Leveraging a gain-of-function allele of Caenorhabditis elegans paqr-1 to elucidate membrane homeostasis by PAQR proteins

Summary

The C. elegans proteins PAQR-2 (a homolog of the human seven-transmembrane domain AdipoR1 and AdipoR2 proteins) and IGLR-2 (a homolog of the mammalian LRIG proteins characterized by a single transmembrane domain and the presence of immunoglobulin domains and leucine-rich repeats in their extracellular portion) form a complex that protects against plasma membrane rigidification by promoting the expression of fatty acid desaturases and the incorporation of polyunsaturated fatty acids into phospholipids, hence increasing membrane fluidity. In the present study, we leveraged a novel gain-of-function allele of PAQR-1, a PAQR-2 paralog, to carry out structure-function studies. We found that the transmembrane domains of PAQR-2 are responsible for its functional requirement for IGLR-2, that PAQR-1 does not require IGLR-2 but acts via the same pathway as PAQR-2, and that the divergent N-terminal cytoplasmic domains of the PAQR-1 and PAQR-2 proteins serve a regulatory function and may regulate access to the catalytic site of these proteins. We also show that overexpression of human AdipoR1 or AdipoR2 alone is sufficient to confer increased palmitic acid resistance in HEK293 cells, and thus act in a manner analogous to the PAQR-1 gain-of-function allele.

Keywords:

Phenotypes – Alleles – Caenorhabditis elegans – Lipids – Protein domains – Cell membranes – Membrane proteins – Fluorescence recovery after photobleaching

Introduction

Maintenance of cell membrane homeostasis relies on regulatory proteins that sense and respond to properties such as lipid composition, thickness, compressibility, lateral mobility and curvature [1, 2]. The best understood example is perhaps the bacterial multi-pass protein DesK that undergoes a conformational change in response to increased membrane rigidification, resulting in the activation of its kinase domain [3–6]. Similarly, the yeast single-pass plasma membrane protein Mga2 rotates along its long axis when the surrounding acyl chains are densely packed, also resulting in its activation [7]. In eukaryotes, the SREBPs are regulated by the availability of specific lipids such as cholesterol in the ER membrane [8], PCYT1A is activated by association with packing defects in the inner nuclear membrane [9–11] and IRE1 is activated by multimerization in response to membrane thickening in the ER [1, 12]. Recently we identified a novel regulator of membrane homeostasis in animal cells, namely the PAQR-2/IGLR-2 complex in the nematode C. elegans [13–18]. The present work helps define the structure-functional basis of fluidity sensing by this complex.

The PAQR family of proteins (named after the founding members Progestin and AdipoQ Receptors) are characterized by the presence of seven transmembrane domains oriented such that their N-terminus is cytosolic [19]. PAQR proteins all likely act as hydrolases but with a wide range of substrates and biological functions [20]. The crystal structures of AdipoR1 and AdipoR2, which are PAQR proteins that may act as receptors for adiponectin, have been partially resolved and show that the transmembrane domains form a barrel with a cavity that opens on the cytosolic side where it may allow the in/out flow of substrates and products of a hydrolytic reaction coordinated by a zinc ion [21, 22]. There are yeast homologs to the AdipoRs that act as ceramidases signaling through sphingolipids to regulate lipid/membrane homeostasis [23–26], and a ceramidase activity has also been documented for the human AdipoRs [22, 27–29], though other substrates may exist. The structures solved so far do not include the cytoplasmic N-terminal domain, which contains most of the sequence divergence among AdipoR-like proteins, and the function of this domain therefore remains unknown.

In C. elegans, there are five PAQR proteins, including the AdipoR homologs PAQR-1 and PAQR-2, with PAQR-2 being best characterized [30]. We previously showed that PAQR-2 is a regulator of membrane fluidity that acts together with its dedicated partner, the single-pass transmembrane protein IGLR-2 [16, 17]. paqr-2 and iglr-2 single and double mutants have the same phenotypes, including a characteristic tail tip defect and intolerance to cold and dietary saturated fatty acids (SFAs) [16, 31]. Additionally, the paqr-2 mutant, and presumably the iglr-2 mutant as well, also exhibits several phenotypes secondary to its primary membrane homeostasis defect, including defects in lifespan [30], vitellogenin trafficking [15], brood size [30], locomotion [30], autophagy [32] and proteostasis [33]. These phenotypes are secondary to the primary membrane fluidity defects of paqr-2 and iglr-2 mutants because they can be suppressed fully or partially by low, fluidizing concentrations of mild detergents [18], by providing supplements of unsaturated fatty acids [16, 18], or by secondary mutations that increase the relative abundance of unsaturated fatty acids (UFAs) among phospholipids, such as mdt-15(et14), nhr-49(et8), fld-1(et48) and several others [14, 17, 18]. PAQR-2 and IGLR-2 mostly act cell non-autonomously: expression in one large tissue (e.g. gonad sheath cells, intestine or hypodermis) is sufficient to rescue the entire worm; the one exception is the tail tip which requires local expression likely because it constitutes a site of poor lipid exchange with the rest of the worm [15]. Additionally, bifluorescence complementation (BiFC) showed that the proteins PAQR-2 and IGLR-2 interact with each other on the plasma membrane, and GFP reporters showed that they are both strongly expressed in the gonad sheath cells and weakly in many other tissues [17].

Less is known about the function or mechanism of action of PAQR-1, though one study indicated that it may modulate lipid metabolism and ER stress [34]. Recently, and to identify additional components of the paqr-2 pathway, we performed a forward genetic screen to isolate mutations that enhance the ability of mdt-15(et14) to suppress the SFA intolerance of the paqr-2(tm3410) null mutant. Of 15 novel mutants isolated in that screen, eight were alleles of the gene fld-1 that functions by limiting the levels of phospholipids containing long chain polyunsaturated fatty acids (PUFAs); fld-1 mutants therefore have elevated PUFA levels among their phospholipids [14]. Another mutation was a loss-of-function allele of the acyl-CoA synthetase acs-13; this mutation also promotes the incorporation of PUFAs into phospholipids [35]. In the present work, one of the remaining suppressor mutations has now been identified as a gain-of-function (gof) allele of paqr-1, i.e. paqr-1(et52). This is interesting because while the paqr-1 single mutant has no obvious phenotype, the double mutant paqr-1;paqr-2 shows much more severe defects than either single mutants, suggesting functional overlap between the two genes [30]. Here, we show that the paqr-1(et52) allele acts through the same pathway as paqr-2. We also show that paqr-1(et52) suppresses the phenotypes of paqr-2 and paqr-2 iglr-2 double mutants better than it suppresses the iglr-2 single mutant, which suggests that the PAQR-2 protein inhibits the gof PAQR-1 only when the IGLR-2 protein is also absent. Finally, we used domain-swapping experiments to show that the distinct intracellular N-terminal domains serve regulatory functions in the PAQR-1 and PAQR-2 proteins. This study is an important advance in our understanding of how PAQR-type proteins are regulated to achieve membrane homeostasis.

