The neural G protein Gαo tagged with GFP at an internal loop is functional in Caenorhabditis elegans

Abstract Gαo is the alpha subunit of the major heterotrimeric G protein in neurons and mediates signaling by every known neurotransmitter, yet the signaling mechanisms activated by Gαo remain to be fully elucidated. Genetic analysis in Caenorhabditis elegans has shown that Gαo signaling inhibits neuronal activity and neurotransmitter release, but studies of the molecular mechanisms underlying these effects have been limited by lack of tools to complement genetic studies with other experimental approaches. Here, we demonstrate that inserting the green fluorescent protein (GFP) into an internal loop of the Gαo protein results in a tagged protein that is functional in vivo and that facilitates cell biological and biochemical studies of Gαo. Transgenic expression of Gαo-GFP rescues the defects caused by loss of endogenous Gαo in assays of egg laying and locomotion behaviors. Defects in body morphology caused by loss of Gαo are also rescued by Gαo-GFP. The Gαo-GFP protein is localized to the plasma membrane of neurons, mimicking localization of endogenous Gαo. Using GFP as an epitope tag, Gαo-GFP can be immunoprecipitated from C. elegans lysates to purify Gαo protein complexes. The Gαo-GFP transgene reported in this study enables studies involving in vivo localization and biochemical purification of Gαo to compliment the already well-developed genetic analysis of Gαo signaling.


Introduction
Ga o , the a subunit of the most abundant heterotrimeric G protein in the brain (Sternweis and Robishaw 1984), is in every neuron and can be activated by G protein-coupled receptors for every neurotransmitter tested (Jiang et al. 2001). Caenorhabditis elegans has a Ga o ortholog named GOA-1 that is >80% identical to mammalian Ga o and that is expressed in most or all neurons. GOA-1 has been shown by genetic analysis to inhibit neurotransmitter release and/or neural activity (Mendel et al. 1995;Sé galat et al. 1995;Nurrish et al. 1999, Ravi et al. 2020, but the molecular mechanisms by which Ga o signals to have these effects remain to be fully defined. While activated Ga o releases Gbc subunits to regulate specific potassium and calcium channels (Lü scher and Slesinger 2010; Proft and Weiss 2015), genetic studies in C. elegans suggest that signaling through Gbc is not likely the sole mechanism by which Ga o has its physiological effects (Koelle 2018). It remains unclear if activated Ga o, like all other Ga proteins in animal cells, may itself bind target "effector" proteins to propagate a signal.
A method to fuse Ga o to fluorescent proteins and/or epitope tags without disrupting its function would enable new experimental approaches to help resolve unanswered questions about Ga o signaling. For example, Ga o -GFP fusion proteins could be visualized in real time in living cells for cell biological studies, and anti-GFP antibodies could be used to immunopurify Ga o protein complexes for biochemical analysis.
The challenge to this approach is that tags at the N-or C-termini would likely disrupt Ga o function since Ga proteins use their N-and C-termini to interact with receptors, Gbc subunits, and membranes (Hynes et al. 2004). For example, a previous study in C. elegans used a multicopy transgene to overexpress Ga q with GFP fused to its N-terminus (Bastiani et al. 2003). This fusion protein was able to fully rescue one behavioral defect of a Ga q partial loss-of-function mutant, while only partially rescuing other defects. The multi-copy transgene also created a gain-of-function effect which a single-copy transgene, not available at the time, might have been able to avoid.
Recent efforts to functionally tag Ga proteins have focused on inserting fluorescent proteins at internal sites. Internally tagged Ga proteins have been shown to be activated by G protein coupled receptors when co-overexpressed in cultured cells with both a receptor and Gbc subunits (Hughes et al. 2001;Yu and Rasenick 2002;Bü nemann et al. 2003;Galé s et al. 2005;Lazar et al. 2011); however, some internal insertions alter Ga function, and overexpressed receptors can promiscuously activate Ga proteins they would not otherwise activate (Gibson and Gilman, 2006). Still, some tagged Ga proteins may be fully functional: in the yeast Saccharomyces cerevisiae and in the slime mold Dictyostelium discoideum, an internally tagged Ga protein can replace the untagged Ga to support physiological functions that depend on activation by a single endogenous receptor (Janetopoulos et al. 2001;Yi et al. 2003). A remaining question is whether a tagged Ga protein could be fully functional in a metazoan, where it must mediate signaling from many different receptors to control diverse, tissuespecific physiological functions.
Here, we demonstrate that C. elegans Ga o with GFP inserted into an internal loop, when expressed at normal levels in the animal, rescues multiple defects in behavior and development caused by loss of native Ga o . We show that this tagged protein can be used to visualize Ga o subcellular localization in living animals and to purify both inactive and activated Ga o protein complexes from C. elegans lysates.

