Cross-species analysis of LZTR1 loss-of-function mutants demonstrates dependency to RIT1 orthologs

  1. Antonio Cuevas-Navarro
  2. Laura Rodriguez-Muñoz
  3. Joaquim Grego-Bessa
  4. Alice Cheng
  5. Katherine A Rauen
  6. Anatoly Urisman
  7. Frank McCormick
  8. Gerardo Jimenez
  9. Pau Castel  Is a corresponding author
  1. Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, United States
  2. Institute for Molecular Biology of Barcelona, Consejo Superior de Investigaciones Científicas, Spain
  3. Centro Nacional de Investigaciones Cardiovasculares, Spain
  4. UC Davis MIND Institute, University of California Davis, United States
  5. Department of Pediatrics, University of California Davis, United States
  6. Department of Anatomic Pathology, University of California San Francisco, United States
  7. Institució Catalana de Recerca i Estudis Avançats (ICREA), Spain
  8. Department of Biochemistry and Molecular Pharmacology, New York University Grossman School of Medicine, United States

Abstract

RAS GTPases are highly conserved proteins involved in the regulation of mitogenic signaling. We have previously described a novel Cullin 3 RING E3 ubiquitin ligase complex formed by the substrate adaptor protein LZTR1 that binds, ubiquitinates, and promotes proteasomal degradation of the RAS GTPase RIT1. In addition, others have described that this complex is also responsible for the ubiquitination of classical RAS GTPases. Here, we have analyzed the phenotypes of Lztr1 loss-of-function mutants in both fruit flies and mice and have demonstrated a biochemical preference for their RIT1 orthologs. Moreover, we show that Lztr1 is haplosufficient in mice and that embryonic lethality of the homozygous null allele can be rescued by deletion of Rit1. Overall, our results indicate that, in model organisms, RIT1 orthologs are the preferred substrates of LZTR1.

Editor's evaluation

Using elegant cross-species biochemistry and genetic approaches, this paper describes the role of the ubiquitin adaptor protein LZTR1 in regulation of the RAS-related GTPase RIT1 as its principal substrate involved in the RASopathy, Noonan syndrome. Although this work does not fully rule out the involvement of canonical RAS isoforms in LZTR1-associated RASopathies in humans, the extensive genetic experiments in Drosophila and mouse presented here suggest that pathological phenotypes observed in LZTR1-linked RASopathy models are mediated primarily by its target RIT1, and not by canonical RAS isoforms.

https://doi.org/10.7554/eLife.76495.sa0

Introduction

Dysregulation of signal transduction by the RAS family of guanosine 5’-triphosphate (GTP) hydrolases (GTPases) can have profound effects on human development and cause genetic disorders collectively termed RASopathies (Castel et al., 2020; Rauen, 2013). Ras GTPases exhibit high affinity toward guanine nucleotides and act as molecular switches by mediating GTP hydrolysis. The nucleotide cycling of RAS GTPases is tightly regulated by GTPase activating proteins (GAPs; e.g. neurofibromin) and guanine nucleotide exchange factors (GEFs; e.g. SOS1) that facilitate nucleotide hydrolysis or loading, respectively (Simanshu et al., 2017). Upon GTP binding, RAS proteins undergo a conformational change, which promotes the interaction with different protein effectors that activate downstream signaling pathways, including Raf/MEK/ERK mitogen activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K) pathways. Although GAPs and GEFs can rapidly affect the nucleotide cycling of RAS proteins, and hence their activity, other accessory proteins can modulate downstream signaling by regulating the stability and/or activity of RAS GTPases.

Noonan syndrome (NS) is a common RASopathy that is characterized by craniofacial dysmorphism, short stature, congenital heart disease, and developmental delays. The cause of NS has been linked to genetic alterations that result in hyperactivation of the Raf/MEK/ERK MAPK pathway, including recently reported gain-of-function mutations in the RAS GTPase RIT1 and loss-of-function mutations in the BTB protein LZTR1 (Aoki et al., 2013; Yamamoto et al., 2015; Johnston et al., 2018). We have previously reported that RIT1 is mostly bound to GTP in cells, suggesting that it either lacks a GAP, uses a less active GAP, or relies on its intrinsic GTPase activity (Castel et al., 2019). An alternative regulatory mechanism of RIT1 activity is through protein degradation; we identified a Cullin-3 RING E3 ubiquitin ligase complex (CRL3) that uses LZTR1 as a substrate receptor (CRL3LZTR1) to bind RIT1 and promote its ubiquitination and proteasomal degradation. CRL3LZTR1 binds to GDP-bound RIT1 and maintains tight regulation of RIT1 protein levels. Importantly, NS-associated RIT1 and LZTR1 missense mutations disrupt CRL3LZTR1-RIT1 binding, thus relieving RIT1 from its negative regulatory mechanism (Castel, 2022).