Results

paqr-1(et52) is a gof allele

The paqr-1(et52) allele was isolated in a previously published screen for suppressors of the paqr-2(tm3410) mutant SFA intolerance phenotype [14]. In this screen, the SFA-rich diet is achieved by cultivating C. elegans on plates containing 20 mM glucose, which is converted to SFAs by the dietary E. coli, also as previously described [16]. The amino acid sequences of PAQR-1 and PAQR-2 proteins are highly conserved throughout the 7 transmembrane domains and membrane-proximal ~100 amino acids of the cytoplasmic domain, but show very little sequence homology within the remainder of the large N-terminal regions, i.e. the first 200 amino acids of PAQR-1 and the first 309 amino acids of PAQR-2 (Fig 1A). The paqr-1(et52) mutation causes an R109C amino acid substitution about halfway within the cytoplasmic N-terminal domain, which suggests a regulatory function for this non-conserved domain (Fig 1A). The levels of protein expression appear mostly unaffected by the R109C amino acid substitution: Western blots against HA-tagged proteins expressed from CRISPR/Cas9-modified endogenous wild-type paqr-1 or mutant paqr-1(et52) loci show similar expression levels throughout development for both allele (both are most abundant in embryos and L1s; Fig 1B) or when cultivated at various temperatures (Fig 1C). In contrast, PAQR-2 expression is highest in adult and increases with temperature (Fig 1B and 1C). The localization of GFP translational reporters is also unchanged between the wild-type PAQR-1 and the mutant PAQR-1(R109C) proteins, which are both expressed in several tissues, but predominantly in the intestine and gonad sheath cells, which is the site of strongest PAQR-2 expression [17] (Fig 1D). In spite of unchanged expression levels and localization, paqr-1(et52) is clearly a gof mutation because providing it as a multicopy transgene in a paqr-2 mutant background, where the endogenous paqr-1 is wild-type, efficiently suppresses the SFA intolerance, cold intolerance and tail tip defect of the paqr-2 mutant (Fig 1E; S1A–S1C Fig). In contrast, providing wild-type paqr-1 as a multicopy transgene efficiently suppresses only the cold intolerance phenotype of the paqr-2 mutant, partially suppresses the tail tip defect and does not suppress at all the intolerance to SFAs, which is a harder challenge for the paqr-2 mutant since many paqr-2 suppressor alleles that fully rescue growth at 15°C do not effectively rescue growth on glucose [14, 18, 35](Fig 1E; S1A–S1C Fig). These results suggest that the R109C amino acid substitution endows the PAQR-1 protein with augmented activity compared to the wild-type protein, i.e. that it is a gof allele.

PAQR-1(R109A) also acts as a gain-of-function allele

The paqr-1(et52) allele corresponds to a R109C substitution that could promote dimerization via disulfide bridges involving the cysteine. This is however not the case: substituting the arginine at position 109 by an alanine (R109A) also results in an equally potent paqr-1 gof allele that efficiently suppresses the glucose and cold intolerance and tail tip defects of the paqr-2 mutant (Fig 1F and S1D and S1E Fig). We conclude that dimerization via disulfide bridges does not explain the gof nature of the R109C allele, and that the bulky and polarized arginine at position 109 limits the activity of the wild-type PAQR-1, perhaps by imposing an inhibitory structure or facilitating an inhibitory interaction that is abolished when the arginine is replaced by another amino acid.

paqr-1(et52) suppresses paqr-2 mutant phenotypes

Having established, using transgenic animals, that paqr-1(et52) is a gof allele, we proceeded to its more detailed characterization. The paqr-1(et52) mutation suppresses, partially or entirely, the defects in pharyngeal pumping rate, brood size, lifespan, defecation rate, and locomotion rate of the paqr-2 mutant (S1F–S1J Fig). Indeed, the paqr-1(et52) mutation partially or entirely suppresses all paqr-2 defects tested so far, including glucose intolerance (Fig 2A; S2A Fig), cold intolerance (Fig 2B; S2A Fig), the tail tip defect (Fig 2C; S2B Fig), the excess SFA and MUFA/PUFA depletion in phosphatidylethanolamines (PEs; Fig 2D and 2E; S2C Fig) as well as the membrane rigidification phenotype measured using Fluorescence Recovery After Photobleaching (FRAP) when paqr-2 mutants are cultivated on glucose or the SFA palmitic acid (Fig 2H and 2I). The paqr-1(et52) allele had no effect on the membrane fluidity of paqr-2 mutants on normal plates or as a single mutant on normal plates or plates containing glucose (S2E–S2H Fig). These results suggest that paqr-1(et52) acts as a complete functional replacement for paqr-2 and that it has no adverse effects under the conditions tested.

In contrast, the loss-of-function allele paqr-1(tm3262) does not act as a paqr-2 suppressor but rather slightly worsens several paqr-2 mutant phenotypes such as poor growth on normal media (Fig 2A), phospholipid composition defects (Fig 2F–2G; S2D Fig), brood size (S1F Fig) and locomotion rate (S1J Fig). These results suggest that the wild-type paqr-1 and paqr-2 genes are partially redundant for these traits.

paqr-1(et52) acts through the same pathway as paqr-2

Our previous work showed that IGLR-2 and PAQR-2 act as a fluidity sensor that signals through NHR-49 (a nuclear hormone receptor related to mammalian PPARα and HNF4β) and/or SBP-1 (the single worm homolog of the mammalian SREBPs) and MDT-15 (a mediator subunit that acts as a co-factor of NHR-49 and SBP-1) to promote the expression of Δ9 desaturases (summarized in Fig 3A) [18]. Using RNAi, we found that mdt-15, sbp-1 and the fat-5/-6/-7 desaturases (which share high sequence similarity and may all be silenced by siRNA against any one) are all required for the maximum activity of paqr-1(et52), scored by measuring growth on plates containing 20 mM glucose (Fig 3B) or on normal media (S3A Fig). Since these are all paqr-2 effectors [18], these results suggest that paqr-1(et52) is a version of paqr-1 that acts via the same pathway as paqr-2. However, paqr-2 and nhr-49 must also have separate functions since the single loss-of-function mutants are viable but the double mutant is lethal (Fig 3C and 3D). Introducing the paqr-1(et52) into the paqr-2;nhr-49 double mutant suppresses the lethality and tail tip defect (Fig 3C and 3D; S3B Fig) but does not suppress the glucose intolerance (glucose is here again used as an expedient way to provide an SFA-rich diet; Fig 3C and 3D) nor the cold intolerance (S3C Fig). These results suggest that nhr-49 is an essential paqr-2 and paqr-1(et52) downstream target for SFA tolerance but that it serves other functions as well. The paqr-2;sbp-1 double mutant is lethal and we were unable to create a paqr-2;sbp-1;paqr-1(et52) triple mutant. This suggests that paqr-1(et52) cannot completely replace all sbp-1-independent functions of paqr-2.

