Introduction

The inositol lipid-specific phospholipase C (PLC) isozymes are key signaling proteins that play critical roles in transducing signals from hormones, growth factors, neurotransmitters, and many extracellular stimuli ^1–3. The PLCs selectively hydrolyze phosphatidylinositol 4,5-bisphosphate (PIP2) into inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG) ^4,5. PIP2 functions as a membrane anchor for numerous proteins and affects membrane dynamics and ion transport ^6–8. The two products, IP3 and DAG, are important intracellular second messengers involved in Ca²⁺ signaling regulation and protein kinase C signaling activation, respectively ^9,10. Hence, PLC orchestrates diverse cellular processes and behaviors, including cell growth, differentiation, migration, and cell death ^11–14. There are at least thirteen PLC isozymes grouped in 6 classes (β, δ, ε, γ, η, ζ) in mammals with similar enzymatic function, but each PLC has its own spectrum of activators, expression pattern, and subcellular distribution ^15–17.

PLCG1 [MIM: 172420] encodes the PLCγ1 isozyme. PLCγ1 is directly activated by receptor tyrosine kinases (RTKs) as well as cytosolic receptors coupled to tyrosine kinases ¹⁸. Upon tyrosine phosphorylation, PLCγ1 undergos conformational changes that release its autoinhibition upon which it associates with the plasma membrane to bind and hydrolyze its substrates ^19–21. There is a second PLCγ isozyme, PLCγ2, encoded by PLCG2 [MIM: 600220]. Although these two isozymes have similar protein structure and activation mechanism, they are differentially expressed and regulated, and play non-redundant roles ^22,23. PLCG2 is mostly expressed in cells of the hematopoietic system and mainly functions in immune response, causing human diseases associated with immune disorders ^24–27. However, PLCG1 is ubiquitously expressed and is enriched in the central nervous system (CNS) ²⁸. Plcg1 is essential in mice, and a null allele causes embryonic lethality with developmental defects in the vascular, neuronal, and immune system ^29,30. PLCG1 has emerged as a possible driver for cell proliferation, and increased expression level of PLCG1 has been observed in breast cancer, colon cancer, and squamous cell carcinoma ^31–34. Moreover, hyperactive somatic mutations of PLCG1 have been observed in angiosarcomas and T cell leukemia/lymphomas ^35–37. However, the genotype-phenotype association of germline PLCG1 variants has yet to be explored.

Results

Individuals with de novo heterozygous missense variants in PLCG1 exhibit hearing impairment, ocular pathology, and cardiac defects

Here, we report three unrelated individuals with de novo heterozygous missense variants in PLCG1 (GenBank: NM_002660.3). Proband 1 (c.3056A>G, p.Asp1019Gly) and proband 2 (c.1139A>G, p.His380Arg) were identified through the Undiagnosed Diseases Network (UDN) ⁴⁵ and proband 3 (c.3494A>G, p.Asp1165Gly) was identified via GeneMatcher ⁴⁶. The probands range in age from 5 years to 20 years. The phenotypes of the probands partially overlap but show a diverse spectrum. Pertinent shared features include deafness (probands 1 and 3), ophthalmologic abnormalities (probands 1 and 2), and cardiac septal defects (probands 1 and 3). Two of the three individuals show abnormal brain MRI (probands 1 and 2), and one has immune defects (proband 3). Additional variants identified in the probands are discussed in supplemental data. [As per medRxiv policy, the whole and detailed case history for the probands have been removed. To obtain more detailed information, please contact the authors]

The proband-associated missense PLCG1 variants are located in conserved protein domains and are predicted to be deleterious

PLCG1 is predicted to be tolerant to loss-of-function alleles with a pLI score ⁴⁷ of 0.16, suggesting that loss of one copy of the gene is unlikely to cause haploinsufficiency in humans and consistent with the presence of many protein truncating variants in gnomAD ⁴⁸. However, the missense constraint Z score ⁴⁷ of PLCG1 is 3.69, suggesting it is intolerant to missense variants. In addition, the prediction based on the DOMINO algorithm indicates that PLCG1 variants are likely to be dominant ⁴⁹. The in-silico pathogenicity predictions suggest that these variants are likely to be pathogenic (Table S1) based on MARRVEL ⁵⁰.

The three variants identified from the probands map to different conserved domains of PLCγ1, and each variant affects an amino acid residue that is conserved from flies to humans (Figure 1A and Figure 1B). The p.Asp1019Gly and p.His380Arg variants map to the catalytic core domains, and the p.Asp1165Gly is in the C-terminal C2 domain. PLCγ1 contains other conserved domains including an N-terminal pleckstrin homology (PH) domain, four EF hand motifs, as well as a PLCγ-specific regulatory array that is composed of a split PH domain (sPH), two Src homology 2 (nSH2 and cSH2) domains and a Src homology 3 (SH3) domain.

