Heterozygous variants in PLCG1 affect hearing, vision, cardiac, and immune function

Abstract

Phospholipase C isozymes (PLCs) hydrolyze phosphatidylinositol 4,5-bisphosphate (PIP₂) into inositol 1,4,5-trisphosphate (IP₃) and diacylglycerol (DAG), important signaling molecules involved in many cellular processes including Ca²⁺ release from the endoplasmic reticulum (ER). PLCG1 encodes the PLCγ1 isozyme that is broadly expressed. Hyperactive somatic mutations of PLCG1 are observed in multiple cancers, but only one germline variant has been reported. Here, we describe seven individuals with heterozygous missense variants in PLCG1 [p.(Asp1019Gly), p.(His380Arg), p.(Asp1165Gly), and p.(Leu597Phe)] who present with hearing impairment (5/7), ocular pathology (4/7), cardiac septal defects (3/6), and various immunological issues (5/7). To model these variants in vivo, we generated the analogous variants in the Drosophila ortholog, small wing (sl). We created a null allele sl^T2A and assessed its expression pattern. sl is broadly expressed, including wing discs, eye discs, and a subset of neurons and glia. sl^T2A mutant flies exhibit wing size reductions, ectopic wing veins, and supernumerary photoreceptors. We document that mutant flies also exhibit a reduced lifespan and age-dependent locomotor defects. Expressing wild-type sl in sl^T2A mutant flies rescues the loss-of-function phenotypes, whereas the variants increase lethality. Ectopic expression of an established hyperactive PLCG1 variant, p.(Asp1165His) in the wing pouch causes elevated Ca²⁺ activity and severe wing phenotypes. These phenotypes are also observed when the p.(Asp1019Gly) or p.(Asp1165Gly) variants are overexpressed in the wing pouch, arguing that these are gain-of-function variants. However, the wing phenotypes associated with p.(His380Arg) or p.(Leu597Phe) overexpression are either mild or only partially penetrant. Our data suggest that the heterozygous missense variants reported here affect protein function differentially and contribute to the clinical features observed in the affected individuals.

Keywords: D. melanogaster; Drosophila; Phospholipase C; genetics; genomics; human; medicine; phenotypic heterogeneity.

Figure 1.. The PLCG1 ortholog is small wing (sl) in Drosophila.

(A) Schematic of human PLCG1 and fly Sl protein domains and positions of the variants identified in the affected individuals. Domain prediction is based on annotation from NCBI. (B) Alignment of protein domains near variants of PLCG1 and PLCG2 with PLCG1 from other species. The variants are marked with boxes. All the variants affect conserved amino acids (labeled in red). Isoforms for alignment: Human PLCG1 NP_877963.1; Human PLCG2 NP_002652.2; Mouse Plcg1 NP_067255.2; Zebrafish plcg1 NP_919388.1; Fly sl NP_476726.2. (C) Schematic of fly sl genomic span, transcript, alleles, and the 92 kb genomic rescue (GR) construct. Loss-of-function alleles of sl including sl² (13 bp deletion, Thackeray et al., 1998), sl^KO (CRISPR-mediated deletion of the gene span; Trivedi et al., 2020), and sl^T2A (T2A cassette inserted in the first intron; Lee et al., 2018) are indicated. The T2A cassette in sl^T2A is flanked by FRT sites and can be excised by Flippase to revert loss-of-function phenotypes. GAL4 expression in sl^T2A is driven by the endogenous sl promoter, allowing assessment of sl gene expression pattern with a UAS-mCherry.nls reporter line. This system also allows in vivo modeling of proband-associated variants by crossing with human PLCG1 cDNAs or corresponding fly sl cDNAs. The primer pair used for real-time PCR is indicated.

Figure 1—figure supplement 1.. Human PLC genes and orthologs in Drosophila.

In humans, at least 13 genes encoding Phospholipase C isozymes have been identified and classified into 6 protein families, whereas in Drosophila, 2 PLCβ genes and 1 PLCγ gene have been identified. The small wing (sl) gene is the fly ortholog of human PLCG1 and PLCG2.

Figure 1—figure supplement 2.. Real-time PCR reveals that sl^T2A is a loss-of-function allele causing severely reduced mRNA expression of sl.