paqr-2 inhibits paqr-1(et52) when iglr-2 is absent

paqr-2 is totally dependent on the presence of a functional iglr-2 for its activity [17]. Similarly, paqr-1(et52) was unable to suppress the glucose intolerance of the iglr-2(et34) mutant (Fig 3E and 3F), though it suppressed the intolerance to cold (15°C), which is a milder stress (Fig 3G). Quite unexpectedly however, we found that paqr-1(et52) was able to suppress both glucose and cold intolerance in the paqr-1(et52);paqr-2(tm3410);iglr-2(et34) triple mutant (Fig 3E–3G), as well as its tail tip defect (S3D Fig). Taken together, these results suggest that the presence of paqr-2 inhibits paqr-1(et52) when iglr-2 is absent. Speculatively, it is possible that the PAQR-2 protein competes with PAQR-1(R109C) for a downstream factor only when IGLR-2 is absent, or that PAQR-2 can freely interact and interfere with PAQR-1(R109C) when IGLR-2 is absent. In either case, it is clear that the iglr-2 gene is not required for the ability of paqr-1(et52) to rescue the paqr-2 null mutant. Indeed, BiFC, a method that detects the interaction between PAQR-2 and IGLR-2 [17] (S4A and S4B Fig), failed to detect any interaction between IGLR-2 and either wild-type PAQR-1 or PAQR-1(R109C) (S4C–S4E Fig). This again is consistent with the PAQR-1 proteins acting independently of IGLR-2.

Domain swapping suggests that the N-terminal cytoplasmic domain is regulatory

The paqr-1(et52) allele corresponds to a R109C amino acid substitution in the N-terminal cytoplasmic domain that is divergent between PAQR-1 and PAQR-2. Given the high degree of similarities in their transmembrane domains and the fact that the paqr-1(et52) allele affects an amino acid outside of the presumed enzymatic domain, it seems likely that PAQR-1 and PAQR-2 have similar enzymatic functions but are regulated differently via their cytoplasmic N-terminal domains. To test this more directly, we proceeded to swap various domains between the two proteins and added an HA tag at the N-terminus of the resulting chimeric proteins, which were all expressed from the paqr-2 promoter to facilitate direct comparisons (Fig 4A). Two independent transgenic lines were generated for each construct, and their expression was confirmed using Western blots against the HA tag (Fig 4B). The constructs were then tested for their ability to rescue three phenotypes of paqr-2 single and paqr-2;iglr-2 double mutants: intolerance to cold, intolerance to glucose (which again is an expedient way to provide a SFA-rich diet[16]) and the tail tip morphology defect.

The full length PAQR-1(R109C) protein expressed from the paqr-2 promoter, but not the full length wild-type PAQR-1, efficiently rescued growth of the paqr-2 single mutant and paqr-2;iglr-2 double mutant at 15°C (Fig 4C) and on glucose (S5A Fig), as well as the tail tip morphology (S5B Fig). This confirms that addition of the HA tag at the N-terminus does not impair function and that driving expression of the PAQR-1(R109C) protein from the paqr-2 promoter supports the ability of this gof allele to function effectively.

A chimeric protein composed of the N-terminal cytoplasmic domain of PAQR-1(R109C) and the transmembrane domains of PAQR-2, was able to rescue growth at 15°C of the paqr-2 single mutant, but not of the paqr-2;iglr-2 double mutant (Fig 4D). This same chimeric protein did not rescue growth of the paqr-2 single mutant on glucose nor its tail tip defect (S5C and S5D Fig). These results indicate that it is the transmembrane domains of PAQR-2 that dictate a requirement for the presence of the IGLR-2 protein, and that combining the N-terminal domain of PAQR-1(R109C) with the transmembrane domains of PAQR-2 results in a protein with reduced activity compared to the wild-type PAQR-2. Interestingly, it was not possible to create a paqr-2;iglr-2 double mutant line expressing a chimeric protein consisting of the N-terminal cytoplasmic domain of wild-type PAQR-1 and the transmembrane domains from PAQR-2 (Fig 4D). This is consistent with the R109C mutation conferring increased activity to the N-terminal domain compared to that of the wild-type PAQR-1 protein.

Swapping the short extracellular C-terminal domain of the PAQR-2 protein with that of PAQR-1(R109C) resulted in a small reduction in the ability of the resulting chimeric protein to rescue the growth of paqr-2 single mutants or paqr-2;iglr-2 double mutants at 15°C (compare Fig 4E with Fig 4C) or on glucose (compare S5E Fig with S5A Fig), as well as the tail tip phenotype (compare S5F Fig with S5B Fig). In other words, the activity of PAQR-1(R109C) is reduced when it carries the longer 44 amino acids C-terminal domain of PAQR-2 rather than its native 5 amino acids extracellular C-terminus.

Finally, a construct bearing the transmembrane domains of PAQR-1 fused to the cytoplasmic N-terminal domain of PAQR-2 was created and found to act as a potent paqr-2 suppressor both in the presence or absence of IGLR-2 (Fig 4F and S5G and S5H Fig). This is an interesting result from which at least two conclusions may be drawn: 1) it is the transmembrane domains of PAQR-2 that impose a requirement for IGLR-2; and 2) fusing the PAQR-1 transmembrane domains to the cytoplasmic domain of PAQR-2 results in a protein that behaves as a PAQR-1 gain-of-function allele. A graphical summary of the main findings from the domain swapping experiments is presented in Fig 4G.