LZTR1 has been reported to regulate the protein stability of other RAS GTPases by similar mechanisms (Bigenzahn et al., 2018; Steklov et al., 2018; Abe et al., 2020). In this study, we employed a molecular evolutionary approach to elucidate the functional relationship and co-evolution of LZTR1 and its cognate RAS GTPase substrate(s), including the evaluation of LZTR1 loss-of-function mutations in invertebrate and mammalian model organisms.

Results and discussion

The RAS family of GTPase proteins has been previously shown to be well-conserved across species and LZTR1 has been proposed to regulate a subset of these proteins (Rojas et al., 2012). Therefore, to understand how LZTR1 has evolved in comparison with other RAS GTPases, we performed protein alignment analysis of the human orthologs of KRAS, RIT1, and LZTR1 orthologs from human and common model organisms. Consistent with previous studies, KRAS was highly conserved in less complex organisms, including fission yeast (Ras1; 60% protein identity) (Figure 1a). In contrast, we found that RIT1 rapidly diverged in less complex organisms and a mildly conserved protein in Drosophila melanogaster (RIC; 60% protein identity) was the most distant ortholog that could be found in these laboratory model species. LZTR1 followed a similar pattern as RIT1; we could identify an ortholog in fruit flies (CG3711/LZTR1; 60% protein identity), but not in the roundworm Caenorhabditis elegans (Figure 1a). These analyses suggest that RIT1 and LZTR1 are more likely to have functionally co-evolved than KRAS and LZTR1. To further assess the biochemical relationship between these proteins across model organisms, we generated recombinant proteins for the KRAS, RIT1, and LZTR1 orthologs from mouse (Mus musculus), zebrafish (Danio rerio), and fruit fly (Drosophila melanogaster) and tested their ability to interact in pull down assays. All LZTR1 orthologs preferentially interacted with RIT1 orthologs (Figure 1b, Figure 1—figure supplement 1), indicating that this interaction that we previously reported for human proteins is also conserved in less complex model organisms.

Figure 1 with 1 supplement see all
Evolutionary analysis of LZTR1 and the RAS proteins KRAS and RIT1.

(a) Phylogenetic trees of KRAS, RIT1, and LZTR1 orthologs based on multiple protein sequence alignments performed with Clustal Omega (Figure 1—source data 1). Orthologs were searched in the following model organisms: chimpanzee (Pan troglodytes), pig (Sus scrofa), chicken (Gallus gallus), mouse (Mus musculus), African clawed frog (Xenopus laevis), zebrafish (Danio rerio), fruit fly (Drosophila melanogaster), roundworm (Caenorhabditis elegans), and budding yeast (Saccharomyces cerevisiae). (b) Pull-down assays using GST-tagged KRAS and RIT1 or their mouse, zebrafish, or fruit fly orthologs produced recombinantly in bacteria. Recombinant proteins were incubated with lysates from HEK-293T cells expressing their corresponding species’ HA-tagged LZTR1 ortholog. Representative results from three biological replicates.

To better understand the role of the LZTR1 ortholog in Drosophila melanogaster, we isolated Lztr1 loss-of-function mutations by using two independent CRISPR/Cas9 gene editing approaches. One resulting allele, Lztr11, causes a frameshift at the beginning of the second exon and encodes a short, truncated product. Another allele, Lztr12, lacks all coding sequences and is, therefore, a null mutation (Figure 2a). Lztr1 null flies were normal and fertile and did not display any obvious phenotype. A previous study showed that flies expressing RNA interference constructs against Lztr1 displayed minor defects in wing vein patterning (Bigenzahn et al., 2018); however, we did not find consistent vein defects in our null mutants (Figure 2b).

Drosophila Lztr1 regulates Ric stability.

(a) A schematic representation of the Drosophila CG3711/Lztr1 gene locus is shown. Coding exons are represented in blue. Two CRISPR/Cas9-mediated approaches were used to isolate the Lztr1 loss-of-function alleles, Lztr11 and Lztr12. (b) Wing vein patterning is not affected in Lztr1 null flies (upper panel), with very few individuals exhibiting small ectopic veinlets in two or three points (asterisks; lower panel) (n = 209 flies). (c) Estimated normalized protein abundance expressed as mean log2 fold change in Lztr12 vs yw (control) comparison. Corresponding single-protein standard error (SE) and t-test p-values are listed. (d) Immunoblot analysis of protein extracts isolated from the indicated transgenic adult flies in a yw or Lztr12 background. ns: non-specific band. (e) Same as panel (d) with protein extracts isolated from third-instar larvae. Representative results from three biological replicates.