Relative sl mRNA expression levels are <5% and<1% in sl^T2A and sl^KO mutant larvae when compared to controls (w¹¹¹⁸). The primers used for real-time PCR are shown in Figure 1C. Each dot represents a replicate per genotype. Unpaired t test, ***p<0.001, mean ± SEM.

Figure 2.. sl^T2A is a loss-of-function allele that affects fly wing and eye development.

(A) sl expression in wing and eye discs. Expression of UAS-mcherry.nls (red) was driven by sl^T2A to label the nuclei of the cells that expressed sl. sl is expressed in the 3rd instar larval wing disc (left) and eye disc (right). A higher magnification image of the wing disc pouch region indicated the by dashed rectangle is shown. The posterior/anterior and dorsal/ventral compartment boundaries are indicated by dashed lines in yellow. Scale bars, 100 μm. (B) sl^T2A cause a wing size reduction and ectopic veins (arrow) in hemizygous mutant male flies. The wing phenotypes can be rescued by introduction of a genomic rescue (GR) construct or the expression of Flippase. Scale bars, 0.5 mm. The quantification of adult wing size is shown in the right panel. Each dot represents the measurement of one adult wing sample. Unpaired t test, ∗∗∗∗p<0.0001, mean ± SEM. (C) sl^T2A causes extra photoreceptors (arrows) in the hemizygous mutant flies. The eye phenotype can be rescued by introduction of a genomic rescue (GR) construct. The photoreceptor rhabdomeres stain positive for phalloidin labeling F-actin. Scale bars, 10 μm. The quantification is shown in the right panel. Each dot represents the measurement of one retina sample. Unpaired t test, ****p<0.0001, mean ± SEM.

Figure 2—figure supplement 1.. Trans-heterozygous sl mutant flies exhibit defects in wing and eye development.

(A) Representative images showing that sl^T2A/sl² trans-heterozygous mutant flies have smaller wing and ectopic veins (indicated by arrow). Scale bars, 0.5 mm. (B) Representative images showing that sl^T2A/sl^KO trans-heterozygous mutant flies have extra photoreceptors (indicated by arrows). The schematic of the section of a normal ommatidia presenting seven photoreceptors (the R8 photoreceptor is not visible in such section) is shown. The photoreceptor rhabdomeres stain positive for phalloidin labeling F-actin. Scale bar, 10 μm.

Figure 3.. sl is expressed in a subset of neurons and glia in the CNS, and loss of sl causes behavioral defects.

(A) Expression pattern of sl in the central nervous system observed by sl^T2A-driven expression of UAS-mCherry.nls reporter (red). In either larval or adult brain, sl is expressed in a subset of fly neurons and glia, which were labeled by pan-neuronal marker Elav (green, upper panel) and pan-glia marker Repo (green, lower panel). Higher magnification images of the regions indicated by dashed rectangles are shown. Scale bars, 20 μm in the magnified images, 50 μm in other images. (B) Loss of sl causes defects in longevity and locomotion. sl^T2A hemizygous flies have a shorter lifespan than w¹¹¹⁸ control flies. The median lifespan of sl^T2A and w¹¹¹⁸ flies is 40 days and 62 days, respectively. The shorter lifespan of sl^T2A flies can be rescued by a UAS transgene that expresses the wild-type sl cDNA (sl^WT). Fly locomotion was assessed by climbing assay (see Materials and methods). sl^T2A flies at the age of 7 days show reduced locomotion and become almost immotile at the age of 35 days. The reduced locomotion ability in sl^T2A flies can be fully rescued by sl^WT. For lifespan assay, Longrank test, ****p<0.0001; sample size n=114, 115, and 125 for w¹¹¹⁸, sl^T2A, and sl^T2A>sl^WT flies, respectively. For climbing assay, each dot represents a measurement of one vial containing 17–22 flies for test. Unpaired t test, ****p<0.0001, mean ± SEM.

Figure 4.. The human and corresponding fly variants are toxic when expressed in flies.