AdipoR1 overexpression can compensates for AdipoR2 depletion

The structure-function studies in C. elegans suggest that the primary difference between PAQR-1 and PAQR-2 lies in their N-terminal regulatory cytoplasmic domains such that PAQR-1 can act without IGLR-2 while PAQR-2 requires an interaction with IGLR-2 for its activation. That PAQR-1 overexpression can suppress paqr-2 mutant phenotypes suggests that it acts either alone or via an interaction partner that is not limiting for PAQR-1 activity. In either case, it is interesting to test whether this relationship is conserved between the human AdipoR1 and AdipoR2. For this purpose, we turned to HEK293 cells where the AdipoR1 and AdipoR2 proteins act as functional homologs of PAQR-2 and are required to prevent membrane rigidification when the cells are cultivated in the presence of palmitic acid [13–16, 35]. Here, we found that overexpression of either AdipoR1 or AdipoR2 in HEK293 cells is sufficient to prevent membrane rigidification in HEK293 cells challenged with 400 μM palmitic acid (Fig 5A–5E). Thus, either AdipoR1 and AdipoR2 act alone in mammalian cells, or the levels of their putative IGLR-2 orthologous partner is not limiting for their activity. Finally, we found that overexpression of AdipoR1 prevents membrane rigidification by 200 μM palmitic acid in cells where AdipoR2 has been silenced (Fig 5F–5H), thus echoing again the C. elegans findings where increased PAQR-1 activity compensates for loss of PAQR-2.

Discussion

Our main findings are: 1) A single amino acid substitution in the cytoplasmic N-terminal domain of PAQR-1, i.e. R109C or R109A, is sufficient to act as a gof mutation. This suggests that the N-terminal cytoplasmic domain has a regulatory function. 2) The paqr-1(et52) gof allele acts through the same pathway as paqr-2 and is a complete paqr-2 mutant suppressor with no obvious adverse effect. Conversely, the paqr-1(tm3262) lof allele has no obvious phenotype on its own but worsens the paqr-2 mutant phenotypes in double mutants. Taken together, these results suggest that paqr-1 and paqr-2 are partially redundant, but that paqr-2 is more important. 3) The paqr-1(et52) gof allele suppresses the mutant phenotypes in paqr-2;iglr-2;paqr-1(et52) triple mutants better than it does in the iglr-2;paqr-1(et52) double mutant, and can also suppress the paqr-2;iglr-2 mutant phenotypes when provided as a transgene. This suggests that the PAQR-1(R109C) protein does not require IGLR-2 for its activity but that PAQR-2 competes with PAQR-1(R109C) for a downstream factor only when IGLR-2 is absent, or that the PAQR-2 protein can interact and interfere with PAQR-1(R109C) when IGLR-2 is absent. 4) Domain swapping experiments suggest that the PAQR-2 transmembrane domains require the presence of IGLR-2 for their activity, but that some essential PAQR-2 activity is still present in the absence of IGLR-2 if the PAQR-1(R109C) N-terminal cytoplasmic domain is attached to the PAQR-2 transmembrane domains. This suggests that the interaction of IGLR-2 and PAQR-2 via their transmembrane domains may allow IGLR-2 to activate PAQR-2 via its regulatory N-terminal cytoplasmic domain. And 5) Overexpression of either AdipoR1 or AdipoR2 confers increased protection against palmitic acid-induced membrane rigidification in HEK293 cells.

The present work sheds new light on the roles of major protein domains in PAQR-1 and PAQR-2. In particular, it is clear that PAQR-1(R109C) and PAQR-2 act through the same pathway and can carry out very similar roles for the maintenance of membrane homeostasis, which suggests that the key physiological functions of these proteins reside within their highly conserved transmembrane domains and/or the conserved membrane-proximal 80 amino acids. The large cytoplasmic N-terminal domain, which is highly divergent between the two proteins, likely has a regulatory function, for example by blocking access to the cytoplasm-facing cavity where the presumed hydrolytic site is located. A similar suggestion was made for the cytoplasmic domain of the AdipoRs: the published crystal structures of the AdipoRs does not include the full-length N-terminal domain but the small portion (i.e. residues 89–120 for AdipoR1) that was included seem to obstruct the access to the cytoplasm-facing cavity, which prompted the authors to suggest that “a much larger opening at helices III–VII would be uncovered on the cytoplasmic side, if the NTR (N-terminal region) was displaced from its present position” [21]. Conceptually, this type of steric regulation is similar to the “ball and chain” mechanism first proposed to regulate sodium channels in neurons, where the cytoplasmic domain of the channel acts as a ball attached by a flexible motif to the channel, which it can block [36–38]. It will be interesting in the future to test whether novel gof alleles may be created by expressing PAQR-1 or PAQR-2 with truncations in their cytoplasmic N-terminal domains, which incidentally harbors membrane localization motifs in the mammalian AdipoRs [39, 40].

PAQR-1, which does not require IGLR-2, may have a low level of constitutive activity; this would explain the observation that over-expression of the wild-type PAQR-1 partially suppresses paqr-2 mutant phenotypes. The PAQR-1(R109C) or PAQR-1(R109A) mutations may displace the regulatory N-terminal domain and thus cause a larger/more frequent access to the active site, and thus act as gof mutations. Arginine residues are important mediators of protein-protein interactions in a wide range of biological processes thanks to the positively charged guanidinium group (-C- (NH2)2⁺) with five hydrogen bond donors [41–44]. Positively charged residues such as arginine are preferentially localized on the surface of proteins where they facilitate the interaction with negatively charged surfaces: about 40% of the salt-bridges in proteins involve ion pairs between arginine and the carboxylate groups of acidic amino acids [45], and the positive charge is also important for planar stacking and polar interactions with aromatic residues [46, 47]. Unsurprisingly, the number of arginines at protein interfaces is larger than expected from random distribution [48], and guanidinium groups are also significantly overrepresented at protein-protein interfaces, as opposed to other polar and charged groups [49]. It is therefore likely that the R109C and R109A substitutions in PAQR-1 have important structural consequences, for example disrupting intramolecular or intermolecular interactions important for the proposed “ball and chain” mechanism. Such a mechanism is also consistent with the gof allele created by fusing the transmembrane domains of PAQR-1 to the cytoplasmic domain of PAQR-2, presumed here to poorly inhibit access to the catalytic site. Our finding that overexpression of AdipoR1 or AdipoR2 is sufficient to protect HEK293 cells from palmitic acid-induced rigidification echoes those of Kupchak et al. who showed that AdipoR1 expression in yeast support its activity and mimic that of the yeast homologs [50], and of Holland et al. who showed that overexpression of AdipoR1 and AdipoR2 in liver also leads to their increased activity, as measured by the ceramidase activity in extracts as well as improved glucose and lipid homeostasis [29]. Clearly, the C. elegans PAQR-1 and the human AdipoR proteins can act in a dose-dependent manner. This suggests that PAQR-1 and the AdipoRs have intrinsic basal activity that is not strictly regulated by the membrane environment. This is interesting given that the tissue expression levels of AdipoR1 and AdipoR2 are quite variable. For example, and although all tissues appear to express either one or both of the AdipoRs, the retina is clearly a tissue with exceptionally high levels of AdipoR1 expression [51]. Not surprisingly then, the retina shows a severe depletion of membrane UFAs associated with retinitis pigmentosa both in mouse models and in human patients [51–53].