Figure 1.

small wing (sl) is the fly ortholog of human PLCG1

To obtain information about the nature of the proband-associated variants, we utilize Drosophila to model the variants in vivo. Flies have three genes encoding PLC isozymes (Table S2). Among them, small wing (sl) is predicted to be the ortholog of PLCG1 with a DIOPT (DRSC Integrative Ortholog Prediction Tool) score of 17/18 (DIOPT version 9.0) ⁵¹. The encoded proteins share 39% identity and 57% similarity and are composed of similar conserved domains (Figure 1A). sl is also predicted to be the ortholog of PLCG2 with a DIOPT score of 12/18. These data suggest that sl corresponds to two human genes encoding the PLCγ isozymes. We generated transgenic flies carrying the UAS-human PLCG1 cDNAs for both the reference (UAS-PLCG1^Reference) and the variants (UAS-PLCG1^D1019G, UAS-PLCG1^H380R, and UAS-PLCG1^D1165G). Given the high level of protein sequence homology and the conservation of the affected amino acids (Figure 1B), we also generated transgenic flies for the reference and analogous variants in the fly sl cDNA, namely UAS-sl^WTand UAS-sl^variants (UAS-sl^D1041G, UAS-sl^H384R, and UAS-sl^D1184G). sl is on the X chromosome, and several alleles of sl have been isolated or generated previously, including sl², sl^KO and sl^T2A (Figure 1C). sl² carries a 13bp deletion in the third exon that leads to a frameshift and early stop gain ⁵². sl² is a strong loss-of-function allele that causes small wing size, ectopic wing veins and extra R7 photoreceptors ⁵². sl^KOwas generated by CRISPR-mediated genomic editing that removes nearly the entire gene ⁵³. sl^T2A allele was generated by inserting an FRT-SA-T2A-GAL4-polyA-FRT cassette as an artificial exon into the first coding intron of sl (Figure 1C) ^54,55. The polyA arrests transcription, and sl^T2Ais a strong loss-of-function allele (Figure S1A). The T2A viral sequence triggers ribosomal skipping and leads to the production of GAL4 proteins ^56,57 that are expressed in the proper spatial and temporal pattern of sl. This allows us to assess the expression pattern of sl by driving the expression of a UAS-fluorescent protein ⁵⁴, or to assess the function of variants by expressing the human UAS-reference/variant cDNAs ^42,58–62. In addition, the cassette is flanked by two FRT sites so that it can be excised from the cells that express the gene in the presence of UAS-Flippase and revert the mutant phenotypes (Figure 1C) ⁵⁴.

We first assessed the expression pattern of sl by driving UAS-mCherry.nls (an mCherry that localizes to nuclei) with sl^T2A. sl is expressed in the 3^rd larval wing discs and eye discs (Figure 2A), consistent with the loss-of-function phenotypes observed in the wings and eyes ⁵². The expression pattern of sl in the wing discs is not homogenous. Higher expression levels are observed in the anterior compartment and along both the anterior/posterior and dorsal/ventral compartment boundaries (Figure 2A). The hemizygous sl^T2A/Y male flies and the trans-heterozygous sl^T2A/sl² or sl^T2A/sl^KO female flies show reduced wing size and ectopic wing veins (Figure 2B and Figure S1B), as well as additional photoreceptors in the eye (Figure 2C and Figure S1C). These phenotypes can be rescued by UAS-Flippase or by introducing a genomic rescue construct (Dp(1;3)DC313 ⁶³, Figure 1C) that covers the sl locus (Figure 2B and Figure 2C). These data show that all the observed phenotypes in sl^T2A mutants can be attributed to the loss of sl.

Figure 2.

sl is expressed in the fly CNS and loss of sl causes longevity and locomotion defects

Given that human PLCG1 is highly expressed in the central nervous system (CNS) ²⁸ and that the probands present with neurologic phenotypes including hearing or vision deficits (Table 1), we investigated the expression pattern and the cell type specificity of sl in the CNS of flies. sl is expressed in the larval CNS as well as the adult brain, and co-staining with the pan-neuronal marker Elav ⁶⁴ and glial marker Repo ⁶⁵ show that sl is expressed in a subset of neurons and glia cells in the CNS (Figure 3A). We therefore assessed the longevity and climbing of sl^T2A flies. Compared to the wild-type w¹¹¹⁸ flies, sl^T2A/Y hemizygous mutant flies show a shortened lifespan and a progressively reduced climbing ability. These phenotypes can be rescued by expression of the wild-type sl cDNA(sl^T2A/Y; UAS-sl^WT) (Figure 3B).

Figure 3.

Functional assays in flies indicate that the PLCG1 variants are toxic

To assess the impact of the variants, we expressed the sl variant cDNAs in the sl^T2A/Y hemizygous mutant males(sl^T2A/Y; UAS-sl^variants) and compared their rescue ability with the wild-type sl(sl^T2A/Y; UAS-sl^WT). As shown in Figure 4A, the sl^T2A/Y mutant flies (or the ones expressing a UAS-Empty control construct) have a slightly reduced eclosion rate but expression of the sl^WT cDNA fully rescues the percentage of eclosing progeny as measured by the Mendelian ratio. In contrast, expressing sl^H384R causes a severe reduction of the number of eclosing flies with very few escapers (sl^T2A/Y; UAS-sl^H384R). Moreover, expression of the sl^D1041G or sl^D1184G leads to 100% lethality (Figure 4A). These data clearly indicate that the three variants are toxic.