(A) Summary of the viability associated with expression of sl cDNAs in sl^T2A mutant or heterozygous flies. Cross strategy: heterozygous sl^T2A female flies were crossed to male flies carrying UAS-cDNAs or control (UAS-Empty) constructs, or crossed to the y w males as an extra control. The percentages of hemizygous sl^T2A/Y male progeny (red) or sl^T2A/yw heterozygous female progeny (blue) that express different UAS-cDNA constructs were calculated. The expected Mendelian ratio is 0.25 (indicated by the green line in the graph). The fly analogue variants of the proband-associated variants were tested. Each dot represents one independent replicate. Unpaired t test, ****p<0.0001, **p<0.01, *p<0.05, ns: not significant, mean ± SEM. (B) Summary of the viability associated with the expression of PLCG1 cDNAs in sl^T2A mutant (red, males) or heterozygous (blue, females) flies. The same cross strategy and progeny ratio measurement described in (A) was applied. The proband-associated variants, as well as three previously reported PLCG1 variants were assessed. We also included the PLCG2 reference cDNA. Each dot represents one independent replicate. Unpaired t test, **p<0.01, *p<0.05, ns: not significant, mean ± SEM.

Figure 4—figure supplement 1.. Proband-associated variants exhibit deleterious impacts in adults.

(A) Cross strategy of the temperature shifting assay to express UAS-sl cDNAs only in the adult stages. GAL80^ts is a temperature-sensitive inhibitor of GAL4, which allows temporal control of UAS- transgene expression by temperature manipulation. GAL80^ts represses GAL4 activity at the permissive temperature (18°C) but becomes inactive at restrictive temperature (29°C), thereby allowing GAL4 to activate expression of UAS-transgenes. In this assay, Tub-GAL4, Tub-GAL80^ts driver was used to express wild-type or variant sl cDNAs specifically in adult flies by shifting the temperature after eclosion. (B) Adult-stage expression of sl variants induces lifespan reduction. The adult flies with expression of sl^D1184G mostly die within 1 week after eclosion. The flies expressing sl^D1041G or sl^H384R have median lifespans of approximately 21 days and 26 days, respectively, both shorter than those expressing sl^WT, which have a median lifespan of 31 days. The sample size for each genotype is indicated (n). Longrank test, ****p<0.0001. (C) Adult-stage expression of sl^D1041G induces some locomotion defect. Flies aged to 20 days were assessed for locomotor ability using a climbing assay. The expression of sl^H384R did not induce significant difference compared to sl^WT. We did not test sl^D1184G as the flies are short-lived. Each dot represents the measurement of one fly. Unpaired t test, *p<0.05, ns: not significant, mean ± SEM. These results suggest that the two strong variants (sl^D1041G and sl^D1184G) contribute to both developmental and acute effects while the sl^H384R variant mainly contributes to developmental stages.

Figure 4—figure supplement 2.. The toxicity of human PLCG1 cDNAs in flies correlates with expression levels.

(A) Summary of the eclosion rate of animals expressing sl cDNAs in sl^T2A mutant (left panel) or heterozygous (right panel) flies at 22°C. The phenotypes are similar to, but slightly milder compared to the same assays performed at 25°C (Figure 4B). Expression of PLCG1^D1019G or PLCG1^D1165G caused lethality in both conditions, whereas expression of PLCG1^Reference, PLCG1^H380R, or PLCG1^L597F led to reduced eclosion rate in mutant hemizygous but not in heterozygous flies. Each dot represents one independent replicate. Unpaired t test, ****p<0.0001, ***p<0.001, *p<0.05, ns: not significant, mean ± SEM. (B) Summary of the viability of expressing PLCG1^Reference ubiquitously using Tub-GAL4. Expression levels of the UAS-PLCG1^Reference transgenes can be manipulated at different temperatures. (C) Summary of the viability of expressing PLCG1 reference and variant cDNAs using a weak ubiquitous driver da-GAL4. Expression levels of the UAS- PLCG1^Reference transgenes can be manipulated at different temperatures. (D) Representative images of the pupae of da-GAL4>UAS-PLCG1 cDNAs flies at 25°C. Expression of PLCG1^D1019G or PLCG1^D1165G caused reduced pupal size. Scale bars, 0.5mm.

Figure 4—figure supplement 3.. Expressing human PLCG1 does not rescue wing or eye phenotypes associated with sl^T2A.