In contrast to PAQR-1, the activity of PAQR-2 is dependent on the presence of IGLR-2 and is only required under conditions of membrane rigidification, such as low temperature or an SFA-rich diet. The domain swapping experiments described here suggest that IGLR-2 and PAQR-2 interact primarily via their transmembrane domains but that an important consequence of that interaction occurs at the level of their cytoplasmic domains. Speculatively, docking of IGLR-2 onto PAQR-2 via the transmembrane domains could allow the cytoplasmic domain of IGLR-2 to cause a conformational change within the cytoplasmic domain of PAQR-2 that results in its activation, i.e. displacement of the “ball” that regulates access of substrates (e.g. ceramides as per [22, 28, 29]) to the active site (see model in Fig 6). IGLR-2 is related to the mammalian LRIG protein family, which are characterized by the presence of a single transmembrane domain, one or more immunoglobulin domains and leucine-rich repeats in their large extracellular N-terminal domain, and a cytoplasmic domain that varies in size and sequence. Many LRIG proteins act as regulators of signaling transmembrane proteins. The LRIG1, -2 and -3 proteins for example are important regulators of receptor tyrosine kinases [54–56]. The mammalian LRIG protein AMIGO, which structurally is the most similar to C. elegans IGLR-2, regulates Kv2.1 voltage-gated potassium channels by facilitating their clustering [57]. IGLR-2 could also act as a promoter of PAQR-2 multimerization to create a signaling nexus. Membrane thickening, which typically accompanies rigidification, could promote PAQR-2/IGLR-2 clustering because of the energy costs of membrane deformation around the fixed-length transmembrane domains of the proteins in a situation analogous to that found in GpA helix-dimers [58, 59] or Ire1[12, 60], another membrane fluidity sensor. Considerations regarding clustering are interesting because several independent lines of evidence suggest that the mammalian AdipoR1 and AdipoR2 proteins can form homo- and heteromultimers; these include co-immunoprecipitation of tagged AdipoR1 and AdipoR2 [61], bifluorescence complementation of AdipoR1 in HEK293 cells [62] and fluorescence resonance energy transfer (FRET) also in HEK293 cells [63]. It will be interesting in the future to determine whether a mammalian homolog of IGLR-2 facilitates the multimerization of the AdipoRs. A better understanding of this membrane homeostasis pathway is important because of its potential therapeutic value in many disease contexts where abnormal membrane properties have been noted, including diabetes [64, 65] and cancer [66–69].

Materials and methods

C. elegans strains and cultivation

The wild-type C. elegans reference strain N2 and the mutant alleles studied (except for the novel paqr-1(et52)) created in the present study) are available from the C. elegans Genetics Center (CGC; MN; USA). Unless otherwise stated, the C. elegans strains maintenance and experiments were performed at 20°C using the E. coli strain OP50 as food source, which was maintained on LB plates kept at 4°C (re-streaked every 6–8 weeks) from which single colonies were picked for overnight cultivation at 37°C in LB medium, then used for seeding NGM plates [70]. OP50 stocks were kept frozen at -80°C and new LB plates were streaked every 3–4 months. NGM plates containing 20 mM glucose were prepared using a 1 M stock solution that was filter sterilized then added to the molten NGM media.

The paqr-1(syb1401) allele in which a HA tag is fused in frame to the start codon of paqr-1 was created by Suny Biotech Co (Fuzhou City, China) using CRISPR/Cas9. The altered sequence is as follows (HA coding sequence is in upper case; underlined sequence is synonymous mutation and endogenous sequence of paqr-1 is in lower case):

catgatcactacaatgtattctatcatcttttcttttcatttttttgcaaacatgaagacaaagttcatatttcaggcaaagaatgTACCCATATGATGTCCCGGATTACGCTaatccagatgaggtcaaccgagcccttgggcactacctcaatgacgctgattcaggcgaattggttgtcgaggacagcacaactgtacaggtaggattgaagaagaaaaataatattgatgttaaaattaaaaaacgttcatattttttctaaattatatatcaatt.

The paqr-1(syb364) allele in which a HA tag is fused in frame to the start codon of paqr-1(et52) was created by Suny Biotech Co (Fuzhou City, China) using CRISPR/Cas9. The altered sequence is as follows (HA coding sequence is in upper case; underlined sequence are synonymous mutations and the endogenous sequence of paqr-1(et52) is in lower case):

catgatcactacaatgtattctatcatcttttcttttcatttttttgcaaacatgaagacaaagttcatatttcaggcaaagaatgTACCCATATGATGTCCCGGATTACGCTaatccagatgaggtcaaccgagcccttgggcactacctcaatgacgctgattcaggcgaattggttgtcgaggacagcacaactgtacaggtaggattgaagaagaaaaataatattgatgttaaaattaaaaaacgttcatattttttctaaattatatatcaatt.

The paqr-2(syb1401) allele in which a HA tag is fused in frame to the start codon of paqr-2 was created by Suny Biotech Co (Fuzhou City, China) using CRISPR/Cas9. The altered sequence is as follows (HA coding sequence is in upper case; and the endogenous sequence of paqr-2 is in lower case): atccgctttattctctcacagttccgattttatttgatttttttctggaatttcttatattctcggttgaaaaaaattttaaaaactaaaattcagctttaacaaaatgTACCCATATGATGTCCCGGATTACGCTgaggaagatgacgtggaatcggcaacaccggcggaatcgcaaaaacttttgcaaaaaagcgttcgaaattcgtttgacgag.

Screen for suppressors of SFA intolerance and whole genome sequencing

paqr-2(tm3410) mdt15(et14) double mutant worms were mutagenized for 4 hours by incubation in the presence of 0.05 M ethyl methane sulfonate according to the standard protocol [70]. The worms were then washed and placed on NGM plate. Two hours later, vigorous hermaphrodite L4 animals were transferred to new NGM plates. Five days later, F1 progeny were bleached, washed and their eggs were allowed to hatch overnight in M9 (22 mM KH2PO4, 42 mM Na2HPO4, 85.5 mM NaCl and 1 mM MgSO4). The resulting L1 larvae were transferred to new plates containing 20 mM glucose and then screened 72 hours later for fertile adults, which were picked onto new NGM plates for further analysis.