Figure 4.

Given that the expression of these sl cDNAs is performed in a null mutant background,we opted to assess the phenotypes associated with ectopic expression assays by overexpressing the wild-type or variant sl cDNAs using a strong ubiquitous driver, Tub-GAL4. Overexpression of sl^WT shows no impact on viability. However, overexpression of sl^H384R leads to ∼36% lethality, while overexpression of sl^D1041G or sl^D1184G results in 100% lethality at the L1-L2 larval stage (Figure 4B). Hence, ubiquitous overexpression assays cause similar but less severe lethal phenotypes than the expression assays driven by sl^T2A. This difference may arise from the different expression patterns and levels of Tub-GAL4 and sl^T2A. Importantly, sl^T2A drives expression in the cells that normally express sl. In summary, the data argue that the sl variants are likely to be gain-of-function or neomorphic alleles and that sl^D1041G and sl^D1184G are very strong toxic alleles whereas sl^H384Ris a milder allele.

To compare the sl and PLCG1 associated phenotypes, we conducted similar assays using human PLCG1 cDNAs. Expression of PLCG1^Referencein the sl^T2A/Y mutant flies (sl^T2A/Y; UAS-PLCG1^Reference) (Figure 4A) reduces viability by 90%, while the ubiquitous expression of PLCG1^Referenceunder the control of Tub-GAL4 (Tub-GAL4 > UAS-PLCG1^Reference) reduces viability by 63% (Figure 4B). These data argue that the reference human PLCG1 cDNA is toxic in flies. Expression of PLCG1^H380R in the sl^T2A/Y mutant flies (sl^T2A/Y; UAS-PLCG1^H380R) leads to a modest increase in lethality when compared to PLCG1^Reference(Figure 4A), whereas Tub-GAL4 > UAS-PLCG1^H380R is not worse than PLCG1^Reference (Figure 4B). In contrast, expression of PLCG1^D1019Gor PLCG1^D1165G results in 100% lethality using either of the drivers (Figure 4A and 4B). In summary, expressing the human PLCG1 in flies induces toxicity, and PLCG1^H380Rhas only slightly increased toxicity when compared to PLCG1^Reference. However, the PLCG1^D1019G or PLCG1^D1165G variants display a severe gain of toxicity.

Since overexpressing fly sl^WT does not cause viability issues (Figure 4B), the observed toxicity associated with PLCG1^Referencein flies may be due to the elevated expression of the human proteins. This hypothesis can be assessed by ectopic expression assays performed by raising the flies at different temperatures. The GAL4-UAS system is highly temperature-dependent since the promoter in the UAS construct contains an Hsp-70 promoter ⁶⁶: at 29°C the expression level is much higher than at 25°C, and at 22°C the expression level is significantly lower than at 25°C ⁶⁷. As shown in Figure 4C, the Tub-GAL4>UAS-PLCG1^Reference flies exhibit a high lethality ratio when raised at 29°C (∼63% lethal). This lethality ratio decreased to ∼43% when the flies are raised at 25°C whereas the flies are viable at 22°C. This shows that the toxicity is highly dependent on protein levels. Indeed, the expression levels can be further lowered by using a weak ubiquitous GAL4 driver, da-GAL4. The da-GAL4 > UAS-PLCG1^Reference or UAS-PLCG1^H380R flies are viable at the three tested temperatures (29°C, 25°C, and 22°C). In contrast, da-GAL4 > UAS-PLCG1^D1019Gor UAS-PLCG1^D1165G flies are lethal at 29°C, semi-lethal (∼80%-85% lethal) at 25°C, and viable at 22°C (Figure 4C, right panel). The animals that escape lethality at 25°C have smaller pupae (Figure S2) and a reduced adult body size, showing that growth is impeded. In summary, these data consistently confirmed that ubiquitous elevated expression levels of the human PLCG1 is toxic in flies.