Materials and methods
All reagents were from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise indicated.

Strains and culture
Caenorhabditis elegans strains were cultured at 20˚C on NGM agar plates with Escherichia coli strain OP50 as a nutrition source (Brenner 1974). All strains were derived from the wild-type strain N2. Generation of transgenic animals and genetic crosses were by standard methods (Evans 2006;Fay 2013). Table 1 shows a list of C. elegans strains used in this study. goa-1::gfp plasmid construction A plasmid to express internally GFP-tagged GOA-1 in C. elegans was generated by first inserting a 9.0 kb C. elegans genomic fragment containing the goa-1 gene into a pBluescript vector and engineering in an SpeI restriction site between the goa-1 codons for T 117 and E 118 . The GFP coding region containing artificial introns was PCR-amplified from the vector pPD95.69 (Addgene plasmid #1491) using primers to add segments encoding SGGGGS and SGGGTS to flank the N-and C-termini of GFP, respectively, and the resulting gfp cassette was inserted into the SpeI site of the goa-1 clone. The resulting plasmid was named pMK376. To generate a clone suitable for miniMos single-copy insertion into the C. elegans genome (Frøkjaer-Jensen et al. 2014), we amplified the GOA-1-GFP coding region from pMK376 along with 4987 bp of 5' promoter and 432 bp of 3' UTR using primers mini909FWD (5'-gagattctgcaggaattccaactgaatttagatttttaaagt-3') and mini909REV (5'-agattaggcctggagtcttttcacccatacttccggaataa-3'), which added StuI and PstI restriction sites (boldface) at the 5' and 3' ends of the amplicon, respectively. The amplicon was then digested and ligated into the pCFJ909 miniMos vector (Frøkjaer-Jensen et al. 2014) using the StuI and PstI restriction sites. The resulting plasmid was named pAO8.

Confocal imaging
Worms were on mounted on 2% agarose pads containing 120 mm Optiprep (Sigma Millipore) to reduce refractive index mismatch (Boothe et al. 2017) on premium microscope slides Superfrost (Thermo Fisher Scientific), and a 22 Â 22-1 microscope cover glass (Fisher Scientific) was placed on top of the agarose pad. Worms were anesthetized using a drop of 150 mm sodium azide (Sigma Millipore) with 120 mm Optiprep. Z-stack confocal images of 24 hour old larvae were taken on a Zeiss LSM 880 microscope using a 63X objective lens.

GOA-1 antibody
The affinity-purified rabbit anti-GOA-1 polyclonal antibody used was from Chase et al. (2001). Whole-mount stains of C. elegans were performed as described by Finney and Ruvkun (1990).