The isolated suppressor alleles were outcrossed 4 to 6 times prior to whole genome sequencing and 10 times prior to their phenotypic characterization or use in the experiments presented in the manuscript.

The genomes of suppressor mutants were sequenced to a depth of 25-40x as previously described [71]. Differences between the reference N2 genome and that of the mutants were sorted by criteria such as non-coding substitutions, termination mutations, splice-site mutations etc. [72] For each suppressor mutant, one or two hot spots, that is small genomic area containing several homozygous mutations, were identified, which is in accordance to previous reports [73]. Mutations in the hot spot that were still retained after 10 outcrosses were considered candidate suppressors and tested experimentally as described in the text.

Pre-loading of E. coli with palmitic acid

A stock of 0.1 M palmitic acid (Sigma) was dissolved in ethanol and diluted in LB media to a final concentration of 2 mM, inoculated with OP50 bacteria, then shaken overnight at 37°C. The bacteria were then washed twice with M9 to remove traces of palmitic acid and growth media, diluted to equal concentrations based on optical density (OD₆₀₀), concentrated 10x by centrifugation, dissolved in M9 and seeded onto NGM plates lacking peptone (200 μl/plate). Worms were added the following day.

Growth, tail tip scoring and other C. elegans assays

For length measurement studies, synchronized L1s were plated onto test plates seeded with E. coli, and worms were mounted and photographed 72, 96 or 144 hours later. The length of 20 worms were measured using ImageJ [74]. Quantification of the withered tail tip phenotype was done on synchronous 1-day old adult populations, that is 72 h post L1 (n >100) [18]. Other assays starting with 1 day old adults have also previously been described in details: total brood size (n = 5) [30], lifespan (n = 100)[30], defecation (n = 5; average interval between ten defecation was determined for each worm) [75], pharyngeal pumping rate (n = 25, each monitored for 20s)[76], speed of locomotion (n = 5) [77].

RNAi in C. elegans

All strains were grown on control L4440 RNAi bacteria for one generation at 20°C, then synchronized and L1s placed onto assay RNAi, incubated at 15°C and scored on day 6. Feeding RNAi clones were from the Ahringer RNAi library and were sequenced to confirm their identity, and used as previously described [78].

Fluorescence recovery after photobleaching (FRAP) in C. elegans and HEK293 cells

FRAP experiments in C. elegans were carried out using a membrane-associated prenylated GFP reported expressed on intestinal cells as described previously and using a Zeiss LSM700inv laser scanning confocal microscope with a 40X water immersion objective [16, 17]. Briefly, the GFP positive membranes were photobleached over a circular area (seven-pixel radius) using 20 iterations of the 488 nm laser with 50% laser power transmission. Images were collected at a 12-bit intensity resolution over 256 x 256 pixels (digital zoom 4X) using a pixel dwell time of 1.58 μsec, and were all acquired under identical settings. For FRAP in mammalian cells, HEK293 cells were stained with BODIPY 500/510 C1, C12 (Invitrogen) at 2 μg/ml in PBS for 10 min at 37°C [16]. FRAP images were acquired with an LSM880 confocal microscope equipped with a live cell chamber (set at 37° and 5% CO₂) with a 40× water-immersion objective as previously described [16]. The recovery of fluorescence was traced for 25 s. Fluorescence recovery and T_half were calculated as described previously [17].

Western blot for C. elegans

Crowded plates of various C. elegans strains were harvested and washed twice in M9 and then lysed using the following lysis buffer: 25 mM Tris (pH 7.5), 300 mM NaCl, 1% triton X-100 and 1X protease inhibitor. 25 μg of total protein was loaded in each lane. For transgenic animals, 25 worms expressing GFP (used as a transformation marker) were picked, boiled in sample buffer (4x Laemmli sample buffer, BIO-RAD) and loaded on an SDS gel. For detection of protein the nitrocellulose membranes (GE Health care) were blocked with 5% skimmed milk (Blotting-Grade Blocker, BIO-RAD) diluted in PBS-T. Antibody dilutions (primary antibody: rabbit monoclonal anti-HA antibody (C29F4; Cell Signaling Technology) 1:5000, mouse monoclonal anti-TUBULIN (T5168; SIGMA) and secondary antibodies: swine anti-rabbit HRP (1:3000, Dako) or goat anti-mouse HRP (1:3000, Dako)) was done in 5% skimmed milk in TBST and washes were carried out in PBS-T. Detection of the bound antibodies was performed using an ECL detection kit (Immobilon Western; Millipore) and visualized with a digital camera (VersaDoc; Bio-Rad).

Construction of plasmids

pQC20.10 and pQC20.11: paqr-1 and paqr-1(et52) rescue constructs

paqr-1 and paqr-1(et52) rescue constructs were created using the following primers 5´ATCCAATTTGCCCCACTGAAT 3´´and 5´ CACAAAACTCTAGACTACTGG 3´ to amplify the paqr-1 and paqr-1(et52) alleles and 2 kb of upstream regulatory sequence using genomic DNA from lysed worms as template. The resulting 5 kb PCR products were cloned into the pCR XL-TOPO vector using TOPO XL PCR cloning kit (Invitrogen). The resulting plasmids, pQC20.10 and pQC20.11, were injected into N2 worms at 10 ng/μl together with 3 ng/ μl pPD118.33 (Pmyo-2::GFP) used as a transformation marker, a gift from Andrew Fire (Addgene plasmid # 1596) [79].

pQC20.8 and pQC20.9: paqr-1 and paqr-1(et52) translational reporters

The translational reporters were generated with a Gibson assembly cloning kit (NEB) with the following two fragments: (1) The promoter and the coding sequence was amplified from the rescue construct pQC20.10 and pQC20.11 with the following primers 5´GCATGGATGAACTATACAAAAATCCAGATGAGGTCAATCG-3´and 5´- CTCCTTTACTCATTGATGCCATTCTTTGCCTGAAATATGAAC-3´, (2) GFP was amplified from the plasmid Ppaqr-2:paqr-2:GFP[30] using the primers 5´- GTTCATATTTCAGGCAAAGAATGGCATCAATGAGTAAAGGAG-3´ and 5´- CGATTGACCTCATCTGGATTTTTGTATAGTTCATCCATGC-3´. The resulting plasmids, pQC20.8 and pQC20.9 support expression of Ppaqr-1::PAQR-1WT::GFP and Ppaqr-1::PAQR-1(et52)::GFP, respectively, and were injected into N2 worms at 10 ng/μl together with the plasmid 30 ng/μl PRF4, which carries the dominant rol-6(su1006) transformation marker [80].