To assess other phenotypes, we used nub-GAL4 to mostly drive the expression of UAS-PLCG1 or UAS-sl cDNAs in the wing disc, a well-established model to study growth and differentiation in a tissue that is dispensable for viability ⁶⁸. Overexpression of either PLCG1^Reference or sl^WT in the wing disc leads to slightly smaller wings with a ∼10% reduction in size when compared to expressing the UAS-Empty control (Figure S3). Hence, both loss and overexpression of sl in the wing lead to a size reduction (Figure 2B and S3). This implies a potential dosage-dependent regulation on wing growth by the PLCγ isozymes, although the underlying mechanism is unknown. Overexpression of the PLCG1^H380R or sl^H384R in the wing results in ∼5% reduction in wing size when compared to the PLCG1^Referenceor sl^WT (Figure S3). However, overexpression of PLCG1^D1019Gor PLCG1^D1165G results in severe wing phenotypes characterized by notched wing margins, fused/thickened veins and reduced wing sizes (Figure 5A). Notably, overexpressing fly sl^D1041Gleads to very similar morphological defects as the corresponding human PLCG1^D1019Gvariant, indicating that the observed wing phenotypes are indeed due to alterations in PLCG1/sl functions. These data also argue that the human PLCG1 functions in the same pathways as sl. Furthermore, overexpressing the sl^D1184G leads to pupal lethality (Figure 5A). This suggests that this variant has a more severe impact on development than the other variants and is not inconsistent with the observation that nub-GAL4 drives some expression of UAS-cDNA in the nervous system as well ⁶⁹.

Figure 5.

We also assessed the effect of ectopic expression of human PLCG1 on eye development using an eye-specific driver eyeless-GAL4 (ey-GAL4). The expression of PLCG1^Reference or PLCG1^H380R in fly eyes leads to a mild reduction in eye size, while the expression of PLCG1^D1019G or PLCG1^D1165Gleads to a severe reduction in eye size (Figure S4). In summary, the eye data are consistent with the wing data, showing that PLCG1^D1019G and PLCG1^D1165G are more toxic than PLCG1^Reference. On the other hand, the toxicity of PLCG1^H380R is milder and can be distinguished from that of the PLCG1^Reference only in certain contexts.

The p.Asp1019Gly and p.Asp1165Gly variants are gain-of-function variants

Previous studies have identified a very strong gain-of-function somatic PLCG1 variant p.Asp1165His in adult T cell leukemia/lymphoma ^20,70,71. This variant has been documented to cause a very dramatic increase in phospholipase activity in vitro ^20,71. To characterize the impact of this hyperactive variant in vivo, we generated transgenic flies with this variant and tested PLCG1^D1165Hin our tissue-specific expression assays. Overexpression of PLCG1^D1165Hin the eye using ey-GAL4 causes lethality at 29°C (Figure S4), arguing that it is highly toxic. As shown in Figure 5B, overexpressing PLCG1^D1165H in the wing using the nub-GAL4 driver causes lethality at 29°C. However, these flies survive when they are raised at 25°C, yet the wings show severe morphological defects, including notched wing margins, thickened veins as well as reduced wing size. These phenotypes are similar to, but more severe than, the defects observed in the wings overexpressing PLCG1^D1019Gor PLCG1^D1165G (Figure 5B). These data provide compelling evidence that these two variants are gain-of-function variants.

Discussion

A recent study reported an individual with a de novo heterozygous gain-of-function germline variant in PLCG1, p.Ser1021Phe ⁷². The proband exhibited an early-onset and severe immune dysregulation with autoimmune and autoinflammatory symptoms. Tao et al. performed in vitro and ex vivo experiments using cultured cell lines transfected with PLCG1 constructs and peripheral blood mononuclear cells from the proband, respectively. They showed that the p.Ser1021Phe variant led to a 1.5-2 fold increase in intracellular IP3 production compared to controls. However, in our in vivo assays, overexpression of PLCG1^S1021Fin wings or eyes does not cause obvious phenotypes when compared to PLCG1^Reference (Figure S5), which is clearly distinct from the other gain-of-function variants tested in our assays (p.Asp1019Gly, p.Asp1165Gly and p.Asp1165His) (Figure 5B). It is notable that the three probands reported here are discordant for immune-related phenotypes. Proband 1 with the p.Asp1019Gly variant has no immune dysregulation symptoms. Proband 2 with the p.His380Arg variant has a relapsing steroid responsive inflammatory encephalomyelitis, which is very different from the autoimmune symptoms reported by Tao et al ⁷². Finally, proband 3 with the p.Asp1165Gly variant presents with a T cell lymphocytopenia with recurrent infections suggesting that the individual is immune compromised. Hence, the immune phenotypes in the individuals reported here are very heterogeneous or absent, and inconsistent with an autoimmune disease. In summary, the immune-related phenotypes and their association with PLCG1 variants will need to be explored in more depth.