Behavioral and worm length assays
Quantitation of unlaid eggs and staging of laid eggs were performed using 30 hour post-L4 adult animals as described in Chase and Koelle (2004). For analysis of worm tracks, reversaltouch behavior, and worm length, worms were staged 24 hours post-L4 and transferred to an NGM agar plate with a thin lawn of OP50 bacteria for imaging. Imaging began 2-20 minutes after transfer of the worm to the new plate and was carried out using a Leica M165FC microscope equipped with a digital camera. For measurements of worm length, >10 second digital video recordings of worms were analyzed using WormLab software from MBF Bioscience. Reversal-touch behavior was defined as a reversal during which the worms bends deeply enough that it contacts itself. Usually the tail or head touches the body, but in some cases two sections of the midbody can touch each other during very deep bends. Reversal-touch behavior was scored by placing a single worm staged 24 hours post-L4 on a new NGM plate with a lawn of OP50 bacteria, waiting 1-10 minutes, and then counting the number of reversal-touches made by the worm during a 15second time period.

Statistical analysis
Error bars shown in the graphs in Figures 2 and 3 represent 95% confidence intervals. All statistical analyses were done using GraphPad Prism version 9.0.1 software. The early-stage egg assay data set was analyzed using Fisher's exact test with two-sided Pvalues. The remaining data sets were analyzed using one-way ANOVA with Sídá k's multiple comparisons test.
Immunoprecipitation of GOA-1::GFP Worm lysates were prepared as described previously (Porter and Koelle 2010) with some modifications. Briefly, C. elegans were grown in 20 ml liquid cultures at 20 C and worms were isolated by flotation on 30% sucrose. Packed worm pellets ($500-600 ml) were resuspended in 4 ml lysis buffer [50 mM HEPES pH7.4, 100 mM NaCl, 1 mM EDTA, 3 mM EGTA, 10 mM MgCl 2 , 1 mM DTT, 1% Triton X-100 and complete protease inhibitor cocktail (Roche #04693159001)]. Resuspended worm pellets were homogenized by passing them two times through a French press (Spectronic Instruments, model number FA078) at 13000 PSI. The resulting lysates were centrifuged at 100,000Xg for 30-60 minutes at 4 C in an Optima TLX tabletop ultracentrifuge using a TLA-110 rotor (Beckman Coulter, Fullerton, CA, USA). The clarified supernatants were removed and transferred into the new tubes. The protein concentrations were determined by Bio-Rad protein assay. Whole worm lysates were aliquoted and snap frozen in liquid nitrogen for storage at À80 C until use. Immunoprecipitation was performed using Pierce crosslink immunoprecipitation kits (Pierce #26147) and the buffers therein, per manufacturer's instructions with some modifications. All spins in this procedure to separate supernatants from beads were at 3000 RPM for 1 minute in a microcentrifuge equipped with a swinging bucket rotor. To prepare beads sufficient for immunoprecipitating three samples, 90 ml of 50% protein A/G agarose slurry/sample was washed three times with 1 ml phosphate buffered saline (PBS, 10 mM sodium phosphate, 0.15 M NaCl, pH 7.5) and then incubated at room temperature for 2 hours on a rotary mixer at room temperature with 10 mg mouse monoclonal anti-GFP antibody (Rockland #600-301-215) diluted in 500 ml PBS. Beads were washed three times with PBS followed by incubation with Pierce DSS crosslinker working solution (150 ml) at room temperature for 1 hour on a rotary mixer. The supernatant was discarded and the beads were washed twice with Pierce elution buffer followed by three washes with IP buffer (50 mM HEPES pH 7.4, 100 mM NaCl, 1 mM EDTA, 3 mM EGTA, 10 mM MgCl 2 and 1% Triton X-100). While the anti-GFP beads were being prepared, protein lysates were pre-cleared: for each immunoprecipitation sample, 1ml of 4mg/ml worm protein lysate was incubated with 15ml packed protein A/G beads (prewashed 3X in PBS) at 4 C for 1 hour on a rotary mixer. Then for each immunoprecipitation, 1 ml of 4 mg/ml protein pre-cleared protein lysate was incubated with 15 ml packed anti-GFP antibody cross-linked beads at 4 C for 2 hours on a rotary mixer. For in vitro activation ( Figure 4E), at the beginning of this 2 hours incubation, 100 mM GDP or GTPcS was added to the protein lysate. After removing the supernatant, the beads bearing the immunoprecipitated products were washed four times with IP buffer at 4 C and eluted with 50 ml of 2X LDS loading buffer (Invitrogen were video recorded and reviewed to quantitate the frequency of episodes of backward locomotion that included a body bend so deep that the animal touched its own body. (G) Videos of adult worms of the same genotypes (n ! 5 per genotype) were analyzed with WormLab software to measure their body length. In panels F and G, error bars represent 95% confidence intervals, comparisons labeled **** were statistically different with P < 0.0001; and all other pairwise comparisons were not statistically different. #NP0007) at 55 C for 30 minutes. The supernatants were transferred into the new tubes and 5% (by volume) b-mercaptoethanol was added and the tubes were incubated at 55 C for 15 minutes. The supernatants were collected, and samples were processed for SDS-PAGE gel electrophoresis and western blotting.