Domain swapping constructs

pQC20.1 and pQC20.2

Constructs carrying full length paqr-1 or paqr-1(et52) driven from the paqr-2 promoter were generated with a Gibson assembly cloning kit (NEB) by assembly of the following fragments: (1) The promoter and the backbone was amplified from Ppaqr-2:paqr-2:GFP plasmid[30] using the primers 5´-CGTCTGAACGAACAATGTCCAGTTAGATAGGGTAGATTTGTTTC-3´and 5´-GGATTAGCGTAATCCGGGACATCATATGGGTACATTTTGTTAAAGCTGAA-3´and (2) the coding sequences were amplified from the rescue constructs using the following primers 5´-TTCAGCTTTAACAAAATGTACCCATATGATGTCCCGGATTACGCTAATCC-3´ and 5´- GAAACAAATCTACCCTATCTAACTGGACATTGTTCGTTCAGACG-3´. The resulting pQC20.1 and pQC20.2 plasmids were injected into N2 worms at 10 ng/μl together with 40 ng/μl sur-5::gfp [81] used as a transformation marker.

pQC20.3 and pQC20.4

The cytoplasmic N terminal domain of paqr-1 and paqr-1(et52) were fused to the transmembrane and C-terminal domains of paqr-2 using a Gibson assembly cloning kit (NEB) with the following fragments: (1) the paqr-2 promoter and coding sequence for the PAQR-1 N-terminal domain was amplified from the pQC20.1 and pQC20.2 plasmids using the primers 5´-CTAAAATTCAGCTTTAACAAAATGAATCCAGATGAGGTCAATCGAGCCCTTGG-3´and 5´-CATATGAGTCCAAATGTTTCCTGTTTCCGTGTGCAGTGACCAAATA-3´and (2) the paqr-2 transmembrane and C-terminal coding sequences were amplified from the Ppaqr-2:paqr-2:GFP plasmid[30] using the primers 5´- TATTTGGTCACTGCACACGGAAACAGGAAACATTTGGACTCATATG-3´ and 5´- CCAAGGGCTCGATTGACCTCATCTGGATTCATTTTGTTAAAGCTGAATTTTAG-3´. The resulting pQC20.3 and pQC20.4 plasmids were injected into N2 worms at 10 ng/μl together with 40 ng/μl sur-5::gfp[81] used as a transformation marker.

pQC20.5 and pQC20.6

The N-terminal cytoplasmic domain and the transmembrane domains encoded by the paqr-1 wild-type and et52 alleles were fused to the C-terminus of paqr-2 using a Gibson assembly cloning kit (NEB) with the following fragments: (1) the promoter and N-terminal/transmembrane domain-coding sequences were amplified from the pQC20.1 and pQC20.2 constructs using the primers 5´-TTCAGTTGATCCGGGCAGGATCCTTTGTTCAGACGAGCAAATGC-3´ and 5´-ACCGGCGGATGTTGGTTTATGAGGTAGATTTGTTTCCAAT-3´ and (2) The paqr-2 -C terminus coding sequence was amplified from the pQC20.3 and pQC20.4 plasmids using the following primers 5´-GCATTTGCTCGTCTGAACAAAGGATCCTGCCCGGATCAACTGAA-3´ and 5´-ATTGGAAACAAATCTACCTCATAAACCAACATCCGCCGGT-3´. The resulting pQC20.5 and pQC20.6 plasmids were injected into N2 worms at 10 ng/μl together with 40 ng/μl sur-5::gfp[81] used as a transformation marker.

pQC20.7

The N terminal cytoplasmic domain encoded by paqr-2 was fused to transmembrane domains and C terminus of paqr-1 wild-type using a Gibson assembly cloning kit (NEB) with the following fragments: (1) the N terminal of paqr-2 was amplified from the wild-type worms that are CRISPR modified and carrying an HA tag at the N terminus (soon after ATG) using the following primers 5`CTAAAATTCAGCTTTACAAAATGTACCCATATGATGTCCCGGATTACGCT 3`and 5`GTGTCCAAATGTTACCAGTTTCAGTGTGCAATGCAAAAATA 3`and (2) the paqr-1 transmembrane domains and C terminus was amplified from pQC20.1 plasmid using the following primers 5`AGCGTAATCCGGGACATCATATGGGTACATTTTGTTAAAGCTGAATTTTAG 3`and 5`TATTTTTGCATTGCACACTGAAACTGGTAACATTTGGACAC 3`. The resulting pQC20.7 plasmid was injected into N2 worms at 10 ng/μl together with 40 ng/μl sur-5::gfp used as a transformation marker.

pQC20.12, a PAQR-1(R109A) expression construct

Arginine at position 109 of the PAQR-1 coding sequence was replaced by alanine using a site directed mutagenesis kit (NEB). The primers used were 5´-GCCCGTAAAAAGGGAGGGCAAT-3´and 5´-ATATCTGAAAAATTATGCG-3´ and the wild-type paqr-1 rescue construct (pQC20.10) was used as template. The resulting plasmid, pQC20.12, was injected into N2 worms at 10 ng/μl together with 40 ng/μl sur-5::gfp[81] used as a transformation marker.

Bimolecular fluorescence complementation (BiFC) constructs and analysis

pCE-IGLR2-VC155, pCE-VN173-PAQR2 and pCE-VN173-PAQR1 WT plasmids were used from a previously published paper [17]. pCE-VN173-PAQR1 R109C construct was generated by Q5 site directed mutagenesis kit (NEB) using the following primers 5`TTTCAGATATTGTCGTAAAAA 3`and 5`AATTATGCGAATTTTTAAAA 3`. Wild type version of PAQR-1 plasmid was used as a template. The different combinations of BiFC plasmids were injected into N2 worms at 15 ng/μl each, together with pRF4(rol-6) at 100 ng/μl as previously described [82]. Expression of the BiFC constructs were induced by heat shocks of 2.5 h and 1.5 h at 33°C, with 2 h recovery at 20°C in between. Scoring of fluorescence was preformed after 16 h of recovery at 20°C.