Representative images showing the adult wings (upper panel) and the photoreceptors (lower panel) expressing PLCG1^Reference or sl^WT. Expression of sl^WT rescues the loss-of-function phenotypes including wing size reduction, ectopic veins (indicated by red arrows) and extra photoreceptors (indicated by yellow arrows), whereas expression of the PLCG1^Reference or UAS-Empty shows no rescue. Scale bars, 0.5 mm for the wing images and 10 μm for the photoreceptor images. The photoreceptor rhabdomeres stain positive for phalloidin labeling F-actin. Quantification of the wing size (upper right panel) and the photoreceptors (lower right panel) is shown. Each dot represents a measurement of one wing or retina sample, respectively. Unpaired t test, ***p<0.001, *p<0.05. ns: not significant.

Figure 5.. Ectopic expression of PLCG1 variants causes variable phenotypes.

(A) The Ca²⁺ reporter CaLexA.GFP was expressed in the wing disc pouch, simultaneously with the PLCG1 cDNAs. Expression of PLCG1^D1019G or PLCG1^D1165G caused elevated CaLexA.GFP signal (green), indicating increased intracellular Ca²⁺ levels, indicating that these variants are hyperactive. Nuclei were labeled with DAPI (blue). Scale bars, 100 μm. (B) Representative images of the adult wing blades showing the morphological phenotypes caused by wing-specific expression of PLCG1 cDNAs. Expression of PLCG1^D1019G or PLCG1^D1165G caused severe wing morphology defects including notched margin (arrows) and fused/thickened veins (arrowheads). Expression of PLCG1^L597F exhibited partial penetrance. Expression of PLCG1^H380R exhibited very mild phenotypes, comparable to PLCG1^Reference. Scale bars, 0.5 mm. (C) Representative images showing that eye-specific expression of PLCG1^Reference or PLCG1^H380R causes an ~15% eye size reduction compared to the UAS-Empty control construct, and expression of PLCG1^L597F further reduced eye size. Expression of PLCG1^D1019G or PLCG1^D1165G causes a severe size reduction by ~30%. Scale bars, 100 μm.

Figure 5—figure supplement 1.. Intracellular Ca²⁺ reporter assay suggests that the p.(Asp1019Gly) and p.(Asp1165Gly) are hyperactive variants.

(A) Three previously reported PLCG1 variants (PLCG1^H380A, PLCG1^D1165H, and PLCG1^S1021F) were tested as controls. The Ca²⁺ reporter CaLexA.GFP was co-expressed with PLCG1 cDNAs in the wing disc pouch region. Expression of PLCG1^D1165H or PLCG1^S1021F caused elevated CaLexA.GFP signal (green), whereas expression of PLCG1^H380A did not. Nuclei were labeled with DAPI (blue). Insets show merged images of DAPI and GFP channels. Scale bars, 100 μm. (B) The fly variants were analyzed using the CaLexA.GFP reporter. Expression of sl^D1041G or sl^D1184G caused elevated CaLexA.GFP signal (green), similar to the corresponding PLCG1 variants. Note that the wing discs expressing sl^D1184G are morphologically aberrant and have a diminished GFP signal. Nuclei were labeled with DAPI (blue). Insets show merged images of DAPI and GFP channels. Scale bars, 100 μm.

Figure 5—figure supplement 2.. Wing and eye phenotypes associated with ectopic expression of PLCG1 variants.

(A) Quantification of the wing blade size in non-notched wings from the samples overexpressing PLCG1^Reference or sl^WT. Wing-specific expression of PLCG1^Reference or sl^WT caused an approximate 15% wing size reduction compared to the UAS-Empty control construct. Each dot represents one measured adult wing. Unpaired t test, ****p<0.0001, mean ± SEM. (B) Quantification of the percentage of adult wings with margin notching phenotype. Wing-specific expression of PLCG1^Reference caused ~10% wings to display a notched margin. Expression of PLCG1^H380R or PLCG1^L597F caused approximately 18% and 23% of the wings to display notched margins, respectively. Expression of the enzymatic-dead PLCG1^H380A caused less than 10% wings with notched margin, whereas expression of the hyperactive PLCG1^S1021F is not significantly different from PLCG1^Reference. Each dot represents one independent replicate. The total number of flies (N) was counted per genotype. Unpaired t test, ***p<0.001, **p<0.01, ns: not significant, mean ± SEM. (C) Representative images of adult wing blades showing morphological phenotypes caused by wing-specific expression of sl cDNAs. Expression of sl^D1041G caused severe wing defects including notched margins (arrows). Expression of sl^H384R or sl^L630F exhibited partial penetrance. Expression of sl^D1184G caused lethality before eclosion. Scale bars, 0.5 mm. (D) Quantification of eye size with expression of PLCG1 cDNAs. Eye size were normalized to the size of the ey-GAL4>UAS-Empty control. Each dot represents the measurement of one adult eye. Unpaired t test, ****p<0.0001, *p<0.05, ns: not significant, mean ± SEM. (E) Representative images showing that wing-specific expression of PLCG1^D1165H caused similar but more severe morphological phenotypes compared to PLCG1^D1019G or PLCG1^D1165G. Expression levels of UAS-cDNA can be manipulated by raising the animals at different temperatures. Scale bars, 0.5 mm. (F) Representative images showing the phenotypes associated with ectopic expression of the control PLCG1 variants. Expression of PLCG1^H380A or PLCG1^S1021F did not induce obvious morphological phenotypes in adult wings or eyes compared to PLCG1^Reference, whereas no eclosed adults were collected when overexpressing PLCG1^D1165H at 29°C in these contexts. Scale bars, 0.5 mm for wing images, 100 μm for eye images.