Previously, studies based on biochemical assays and protein structures provided insights into how the variants studied here may affect the enzymatic activity of PLCγ1 (the established protein structure of full-length rat Plcg1 is shown in Figure 6A). In its basal state, the PLCγ-specific regulatory array (sPH-nSH2-cSH2-SH3) forms autoinhibitory interfaces with the catalytic domains. Upon activation by the RTKs through binding to nSH2, PLCγ1 is phosphorylated, which in turn induces the dissociation of the inhibitory cSH2 domain from the C2 domain. This triggers conformational rearrangements, allowing the enzyme to associate with the membrane and exposing the catalytic domains to allow hydrolysis of PIP2 ^{19–21,73,74}. As shown in Figure 6A, the location of the three variants modeled in this study (Asp1019Gly, His380Arg and Asp1165Gly) are either within the catalytic domains or at the intramolecular interfaces. The p.Asp1019Gly and p.Asp1165Gly variants impact crucial residues involved in autoinhibition. Specifically, the p.Asp1019Gly variant affects a residue located at the apex of the hydrophobic ridge within the Y box (Figure 6B), which is important for the interaction between the sPH domain and the Y box. This interaction is critical for the autoinhibition by blocking the membrane engagement of the catalytic core domain prior to enzymatic activation ^20,75. Notably, substitution of Asp1019 with Lys (D1019K) has been demonstrated to enhance basal phospholipase activity by approximately 15 fold in vitro ⁷⁶. Similarly, another hotspot somatic variant, p.Ser345Phe, located at the hydrophobic ridge within the X box, involved in the interaction between the sPH domain and X box, has also been verified to be hyperactive ^37,77. In contrast, the p.Ser1021Phe variant described by Tao et al. ⁷² lies outside the hydrophobic ridge of the interface between the sPH domain and Y box (Figure 6B). On the other hand, the p.Asp1165Gly variant affects a residue situated within a loop of the C2 domain (Figure 6C). The Asp1165 residue plays a key role in stabilizing the interaction between the cSH2 domain and the C2 domain to maintain the autoinhibited state ⁷⁸. As mentioned above, the somatic variant p.Asp1165His leads to significantly elevated phospholipase activity in vitro ^71,73, and results in severe phenotypes in vivo (Figure 5B). In contrast, the p.His380Arg variant impacts the His380 residue within the X box, situated near the bound Ca²⁺ cofactor in the catalytic core (Figure 6D). His380 plays a role in coordination of phosphate 1 of IP3 ⁷⁴. While this residue may not be key to the autoinhibition, it is important to the phospholipase activity. Substitution of His380 with Phe or Ala (H380F, H380A) has been reported to suppress PIP2 hydrolysis and IP3 production ^79,80. Substitution of the His380 with Arg in p.His380Arg variant may create a more basic environment, impacting the lipase activity in a distinct way. Our in vivo data are consistent with the published in vitro data, strengthening our conclusion that the variants are pathogenic and impact the protein function.

Figure 6.

In summary, we report three individuals with de novo heterozygous missense variants in PLCG1 presenting with disease features that encompass ophthalmologic, hearing, and cardiac defects with variable expressivity. Our functional assays provide compelling in vivo evidence that the PLCG1 variants alter normal protein function. However, additional genetic variants (see Table 1) and environmental factors may contribute to some of the diverse phenotypes observed in these three individuals. Moreover, the PLCγ1 isozyme is an integral component of multiple signaling pathways, and the outcomes of its dysregulation are expected to be context dependent. Hence, different variants of PLCG1 may impact diverse cellular processes in different tissues and cells, leading to a range of pathological changes. Even when the affected residues are in close proximity to each other, their impact on protein function can be different ⁷³. Moreover, given that the p.Asp1019Gly and p.Asp1165Gly variants are hyperactive, potential therapeutic targets include specific inhibitors of PLCG1, such as antisense approaches that target the altered nucleotides ^81,82.

Data and code availability

This study did not generate datasets. All reagents developed in this study are available upon request.

Conflict of Interest Statement

The Department of Molecular and Human Genetics at Baylor College of Medicine receives revenue from clinical genetic testing completed at Baylor Genetics Laboratories.

Data Availability

All data produced in the present study are available upon reasonable request to the authors.

Acknowledgements

We thank the probands and families for their participation in this study. We thank Ms. Hongling Pan for transgenic fly lines. We thank Dr. Meisheng Ma for suggestions about protein structure interpretation. We thank the Bloomington Drosophila Stock Center (BDSC) for providing stocks.

Funding Statement

This work was supported by the Huffington Foundation; the Jan and Dan Duncan Neurological Research Institute at Texas Children’s Hospital, and the Undiagnosed Diseases Network funded by grants from the National Institutes of Health (U01 HG010233, U01 NS134355, U01 HG007709, U01 HG007942). Sequence data analysis was supported by the University of Washington Center for Rare Disease Research (UW-CRDR; U01 HG011744, UM1 HG006493, U24 HG011746). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. H.J.B. receives support from the NIH Common Fund through the Office of Strategic Coordination/Office of the NIH Director and the NINDS (U54 NS093793) as well as ORIP (R24 OD022005 and R24 OD031447). Confocal microscopy was performed in the BCM IDDRC Neurovisualization Core, supported by the NICHD (U54 HD083092).