C. elegans lipidomics

For worm lipidomics, samples were composed of synchronized L4 larvae (one 9 cm diameter plate/sample; each treatment/genotype was prepared in four independently grown replicates) grown overnight on OP50-seeded NGM or NGM containing 20 mM glucose. Worms were washed 3 times with M9, pelleted and stored at -80°C until analysis. For lipid extraction, the pellet was sonicated for 10 minutes in methanol and then extracted according to BUME method [83]. Internal standards were added in the chloroform phase during the extraction. Lipid extracts were evaporated and reconstituted in chloroform: methanol (1:2) with 5 mM ammonium acetate. This solution was infused directly (shotgun approach) into a QTRAP 5500 mass spectrophotometer (ABSciex) equipped with the Nanomate Triversa (Advion Bioscience) as described previously[84]. Phospholipids were measured using multiple precursor ion scanning [85, 86]. The data was evaluated using the LipidProfiler software [85]. The full lipidomics data is provided as in the file S1 Lipidomics.

Cultivation of HEK293 cells

HEK293 (identity verified by STR profiling and tested free of mycoplasma) cells were grown in DMEM containing 1 g/l glucose, pyruvate, and GlutaMAX and supplemented with 10% FBS, 1% nonessential amino acids, 10 mM HEPES, and 1% penicillin and streptomycin (all from Life Technologies) at 37°C in a water-humidified 5% CO2 incubator. Cells were sub-cultured twice a week at 90% confluence. For FRAP, the cells were seeded in glass-bottom dishes (Ibidi) pre-coated with 0.1% porcine gelatin (Sigma-Aldrich).

AdipoR1/2 constructs and over-expression

The human AdipoR1 cDNA (NP_001277482.1) was cloned between the NheI and BspEI sites of the pIRESneo2 vector (Clontech, Takara Bio) including the N-terminal VSV-tag and C-terminal FLAG-tag. The human AdipoR2 cDNA (NP_001362293.1) was cloned between the Nhe I and XmaI sites of the pIREShyg2 vector (Clontech, Takara Bio) including the N-terminal HA tag and C-terminal c-Myc tag. HEK293 cells were transfected using Viromer Red according to the manufacturer’s instructions 1X protocol (Lipocalyx).

siRNA in HEK293 cells

The predesigned siRNA: AdipoR2 J-007801-10 and Non-target D-001810-10 were purchased from Dharmacon. HEK293 cell transfection was performed in complete media using 25 nM siRNA and Viromer Blue according to the manufacturer’s instructions 1X (Lipocalyx). Knockdown gene expression was verified 48 h after transfection.

Quantitative PCR in HEK293 cells

Total cellular RNA was isolated using RNeasy Kit according to the manufacturer’s instructions (Qiagen) and quantified using a NanoDrop spectrophotometer (ND-1000; Thermo Fisher Scien- tific). cDNA was obtained using a RevertAid H Minus First Strand cDNA Synthesis Kit with random hexamers. Quantitative PCR (qPCR) was performed with a CFX Connect thermal cycler (Bio-Rad) using HOT FIREpol EvaGreen qPCR Super- mix (Solis Biodyne) and standard primers. Samples were measured as triplicates. The relative expression of each gene was calculated according to the delta-deltaCT method[87]. Expression of the housekeeping gene PPIA was used to normalize for variations in RNA input. PPIA and AdipoR2 primers were: AdipoR2 forward, TCATCTGTGTGCTGGGCATT; Adipo2 reverse, CTATCTGCCCTATGGTGGCG; PPIA forward, GTCTCCTTTGAGCTGTTTGCAG; PPIA reverse, GGACAAGATGCCAGGACCC.

Western blot for HEK293 cells

Cellular proteins were extracted using a lysis buffer (1% Nonidet P-40, 0.1% SDS, 10% glycerol, 1% sodium deoxycholate, 1 mM DTT, 1 mM EDTA, 100 mM HEPES, 100 mM KCl) containing Halt Protease Inhibitor Cocktail (1X; Pierce) on ice for 10 min. Upon lysis completion, cell lysates were centrifuged at 13 000 rpm for 10 min at 4°C. The soluble fraction was kept for further analysis, and the protein sample concentration was quantified using the BCA protein assay kit (Pierce) according to the manufacturer’s instructions. Twenty micrograms of protein were mixed with Laemmli sample loading buffer (Bio-Rad), heated to 37°C for 10 min, and loaded in 4% to 15% gradient precast SDS gels (Bio-Rad). After electrophoresis, the proteins were transferred to nitrocellulose membranes using Trans-Blot Turbo Transfer Packs and a Trans-Blot Turbo apparatus/predefined mixed-MW program (Bio-Rad). Blots were blocked with 5% nonfat dry milk in PBS-T for 1 h at room temperature. Blots were incubated with primary antibodies overnight at 4°C: rabbit monoclonal anti-HA antibody (C29F4; Cell signaling Technology) 1:5000 dilution, mouse monoclonal anti-FLAG antibody (F3165; SIGMA) 1:3000 dilution, mouse monoclonal anti-VSV antibody (V5507; SIGMA) 1:3000 and rabbit anti-GAPDH antibody (14C10; Cell Signaling Technology). Blots were then washed with PBS-T and incubated with either swine anti-rabbit HRP (1:3000, Dako) or goat anti-mouse HRP (1:3000, Dako)) and washed again with PBS-T. Detection of the hybridized antibody was performed using an ECL detection kit (Immobilon Western; Millipore), and the signal was visualized with a digital camera (VersaDoc; Bio-Rad).

Palmitic acid treatment of HEK293 cells

Palmitic acid was dissolved in sterile DMSO (Sigma-Aldrich) and then mixed with fatty acid-free BSA (Sigma-Aldrich) in serum-free medium for 15 min at room temperature. The molecular ratio of BSA to palmitic acid was 1:5.3 when PA 400 μM and 1:2.65 when PA was 200 μM. Cells were then cultivated in serum-free media containing palmitic acid for 6 h prior to analysis.

Statistics

Unless otherwise stated, means and standard error of the means are presented, and t-tests were used to identify significant differences. For the tail tip defect, significant differences were determined using Z-tests. All experiments were repeated several times with similar results. Asterisks are used in the figures to indicate various degrees of significance, where *: p<0.05; **: p<0.01; and ***: p<0.001.

Supporting information

S1 Fig [a]
The allele suppresses most or all mutant phenotypes, and an R109A substitution also acts as a allele.

S2 Fig [a]
The allele suppresses several membrane-related phenotypes of the mutant.

S3 Fig [a]
Genetic requirements for function.

S4 Fig [a]
BiFC shows that PAQR-2, but not PAQR-1, interacts with IGLR-2.

S5 Fig [tiff]
Domain swapping experiments indicate that the intracellular domains of PAQR-2 and IGLR-2 are likely regulatory.

S1 Lipidomics [xlsx]
Original lipidomics data.