Figure 6.. PLCG1 variants affect important residues.

(A) 3D structure of full-length rat Plcg1 (rat Plcg1 shares 97% amino acid identity with human PLCG1). The conserved protein domains are labeled with different colors. Two major intracellular interfaces are circled by dashed lines: 1-The hydrophobic ridge between the sPH domain and the catalytic core (X-box and Y-box); and 2-The interface between the cSH2 domain and the C2 domain. The four amino acids affected by the variants are shown as bolded black and indicated by yellow balls. (B) Enlarged views of the Asp1019 residue within the autoinhibition interface between sPH domain and the Y box. The potential interactions with nearby residues are indicated. (C) Enlarged view of the Asp1165 residue within the autoinhibition interface between the cSH2 domain and the C2 domain. The potential interactions with nearby residues are indicated. (D) Enlarged view of the His380 residue within the X-box catalytic domain, in proximity to the Ca²⁺ cofactor. (E) Enlarged view of the Leu597 and nearby residues in the nSH2 domain. Structural analysis was performed via UCSF Chimera (Pettersen et al., 2004).

Figure 6—figure supplement 1.. In silico analyses of PLCG1 variants.

(A) Molecular dynamics simulations for the p.(Asp1019Gly) and p.(Asp1165Gly) variants. These variants exhibit increased disorganization with a higher root mean square deviations (RMSD) compared to the reference PLCG, similar to the hyperactive p.(Asp1165His) variant. Color code: PLCG1 reference - red; p.(Asp1019Gly) - green; p.(Asp1165Gly) - cyan; p.(Asp1165His) - purple. (B) Prediction of the effects of PLCG1 variants on protein stability and folding using DDMut. DDMut is a platform designed to predict protein stability based on the Gibbs Free Energy change (ΔΔG), and larger ΔΔG indicates a more pronounced impact on protein stability. The protein structure of rat Plcg1 [Protein Data Bank (PDB) ID 6PBC] was used as a reference. The p.(Asp1019Gly), p.(Asp1165Gly), and p.(Asp1165His) variants are predicted to induce destabilizing, with ΔΔG^{Stability wt->mt} values of –0.91 kcal/mol, –1.54 kcal/mol, and –1.59 kcal/mol, respectively, suggesting impaired protein folding or altered conformational dynamics. These results are consistent with their proposed roles in disrupting autoinhibitory interactions (Hajicek et al., 2013; Siraliev-Perez et al., 2022). The ΔΔG^{Stability wt->mt} value for p.(His380Arg) and the enzymatically inactive p.(His380Arg) are predicted to be –0.22 kcal/mol and –0.13 kcal/mol, respectively. Given that the His380 residue’s influence on the phospholipase activity might not be through the autoinhibition mechanism, the impact of His380 residue variants on the protein folding and stability may not elucidate their effects on protein function. The ΔΔG^{Stability wt->mt} value for p.(Leu597Phe) is predicted to be –1.26 kcal/mol, which is not inconsistent with the potential impact of this variant on interaction between nSH2 domain and RTK activators. However, the p.(Ser1021Phe) variant has a predicted ΔΔG^{Stability wt->mt} value of 0.14 kcal/mol, indicating a slight stabilizing effect. These analyses indicate that the variants may act through various pathogenic mechanisms, but further investigations are needed.