SDS-PAGE gel electrophoresis and Western blotting
For SDS-PAGE gel electrophoresis followed by total protein staining, IP samples were loaded on 4 À 12% Bis-Tris gels (NuPAGE #NP0322BOX) and separated using MOPS buffer (Novex #NP0001). Gels were stained with Imperial Protein Stain (Thermofisher Scientific #24615) per manufacturer's instructions and images were captured using Epson Perfection V800 Photo Color Scanner. Images were processed by ImageJ software. For western blots, IP samples were loaded (20% of an IP sample/well) and separated on 10% SDS-PAGE gels. The protein was transferred onto a nitrocellulose membrane, and the blot was blocked and incubated with a primary anti-GOA-1 antibody [1:1000 diluted affinity-purified rabbit anti-GOA-1 polyclonal antibody (Patikoglou and Koelle 2002) at 4 C for overnight], washed, incubated with a secondary antibody (1:3000 HRP-linked Anti-Rabbit antibody Bio-Rad) and protein bands were visualized with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermofisher Scientific #34580) using a BioRad ChemiDoc MP system. Blots were reprobed for Gb by stripping and then incubating overnight with 1:200 diluted mouse monoclonal anti-Gb antibody (Santa Cruz #sc-166123) followed by a secondary incubation with 1:1000 diluted m-IgGk BP-HRP (Santa Cruz #sc-516102), and bands were again visualized by chemiluminescence.

Data availability
Strains and plasmids used in this study are available upon request.

Design of internally GFP-tagged Ga o and expression in C. elegans
We designed a functionally tagged C. elegans Ga o -GFP fusion protein, modeled in Figure 1A. The design inserts GFP that is flanked on either side by six amino acid flexible linkers into an internal loop of the alpha-helical domain of GOA-1, the C. elegans ortholog of Ga o (Mendel et al. 1995;Sé galat et al. 1995). An analogous GFP insertion site was used by Hughes et al. (2001) to generate a mammalian Ga q-GFP fusion protein that was capable of mediating signaling in cultured cells when overexpressed with the a 2aadrenergic receptor. Gibson and Gilman (2006) further showed that insertion of YFP in the analogous loop of mammalian Ga i1 did not alter the nucleotide exchange or GTPase reaction rates of the purified protein, and that this Ga i1 -YFP fusion protein could be activated in cultured cells by overexpressed and endogenous a 2 -adernergic receptors.
To express Ga o -GFP in C. elegans, we modified a 9.0 kb genomic clone containing the entire Ga o gene (including its promoter, exons and introns, and 3' region) to insert the GFP coding sequences between the codons for Ga o amino acids T117 and E118. We also generated an activated mutant version of this Ga o -GFP clone in which we altered codon 205 to encode L instead of Q, a change that disrupts GTPase activity of Ga o and renders the protein constitutively active (Mendel et al. 1995). The wild-type and Q205L versions of the Ga o -GFP transgene were separately inserted as single-copy transgenes into the C. elegans genome using Mos1 transposase (Frøkjaer-Jensen et al. 2008. We also crossed the wild-type Ga o -GFP transgene into a mutant strain of C. elegans lacking endogenous Ga o protein due to the Ga o gene carrying an early stop codon mutation (Robatzek and Thomas 2000).  Figure 1B shows Western blots of whole-worm protein lysates probed with an antibody against Ga o . The endogenous Ga o protein and the higher molecular weight Ga o -GFP fusion proteins gave signals of similar intensity, suggesting that insertion of GFP into Ga o did not interfere with its expression or stability.
We examined the localization of Ga o -GFP in C. elegans animals. Previous studies demonstrated that the C. elegans Ga o gene is expressed in most or all neurons (Mendel et al. 1995;Sé galat et al. 1995). Using an antibody against C. elegans Ga o to stain wild-type animals, we visualized endogenous Ga o protein concentrated in bundles of neural processes, such as the nerve ring in the head, as well as in what appeared to be the plasma membranes of neural cell bodies ( Figure 1C). Green fluorescence in transgenic animals carrying the Ga o -GFP or Ga o (Q205L)-GFP transgenes was localized in patterns that closely mimicked the localization of endogenous Ga o (Figure 1, D and E).