Author contributions

Conceptualization: M.M., Y.Z., S.R.L., I.A.G., P.P., H.J.B.; Data curation: M.M., Y.Z., S.R.L., I.A.G., E.B., P.P., H.J.B.; Formal analysis: M.M., Y.Z., S.L., J.A.R., A.A.; Funding acquisition: UDN, H.J.B.; Investigation: M.M., Y.Z., S.L.; Resources: M.M., Y.Z., W.K.C., B.L.C., B.L.S., M.G., M.B., S.Y., M.F.W., S.R.L., I.A.G., P.P., H.J.B.; Supervision: H.J.B.; Visualization: M.M., Y.Z. S.L., X.P.; Writing-original draft: M.M., Y.Z.; Writing-review & editing: M.M., Y.Z., X.P., S.L., L.D., D.D., J.A.R., S.S., W-L.C., E.B., S.R.L., I.A.G., P.P., H.J.B..

The Undiagnosed Diseases Network Consortia

Carlos A. Bacino, Ashok Balasubramanyam, Lindsay C. Burrage, Hsiao-Tuan Chao, Ivan Chinn, Gary D. Clark, William J. Craigen, Hongzheng Dai, Lisa T. Emrick, Shamika Ketkar, Seema R. Lalani, Brendan H. Lee, Richard A. Lewis, Ronit Marom, James P. Orengo, Jennifer E. Posey, Lorraine Potocki, Jill A. Rosenfeld, Elaine Seto, Daryl A. Scott, Arjun Tarakad, Alyssa A. Tran, Tiphanie P. Vogel, Monika Weisz Hubshman, Kim Worley, Hugo J. Bellen, Michael F. Wangler, Shinya Yamamoto, Oguz Kanca, Christine M. Eng, Pengfei Liu, Patricia A. Ward, Edward Behrens, Marni Falk, Kelly Hassey, Kosuke Izumi, Gonench Kilich, Kathleen Sullivan, Adeline Vanderver, Zhe Zhang, Anna Raper, Vaidehi Jobanputra, Mohamad Mikati, Allyn McConkie-Rosell, Kelly Schoch, Vandana Shashi, Rebecca C. Spillmann, Queenie K.-G. Tan, Nicole M. Walley, Alan H. Beggs, Gerard T. Berry, Lauren C. Briere, Laurel A. Cobban, Matthew Coggins, Elizabeth L. Fieg, Frances High, Ingrid A. Holm, Susan Korrick, Joseph Loscalzo, Richard L. Maas, Calum A. MacRae, J. Carl Pallais, Deepak A. Rao, Lance H. Rodan, Edwin K. Silverman, Joan M. Stoler, David A. Sweetser, Melissa Walker, Jessica Douglas, Emily Glanton, Shilpa N. Kobren, Isaac S. Kohane, Kimberly LeBlanc, Audrey Stephannie C. Maghiro, Rachel Mahoney, Alexa T. McCray, Amelia L. M. Tan, Surendra Dasari, Brendan C. Lanpher, Ian R. Lanza, Eva Morava, Devin Oglesbee, Guney Bademci, Deborah Barbouth, Stephanie Bivona, Nicholas Borja, Joanna M. Gonzalez, Kumarie Latchman, LéShon Peart, Adriana Rebelo, Carson A. Smith, Mustafa Tekin, Willa Thorson, Stephan Zuchner, Herman Taylor, Heather A. Colley, Jyoti G. Dayal, Argenia L. Doss, David J. Eckstein, Sarah Hutchison, Donna M. Krasnewich, Laura A. Mamounas, Teri A. Manolio, Tiina K. Urv, Maria T. Acosta, Precilla D’Souza, Andrea Gropman, Ellen F. Macnamara, Valerie V. Maduro, John J. Mulvihill, Donna Novacic, Barbara N. Pusey Swerdzewski, Camilo Toro, Colleen E. Wahl, David R. Adams, Ben Afzali, Elizabeth A. Burke, Joie Davis, Margaret Delgado, Jiayu Fu, William A. Gahl, Neil Hanchard, Yan Huang, Wendy Introne, Orpa Jean-Marie, May Christine V. Malicdan, Marie Morimoto, Leoyklang Petcharet, Francis Rossignol, Marla Sabaii, Ben Solomon, Cynthia J. Tifft, Lynne A. Wolfe, Heidi Wood, Aimee Allworth, Michael Bamshad, Anita Beck, Jimmy Bennett, Elizabeth Blue, Peter Byers, Sirisak Chanprasert, Michael Cunningham, Katrina Dipple, Daniel Doherty, Dawn Earl, Ian Glass, Anne Hing, Fuki M. Hisama, Martha Horike-Pyne, Gail P. Jarvik, Jeffrey Jarvik, Suman Jayadev, Emerald Kaitryn, Christina Lam, Danny Miller, Ghayda Mirzaa, Wendy Raskind, Elizabeth Rosenthal, Emily Shelkowitz, Sam Sheppeard, Andrew Stergachis, Virginia Sybert, Mark Wener, Tara Wenger, Raquel L. Alvarez, Gill Bejerano, Jonathan A. Bernstein, Devon Bonner, Terra R. Coakley, Paul G. Fisher, Page C. Goddard, Meghan C. Halley, Jason Hom, Jennefer N. Kohler, Elijah Kravets, Beth A. Martin, Shruti Marwaha, Chloe M. Reuter, Maura Ruzhnikov, Jacinda B. Sampson, Kevin S. Smith, Shirley Sutton, Holly K. Tabor, Rachel A. Ungar, Matthew T. Wheeler, Euan A. Ashley, William E. Byrd, Andrew B. Crouse, Matthew Might, Mariko Nakano-Okuno, Jordan Whitlock, Manish J. Butte, Rosario Corona, Esteban C. Dell’Angelica, Naghmeh Dorrani, Emilie D. Douine, Brent L. Fogel, Alden Huang, Deborah Krakow, Sandra K. Loo, Martin G. Martin, Julian A. Martínez-Agosto, Elisabeth McGee, Stanley F. Nelson, Shirley Nieves-Rodriguez, Jeanette C. Papp, Neil H. Parker, Genecee Renteria, Janet S. Sinsheimer, Jijun Wan, Justin Alvey, Ashley Andrews, Jim Bale, John Bohnsack, Lorenzo Botto, John Carey, Nicola Longo, Paolo Moretti, Laura Pace, Aaron Quinlan, Matt Velinder, Dave Viskochil, Gabor Marth, Pinar Bayrak-Toydemir, Rong Mao, Monte Westerfield, Anna Bican, Thomas Cassini, Brian Corner, Rizwan Hamid, Serena Neumann, John A. Phillips III, Lynette Rives, Amy K. Robertson, Kimberly Ezell, Joy D. Cogan, Nichole Hayes, Dana Kiley, Kathy Sisco, Jennifer Wambach, Daniel Wegner, Dustin Baldridge, F. Sessions Cole, Stephen Pak, Timothy Schedl, Jimann Shin, and Lilianna Solnica-Krezel.

