FORWARD GENETICS IN C. ELEGANS REVEALS GENETIC ADAPTATIONS TO POLYUNSATURATED FATTY ACID DEFICIENCY

Abstract

Polyunsaturated fatty acids (PUFAs) are essential for mammalian health and function as membrane fluidizers and precursors for signaling lipids though the primary essential function of PUFAs within organisms has not been established. Unlike mammals who cannot endogenously synthesize PUFAs, C. elegans can de novo synthesize PUFAs starting with the Δ12 desaturase FAT-2 which introduces a second double bond to monounsaturated fatty acids to generate the PUFA linoleic acid. FAT-2 desaturation is essential for C. elegans survival since fat-2 null mutants are non-viable; the near-null fat-2(wa17) allele synthesizes only small amounts of PUFAs and produces extremely sick worms. Using fluorescence recovery after photobleaching (FRAP), we found that the fat-2(wa17) mutant has rigid membranes and can be efficiently rescued by dietarily providing various PUFAs, but not by fluidizing treatments or mutations. With the aim of identifying mechanisms that compensate for PUFA-deficiency, we performed a forward genetics screen to isolate novel fat-2(wa17) suppressors and identified four internal mutations within fat-2, and six mutations within the HIF-1 pathway. The suppressors increase PUFA levels in fat-2(wa17) mutant worms and additionally suppress the activation of the daf-16, UPR^er and UPR^mt stress response pathways that are active in fat-2(wa17) worms. We hypothesize that the six HIF-1 pathway mutations, found in egl-9, ftn-2, and hif-1 all converge on raising Fe²⁺ levels and in this way boost desaturase activity, including that of the fat-2(wa17) allele. We conclude that PUFAs cannot be genetically replaced and that the only genetic mechanism that can alleviate PUFA-deficiency do so by increasing PUFA levels.

Fig 1.. C. elegans fatty acid synthesis pathway and FAT-2 desaturase.

(A) Simplified pathway of fatty acid synthesis and desaturation in C. elegans. Boxes indicate the name of the enzymes, FAT-2 desaturase is indicated in a red box. Fatty acids of which the synthesis is dependent on FAT-2 are indicated in red. Fatty acid abbreviations are as follow: palmitic acid (PA), palmitoleic acid (POA), vaccenic acid (VA), stearic acid (SA), oleic acid (OA), linoleic acid (LA), alpha-linolenic acid (ALA), gamma-linolenic acid (GLA), stearidonic acid (STA), dihomo-gamma-linolenic acid (DGLA), eicosatetraenoic acid (ETA), arachidonic acid (AA), and eicosapentaenoic acid (EPA). (B) AlphaFold2 predicted FAT-2 structure with the serine at position 101 indicated with a red arrow. (C) Same structure as in B, zoomed in and angled to show that the S101 position that is mutated to phenylalanine in the fat-2(wa17) allele lies in a loop connecting two alpha helices.

Fig 2.. Characterization and rescue of fat-2(wa17).

(A) Introduction of the wild-type fat-2(+) allele on an extrachromosomal array rescues the fat-2(wa17) growth defect. (B-C) FRAP curve and T_half value show that fat-2(wa17) has rigid membranes similar to paqr-2(tm3410) control. (D-I,K-M) The lengths of fat-2(wa17) worms grown from L1 stage for 72h in the indicated conditions; horizontal dashed lines indicate the approximate lengths of the synchronized L1s at the start of the experiments. n=20 for each genotype/condition. (J) FRAP T_half values show that NP-40 rescues fat-2(wa17) rigid membranes similar to paqr-2(tm3410). Error bars show the standard error of the mean. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control.

Fig 3.. Lipidomic analysis of fat-2(wa17) mutant.

(A) SFA, MUFA, and PUFA levels in phosphatidylcholines (PCs) of fat-2(wa17) grown in various conditions. Note that cultivation on 2mM LA boosts PUFA levels. LA to NGM worms were grown on 2 mM LA before being transferred to NGM 6 h prior to harvesting. (B) Heatmap of PC species in fat-2(wa17) in all conditions. (C) Levels of individual FA species in PCs for all conditions. Inset shows levels of 20:5 FA are increased by providing fat-2(wa17) with linoleic acid. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control.

Fig 4.. Membrane fluidizing mutations partially rescue fat-2(wa17).

(A-D) Fluidizing paqr-2(tm3410) suppressor mutations only slightly rescue fat-2(wa17) growth. Dashed horizontal lines indicate approximate length of L1s at the start of the experiments; length was measured 72 h post-synchronization. Error bars show the standard error of the mean. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control. (E-G) Oil Red O staining of day 1 adults shows that the high lipid abundance in fat-2(wa17) is not suppressed by paqr-2(tm3410) fluidizing mutations.

