLP recombinases and conditional ChR2 expression at the intersection of two promoter expression domains, i.e. in the cell of interest only. Success of this method depends on precise knowledge of the individual promoters' expression patterns and on relative expression levels of recombinase and ChR2. A bicistronic expression cassette with GFP helps to identify the correct expression pattern. Here we show specific expression in the AVA reverse command neurons and the aversive polymodal sensory ASH neurons. This approach shall enable to generate strains for optogenetic manipulation of each of the 302 C. elegans neurons. This may eventually allow to model the C. elegans nervous system in its entirety, based on functional data for each neuron.

Figure 1. Conditional expression of ChR2 in single neurons using two promoters of intersecting expression domain, and DNA recombinases.

A) Two promoters, each expressing in a different set of neurons, with overlapping expression in one neuron only. B) Concept of achieving conditional expression using split proteins that can reconstitute functional ChR2 or NpHR in the neuron of interest. See also Fig. S1. C) Conditional expression is achieved by encoding ChR2 form a construct (driven by promoter I) that is blocked by a transcriptional stop cassette (red hexagon) between promoter and ChR2 start codon, and is flanked by recombinase recognition sites (loxP or FRT sites, recognized by Cre or FLP recombinases, respectively). The respective recombinase is driven by promoter II, to generate a transcription-competent DNA construct encoding ChR2 and (optionally) soluble GFP from a bicistronic expression cassette (via SL2 trans-splicing).

Figure 2. Promoter combinations chosen for expression of ChR2 in PVC neurons, using Cre recombinase, are not PVC-specific.

A) Confocal stack of an animal expressing ChR2::mCherry in head neurons and PVC, using the glr-1 and des-2 promoters, resembling the pglr-1 expression pattern. B) The promoter combination nmr-1 and des-2 led to an expression in four neurons in the head (AVA, AVB, AVD, AVE), in addition to expression in PVC in the tail (C). Scale bars = 20 µm.

Figure 3. Expression pattern and behavioral responses of animals generated towards ChR2 expression in ASH by using Cre recombinase.

GFP fluorescence (confocal stacks) in animals expressing ChR2::mCherry and GFP (bicistronically co-expressed) using the promoter combinations osm-10 and gpa-11 A), osm-10 and nhr-79 B), or sra-6 and gpa-13 C). Scale bar = 20 µm. D) Behavioral assay testing withdrawal reactions in response to 470 nm blue light illumination of the indicated conditional expression strains (all lite-1(ce314) background). Strain AQ2334 (pmec-4::ChR2) and lite-1(ce314) were used as positive and negative controls, respectively. n = 14 animals of each strain were tested, with 5 consecutive light pulses each; displayed are means ± SEM. Stastistically significant differences were determined by t-test, relative to the lite-1(ce314) control. * p<0.05; *** p<0.001.

Figure 4. Expression pattern and behavioral responses of animals generated towards ChR2 expression in AVA by using FLP recombinase.

A) Fluorescence images of animals expressing ChR2::YFP by using flp-18 and rig-3 promoter pair with the FLP recombinase (top). Bottom panel shows pflp-18::mCherry expression. Scale bar = 20 µm B) Behavioral assay testing withdrawal reactions in response to 470 nm blue light illumination were as described in Fig. 3D (n≥15).

Figure 5. Expression pattern and behavioral responses of animals generated towards ChR2 expression in AVA by using Cre recombinase.

GFP fluorescence (confocal stacks) in animals expressing ChR2::mCherry and GFP (bicistronically co-expressed) using the promoter combinations glr-1 and gpa-14 A), flp-18 and rig-3 B), or flp-18 and gpa-14 C). Scale bar = 20 µm. D) Behavioral assay testing withdrawal reactions in response to 470 nm blue light illumination were as described in Fig. 3D (n = 14).

Figure 6. ChR2::mCherry::SL2::GFP expression in AVA (and RIG) neurons using flp-18 and gpa-14 promoters and Cre recombinase from an integrated transgene.

A) Confocal stack showing GFP fluorescence of the specific expression pattern. Scale bar = 100 µm (20 µm in inset). Inset: Overlay of the ChR2::mCherry (red) and the much brighter SL2::GFP (green) expression. B) Withdrawal in response to whole-animal blue illumination, compared in the transgenic line before (Ex) and after chromosomal integration (Is) was compared after normalization to the responses of pmec-4::ChR2 animals tested alongside. C) Locomotion speed traces of animals of the integrated strain expressing ChR2 in AVA (and RIG) before, during and after patterned illumination of either of two different segments in the head region (blue, segment harboring AVA cell bodies; red, segment harboring RIG cell bodies, and parts of the AVA axons). In addition, the whole animal was illuminated (green). Displayed are means ± SEM (n≥15 in B and C). Duration of the light stimulus is indicated by a blue bar.