Supplemental Data

Additional variants identified in the probands

Proband 1 carries an intragenic duplication in PSD3. PSD3 has not been associated with a Mendelian disorder but is potentially associated with an autosomal dominant arthrogryposis ¹. Hence, it may underlie the joint defects observed in proband 1.

Proband 2 has compound heterozygous missense variants in ERAP2 and SEMA3G. ERAP2 [MIM: 609497] has not been associated with a Mendelian disorder. It encodes an ER-residential metalloaminopeptidase that functions in the major histocompatibility class I antigen presentation pathway. Some variants in ERAP2 are associated with a susceptibility to autoimmune diseases such as ankylosing spondylitis and Crohn’s disease ^2–4. Given that proband 2 exhibits neuroinflammation and encephalitis, these phenotypes may be associated with the ERAP2 variants. SEMA3G (Semaphorin 3G) has not been associated with a Mendelian disorder. However, a homozygous missense variant in SEMAG3 was observed in two affected siblings from a consanguineous family. The siblings exhibited dysmorphic features as well as developmental delay ⁵.

Proband 3 carries a de novo missense variant in PKP2 [MIM: 602861]. PKP2 encodes Pakophilin-2 and has been associated with dominant arrhythmogenic right ventricular dysplasia 9 [MIM: 609040] ^6–8. However, this proband was born with septal defects.

Expression of human PLCG1 reduces the viability of sl^T2A hemizygotes, and PLCG1^Reference shows no obvious rescue of the phenotypes in the wings and eyes caused by loss of sl

We assessed if human PLCG1 could effectively serve as a functional substitute for fly sl and rescue the loss-of-function phenotypes observed in sl mutant flies. However, only a small fraction of the sl^T2A/Y mutant hemizygotes expressing PLCG1^Reference can survive to adults (Figure 4A). In addition to the obvious toxicity, PLCG1^Referencefails to rescue the phenotypes observed in the wings or eyes of the sl mutant flies (Figure S6), which is fully rescued by fly sl^WT (Figure 2). This suggests that despite their high DIOPT score, human PLCG1 cannot fully substitute for fly sl. It is possible that during the course of evolution, PLCG1 has acquired more specialized functions. For example, an essential step for enzymatic activation of mammalian PLCγ is the binding of its nSH2 domain to specific phosphotyrosines on the RTKs through a specific binding motif ⁹. However, Thackeray et al. found that the consensus motif is absent in the intracellular domain of the Drosophila EGF receptor homolog DER ¹⁰, one of the three RTKs in Drosophila which is required for wing vein differentiation and photoreceptor formation ^11–13.

Supplemental Figures and Legends

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.

Supplemental Tables

Table S1. Pathogenicity prediction of the proband variants

Table S2. Mammalian PLC coding genes and their fly orthologs

Table S3. Fly strains used in the experiments

Table S4. Primers used in the experiments

Material and methods

Recruitment of the probands

Formal consents for genetic testing and publication were obtained from all probands or their family members. Probands 1 and 2 were recruited through the Undiagnosed Diseases Network (UDN) and were evaluated through the clinical research protocol of the National Institutes of Health Undiagnosed Diseases (15-HG-0130), which was approved by the National Human Genome Research Institute (NHGRI). Proband 3 was recruited through GeneMatcher.