Ga o -GFP rescues the locomotion and body morphology defects of a Ga o null mutant
Ga o null mutants have defects in locomotion behavior (Mendel et al. 1995;Sé galat et al. 1995), and we tested whether these defects could be rescued by transgenically expressed Ga o -GFP. Figure 2, A-E shows photographs of Petri plates on which individual worms have left tracks that reveal features of their locomotion behavior. We also analyzed video of worms moving on such Petri plates to quantitate the abnormally deep body bends that Ga o null mutants make during backward locomotion ( Figure 2F). Wild-type worms (Figure 2A) move forward with smooth sinusoidal body bends and rarely reverse direction, but Ga o null mutants make abnormal body bends that leave abnormal tracks ( Figure  2B) and that results in the animals sometimes bending so deeply during backward locomotion that they touch their own body, events we term "reversal-touches" ( Figure 2F). We found that expression of Ga o -GFP in the Ga o null mutant background qualitatively rescued the Ga o null mutant locomotion defects as judged by the tracks worms made (Figure 2, A-C), and quantitation showed that the reversal-touch frequency defect was fully rescued ( Figure 2F). The locomotion defects in Ga o mutants arise at least in part because Ga o is required to inhibit neurotransmitter release in ventral cord motor neurons that control locomotion behavior (Nurrish et al. 1999). The rescue of these defects suggests Ga o -GFP is functional in regulating neurotransmitter release.
Ga o (Q205L) is an activated mutant of Ga o with a mutation thought to block the GTPase activity of this G protein, thus trapping Ga o in its active GTP-bound state. Expression of Ga o (Q205L) in C. elegans leads to gain-of-function phenotypes, such as shallow body bends (Mendel et al. 1995). Transgenic expression of Ga o (Q205L)-GFP in otherwise wild-type C. elegans also caused a gain-of-function phenotype, since the tracks left by these worms show very shallow bends ( Figure 2E), while similar transgenic expression of Ga o -GFP without the Q205L mutation did not have this effect ( Figure 2D). Thus Ga o -GFP, like wild-type Ga o , can be activated by the Q205L mutation.
Ga o null mutants have a scrawny body morphology which is seen in Figure 3, A and B and was also evident as a decrease in the length of the worms we photographed and tracked for the experiments shown in Figure 2, A and B. This body length defect was fully rescued by expression of Ga o -GFP (Figure 2, C and G).
Ga o -GFP partially rescues egg-laying behavior defects of a Ga o null mutant Ga o null mutants show a hyperactive egg-laying behavior defect (Mendel et al. 1995;Sé galat et al. 1995;Koelle and Horvitz 1996) due at least in part to Ga o acting in the HSN motor neurons to inhibit their release of serotonin (Tanis et al. 2008). This defect leads Ga o null mutant adult animals to retain very few unlaid eggs, since their eggs are laid almost as soon as they are produced (Figure 3, A, B, and F). Another way to measure hyperactive egglaying behavior is to count the fraction of freshly laid eggs that are at early stages of development (Chase and Koelle 2004). Hyperactive egg-laying mutants such as the Ga o null mutant lay eggs so soon after they are fertilized that the laid eggs are often at the eight-cell stage or earlier, whereas wild-type animals rarely lay such early-stage eggs ( Figure 3G). Both assays of egg-laying behavior show that the hyperactive egg-laying defect in the Ga o null mutant was substantially although not fully rescued by expression of Ga o -GFP (Figure 3, C, F, and G). The partial rescue of the Ga o egg-laying defect seen in Figure 3 as opposed to the full rescue of the Ga o locomotion reversal defect seen in Figure 2D may reflect differences in the ability of Ga o -GFP fully function in egg-laying vs locomotion neurons; alternatively, it could simply reflect an ability of the egg-laying assays to detect smaller differences in Ga o function.
Transgenic expression of Ga o (Q205L)-GFP in otherwise wildtype C. elegans caused a gain-of-function phenotype in which animals fail to lay eggs, resulting in an accumulation of unlaid eggs, while similar transgenic expression of Ga o -GFP without the Q205L mutation did not have this effect (Figure 3, A and D-F). Thus, Ga o -GFP not only rescues the behavioral defects of a Ga o null mutant, but can also be activated by the Q205L mutation to induce gain-of-function defects that are opposite to the defects of the null-mutant (Figures 2 and 3).