Fig 5.. A forward genetic screen reveals fat-2(wa17) is suppressed by mutations in the HIF-1 pathway.

(A) Overview of the forward genetics screen strategy to isolate fat-2(wa17) suppressors. (B) Identity and position of the fat-2(wa17) suppressors as well as the positions of functional domains. Novel mutations are marked by a black triangle with corresponding allele name and mutation effect; the red triangle in FAT-2 indicates the original wa17 allele. Gene names in red represent loss- or reduction-of-function mutations; gene names in green represent gain-of-function mutations. (C) Proposed pathway of fat-2(wa17) suppression by mutations in the HIF-1 pathway. Reduction of EGL-9 constitutively activates HIF-1, HIF-1 activation inhibits FTN-2. The loss of FTN-2 increases the levels of Fe²⁺ outside of iron pools, thus boosting FAT-2 desaturase activity. Gain-of-function mutations are labeled in green, loss- or reduction-of-function mutations are labeled in red. (D) Length of all fat-2(wa17) suppressors measured 72 h after L1 stage. (E) Representative images of fat-2(wa17) suppressors after 72 h of growth. (F-H) Null alleles of egl-9 and hif-1 do not rescue fat-2(wa17), but the null allele of ftn-2 does, confirming that ftn-2(et67) and ftn-2(et68) are loss-of-function alleles. Lengths measured 72 h after L1 synchronization. (I) ftn-2(et68) rescue of fat-2(RNAi) worms confirming that the suppressors are not wa17 specific. The horizontal dashed line indicates the approximate length of L1s at the start of each experiment. Error bars show the standard error of the mean. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control.

Fig 6.. fat-2(wa17) suppressors belong in HIF-1 pathway and influence HIF-1 levels.

(A) Western blot confirming that hif-1::3xFLAG levels in fat-2(wa17) is increased by egl-9(et60), but not by ftn-2(et68). Hypoxia treatment increases HIF-1 levels in WT and fat-2(wa17), confirming successful protein tagging. (B) Quantification of Western blot in A showing normalized relative intensity of the HIF-1 signal to that of tubulin. (C) mRNA expression of FTN-2 confirming that hif-1(et69) reduces FTN-2 levels. (D) Western blot confirming that fat-2::HA levels in fat-2(wa17) are greatly reduced but increased in suppressor strains. (E) Quantification of Western blot in D showing normalized relative intensity of the FAT-2 signal to that of tubulin.

Fig 7.. Lipidomic analysis of fat-2(wa17) suppressors reveals that PUFA levels are increased.

(A) Levels of SFAs, MUFAs, and PUFAs in PCs measured in fat-2(wa17) suppressors confirming that the suppressors increase PUFA levels in fat-2(wa17). Worms were homozygous for all indicated genotypes but note that the hif-1(et69) allele suppresses fat-2(wa17) best in a heterozygous state. (B) Heat map analysis of PC species in suppressor mutants. (C) Levels of individual FA species in PCs in fat-2(wa17) suppressors, insert shows that levels of C20:5 are significantly increased in all double mutant strains. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control. Note that the N2 and fat-2(wa17) samples are the same as in Fig 3.

Fig 8.. ftn-2(et68) rescues fat-2(wa17)’s stress responses.

(A) Representative image of a FRAP experiment, showing pGLO-1::GFP-CAAX-positive intestinal membranes. The rectangle indicates the bleached area. (B-C) T_half values and FRAP curves showing that ftn-2(et68);fat-2(wa17) has less rigid membranes than fat-2(wa17). (D-E) Representative images and quantification of ftn-2(et68) rescue of fat-2(wa17) mitochondrial stress with a hsp-60::gfp reporter. atfs-1(et15) serves as a control for high mitochondrial UPR activation. (F-G) Representative images and quantification of DAF-16::GFP localization showing that the DAF-16 stress response is constitutively active in the fat-2(wa17) mutant but normalized by ftn-2(et68). Chi-squared test shows that fat-2(wa17) is significantly different from WT. (H-I) Representative images and quantification of mild ER stress in fat-2(wa17) that is slightly rescued by ftn-2(et68) using a hsp-4::gfp reporter. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control.

Fig 9.. Exogenous treatments that mimic fat-2(wa17) suppressors partially rescue fat-2(wa17).

(A-E) Length assay of fat-2(wa17) cultivated with different treatments for 72 h after L1 stage synchronization. The horizontal dashed line represents the approximate length of L1 worms at start of each experiment. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant differences compared to the fat-2(wa17) control.