Drosophila husbandry and generation of transgenic flies

The fly strains used in this study (listed in Table S3) were generated in house or obtained from the Bloomington Drosophila Stock Center (BDSC). All the flies were raised and maintained on standard fly food at room temperature unless specified. The sl^T2A allele was outcrossed with w¹¹¹⁸ to clean up the genetic background. The strains used in this study were listed in Supplemental Table S3.

To generate the UAS-cDNA transgenic lines, human PLCG1 cDNA was obtained from Horizon Discovery (MHS6278-213246131, clone ID 9052656), and fly sl cDNA was obtained from Drosophila Genomics Resource Center (DGRC, RE62235). The coding sequence (CDS) of PLCG1^Referenceand sl^WT were amplified using iProof™ High-Fidelity DNA Polymerase Kit (BioRad, #1725301), purified using QIAEX II Gel Extraction Kit (QIAGEN, #20021), sub-cloned into the Gateway compatible entry vector pDONR223 by BP cloning (BP clonase II, Thermo Fisher Scientific, #11789020) and sequentially cloned into the destination vector pGW-attB-HA by LR cloning (LR clonase II, Thermo Fisher Scientific, #11791100) ¹⁴. The variants were generated by site-directed mutagenesis strategy using Q5 Hot Start High-Fidelity 2x Master Mix (NEB, #M0494S) and DpnI restriction enzyme (NEB, # R0176L). All the constructs were sanger verified and injected, and inserted into the VK33 (PBac{y[+]-attP}VK00033) docking site using ϕC31 mediated transgenesis ^15,16. Primers are listed in Supplemental Table S4.

Drosophila behavioral assays

The climbing assay to examine the negative geotaxis and locomotion ability of the flies was performed as previously described ^17,18 with some modifications. 18-22 flies per vial were transferred to an empty plastic vial and given 20min to rest prior to being tested. The flies were tapped to the bottom of the vial and were allowed to climb for 15s. The percentage of flies per vial that climbed over 5cm were calculated. The maximum distance from the bottom to the top is 18.5cm.

For the lifespan assay, newly eclosed male flies were collected and maintained at 25 °C (10 flies per vial). The flies were transferred to a new vial and the number of dead flies was counted every two days.

Immunostaining

Fly tissues were dissected in 1x PBS, fixed in 4% paraformaldehyde for 20min at room temperature, and washed in PBS (3 x 10min). For antibody staining, samples were treated with PBST (Triton X-100 in PBS, 0.1% for larval tissues, 2% for adult brain), 5% normal goat serum, and incubated in primary antibody overnight at 4°C. Samples were washed with 0.1% PBST (3 x 10min) and incubated with secondary antibody for 2h at room temperature (in darkness) and washed in 0.1% PBST (3 x 10min). Primary antibodies: rat anti-Drosophila Elav (1:250, DSHB, #7E8A10); mouse anti-Drosophila Repo (1:50, DSHB, #8D12). Secondary antibodies: goat anti-rat-647 (1:250, Jackson ImmunoResearch, #112-605-003), goat anti-mouse-Cy5 (1:250, Invitrogen, #A10524). Larval discs were mounted in Vectashield (Vector Labs #H1200 and #H1000). Larval CNS and adult brain were mounted in Rapiclear (Cedarlane, #RC147001). For adult retinas, flies are reared at 25 °C under 12-h light/dark conditions. Retinas were isolated from 5-7 day old flies. Heads were dissected in PBS and fixed in 3.7% formaldehyde overnight at 4°C. The samples were rinsed with 0.1% PBST and the retinas were subsequently dissected and incubated with PBST-diluted phalloidin 647 (1:100, Invitrogen, #A22287) for 1h. Retinas were washed in 0.1% PBST and mounted in Vectashield. The images were obtained with a confocal microscope (Leica SP8X or Zeiss Airyscan LSM 880) and processed using the ImageJ-FIJI software ¹⁹.

Imaging of adult fly wings and eyes

To prepare the samples of adult fly wings, the wing blades were dissected and mounted in a glycerol/ethanol 1/1 mixture. Only wings from the same gender were compared to each other since females have larger wings than males when raised in the same conditions. To prepare the samples of adult fly eyes, the flies were frozen and placed onto a double-side stick tape with one eye facing up. The samples were imaged using bright field Stereomicroscope (Leica MZ16 or Leica Z16 APO). Image Pro Plus 7.0 software was used to for extended depth-of-field images. The image processing and the measurement of the total areas of the wing blades or eyes were conducted using the ImageJ-FIJI software¹⁹.

Real-time PCR

Real-time PCR was performed as previously described ²⁰ with modifications. All-In-One 5X RT MasterMix (abm, #G592), iTaq Universal SYBR Green Master Mix (BioRad#1725120) and BioRad C1000 Touch Cycler were used. Primers are listed in Supplemental Table S4.