Immunoprecipitation of Ga o -GFP in both its inactive and active states
Although an antibody that recognizes C. elegans Ga o on Western blots (Patikoglou and Koelle 2002) and in whole-mount stains of C. elegans animals ( Figure 1C) is available, no antibody we have tested can be used to immunoprecipitate C. elegans Ga o , and this has remained an obstacle to biochemical studies of this protein. Therefore, we tested whether Ga o -GFP can be immunoprecipitated from C. elegans lysates using an anti-GFP monoclonal antibody. In these experiments, we assessed the activation state of Ga o -GFP immunoprecipitated from whole-worm lysates by testing if the Gb subunit co-precipitates, since Gb should be in a complex with inactive but not with activated Ga o ( Figure 4A).
We found that Ga o -GFP, along with its associate Gb subunit, were the major proteins found in anti-GFP immunoprecipitates of worm lysates expressing Ga o -GFP (Figure 4, B-D). When worm lysates expressing Ga o (Q205L)-GFP were immunoprecipitated with the same antibody, the Ga o -GFP protein but not Gb were precipitated, confirming that the Q205L mutation locked Ga o -GFP in its active, GTP-bound state.
We were also able to fully activate the Ga o -GFP protein in a whole-worm lysate by incubating it with the nonhydrolysable GTP analog, GTPcS. The left lane of the gel in Figure 4E shows a total protein stain of a control immunoprecipitate of Ga o -GFP from a C. elegans lysate treated with GDP to maintain the G protein in an inactive state. An approximately equal amount of the Ga o -GFP and Gb proteins were coprecipitated, as expected if the Ga o -GFP present in this lysate was close to 100% in the inactive, GDP-bound heterotrimer state in which Ga o -GFP and Gb associate in a 1:1 stoichiometry, as modeled in Figure 4A. The identity of the major protein bands in Figure 4E as Ga o -GFP and Gb were confirmed in the Western blots in Figure 4, F and G. The right lanes in Figure 4, E-G shows a parallel analysis of Ga o -GFP immunoprecipitates from the same protein lysate, but this time after treatment of the lysate with the nonhydrolysable GTP analog GTPcS. In this experiment, no detectable Gb was coprecipitated with Ga o -GFP, consistent with the Ga o -GFP protein being close to 100% converted to the activated GTPcS-bound form, in which it is dissociated from Gb as modeled in Figure 4A.

Discussion
The goal of this study was to functionally tag Ga o in C. elegans with a fluorescent protein to facilitate cell biological and biochemical studies of this major neural signaling protein. We adapted the approach of Hughes et al. (2001) by inserting GFP flanked by flexible linkers into a specific internal loop of the alpha-helical domain of the Ga protein. We found that single-copy transgenes express such Ga o -GFP fusion proteins in C. elegans at levels similar to that of the endogenous Ga o protein, and that these Ga o -GFP fusion proteins appear to be localized to the processes and plasma membranes of neurons, similar to the localization of the endogenous Ga o protein. The wild-type version of the Ga o -GFP protein can rescue the behavioral and body morphology defects of a Ga o null mutant, although the extent of this rescue ranges from full to partial depending on the specific phenotypic defect analyzed. An activated mutant version, Ga o (Q205L)-GFP, can induce a gain-of-function phenotype. Immunoprecipitated wild-type Ga o -GFP appears to be stoichiometrically associated with Gb, but can be fully activated and dissociated from Gb either with the activating Q205L mutation or by incubation with the nonhydrolysable GTP analog GTPcS.
In this study, we inserted Mos1 transposon-based single-copy Ga o -GFP transgenes into the genome, which allowed us to cross these transgenes either into a wild-type background that also expresses endogenous Ga o , or into a Ga o knockout background that lacks endogenous Ga o . During the construction of our Mos1 transgenes, Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) emerged as a powerful technique to edit the C. elegans genome (Nance and Frøkjaer-Jensen 2019). CRISPR editing has now been used to insert GFP coding sequences directly into the endogenous C. elegans Ga o gene using the same insertion point and the same flanking linker sequences we used in constructing the Mos1 Ga o -GFP transgene in this work (Catharine Rankin, personal communication). Preliminary analysis suggests that the Mos1 Ga o -GFP transgene and the CRISPR Ga o -GFP edit produce similar GFP fluorescence and behavioral effects (data not shown).
As of this writing, over 100 research articles have been published analyzing Ga o function in C. elegans, reflecting the importance of this neural signaling protein (WormBase, http://www. wormbase.org, release WS279). Because this past work has relied almost exclusively on genetic methods to make indirect inferences about the molecular mechanisms of Ga o signaling, it has had a limited ability to make definitive conclusions about such mechanisms. For example, a several studies have speculated as to whether Ga o signals by directly binding and activating the diacylglycerol kinase DGK-1 (the worm ortholog of mammalian DGK) based on genetic results consistent with this hypothesis (Miller et al. 1999;Nurrish et al. 1999;Jose and Koelle 2005;Koelle 2018); however, the lack of methods to immunoprecipitate Ga o protein complexes from worm lysates have prevented a clear test of this hypothesis. Another line of genetic work in C. elegans has suggested that major neural protein kinase CaMKII (known as UNC-43 in C. elegans) may phosphorylate Ga o to regulate Ga o signaling (Robatzek and Thomas 2000), but this hypothesis has not been tested using biochemical approaches for lack of tools to isolate and directly examine the C. elegans Ga o protein from lysates of wild-type vs unc-43 mutants.
Beyond its role in neural signaling, Ga o also plays a central role in mitotic spindle positioning during asymmetric cell divisions in early development (Gotta and Ahringer 2001). Cell biological studies of asymmetric cell division in C. elegans embryos have depended heavily on the use of functional GFP fusions to proteins that control cell polarity and mitotic spindle positions, as these tools make it possible to track the positioning and dynamic movements of these proteins during cell divisions (Rose and Gö nczy 2014). Antibody stains suggest that Ga o may be localized to spindle asters in dividing embryonic cells (Gotta and Ahringer 2001), but this early finding has not been followed up. Our development of functional Ga o -GFP transgenes should enable a more definitive analysis of the role of Ga o in asymmetric cell divisions.