Next, we undertook label-free proteomics to quantify the levels of Ras family proteins in head extracts from Lztr12 and background matched control (yw) flies. Ric levels were upregulated ~11 times in Lztr12 fly heads relative to the control, while Ras85D (the HRAS, KRAS, and NRAS ortholog, hereafter referred to as Ras) levels were only modestly altered in both backgrounds (~1.5-fold difference) (Figure 2c). Similarly, Ras64D (the MRAS ortholog) was only upregulated ~1.6-fold in Lztr12 mutant heads. To validate these results, we generated Drosophila transgenic lines carrying N-terminally HA tagged Ric or Ras genes under the control of their natural genomic sequences, thereby facilitating detection of both products at their endogenous levels. Both transgenes were then placed in the Lztr12 mutant background to assess the effect of Lztr1 depletion on Ric and Ras protein stability. Consistent with our mass spectrometry results, HA-Ric levels were elevated in the absence of Lztr1, while HA-Ras levels remained barely altered in either mutant adult flies or third-instar larvae (Figure 2d–e). Altogether, these experiments show that Drosophila Lztr1 preferentially regulates Ric rather than Ras levels, as seen with the corresponding mammalian orthologs.

Given the absence of a phenotype in fruit flies, we examined the effect of Lztr1 deletion in mice. Steklov et al had previously reported that Lztr1 heterozygous knockout mice display typical features of NS, suggestive of haploinsufficiency (Steklov et al., 2018). In humans, NS-associated variants are most commonly found as bi-allelic loss-of-function variants that are inherited in an autosomal recessive manner; although there are few heterozygous variants that segregate as autosomal dominant, these are single nucleotide variants that likely act as dominant negatives (Johnston et al., 2018; Yamamoto et al., 2015; Motta et al., 2019). Hence, we further analyzed the potential NS-like phenotype of heterozygous Lztr1 deletion in the mouse as previously described in other mouse models (Araki et al., 2004; Wu et al., 2011; Hernández-Porras et al., 2014; Castel et al., 2019). Heterozygous mice did not exhibit significant changes in body weight and, overall, their morphology appeared normal (Figure 3a, Figure 3—figure supplement 1). The morphology of the skull, which generally shows dysmorphic features in mouse models of NS (round skull, blunt snout, and hypertelorism), was assessed using micro-computed tomography (µCT) and did not have significant differences when compared to wild-type littermates (Figure 3b). In addition, we assessed whether heterozygous mice displayed either cardiomegaly or splenomegaly, as these signs have been observed in other NS-like mouse models, but we failed to observe differences with wild-type animals (Figure 3c–d). No significant changes in RAS GTPases levels were seen either in tissue extracts from these mice (Figure 3—figure supplement 1). These results indicate that Lztr1 heterozygous mice do not exhibit the typical features seen in other NS mouse models and indicate that Lztr1 is haplosufficient for this phenotype. In support of these observations in mice, NS families carrying LZTR1 alleles generally exhibit an autosomal recessive inheritance pattern and bi-allelic inactivation of LZTR1 is required to exhibit a NS phenotype (Johnston et al., 2018). In addition, analysis of LZTR1 variants in the genome aggregation database (gnomAD) shows that many loss-of-function variants are observed in healthy individuals (pLI = 0; o/e = 2.28) (Karczewski et al., 2020). This indicates that these heterozygous loss-of-function variants are tolerated and are found in non-syndromic individuals. Thus, these data suggest haploinsufficiency of LZTR1 in humans is unlikely to lead to NS.

Figure 3 with 1 supplement see all
Lztr1 is haplosufficient in mice and its null phenotype can be modified by strain background.

(a) Weight (left panel) and length (right panel) of 4-week-old male Lztr1 wild type (n = 18) and heterozygous mutant (n = 20) mice. (b) Representative µCT imaging of the skull of an 8-week-old male Lztr1 wild type and heterozygous mutant mouse. The indicated values show the average measurement (mm) of length, width, and inner intercanthal distance in Lztr1 wild type (n = 5) and heterozygous mutant (n = 5) mice. Mann-Whitney p values were not significant for all the measurements. (c) Heart weight was similar between 8-week-old male Lztr1 wild type (n = 8) and heterozygous mutant (n = 6) mice, as assessed by heart to body weight ratio (HW/BW). Mann-Whitney test p value was not significant. (d) Same as panel (c), for the spleen of these mice. (e) Pie charts indicate the percentage of obtained genotypes upon weaning (21 days of age) the offspring of Lztr1 heterozygous mutant intercrosses. Each pie chart represents a different strain background and/or mixed background. (f) Representative image of female littermates with the indicated Lztr1 genotypes (C57BL/6N-129Sv F3 background). Note the decreased size, round skull, and proptosis of the homozygous Lztr1 knockout mouse. (g) Immunoblot analysis of RIT1, RAS, and Tubulin proteins isolated from the indicated tissues of Lztr1 wild type, heterozygous, and homozygous mice (C57BL/6N-129Sv F3 background). Protein lysates from two different mice were used for each genotype.

We and others have previously shown that Lztr1 knockout mice are embryonically lethal (Steklov et al., 2018; Castel et al., 2019); however, the reason for such lethality remains largely unknown. A previous study showed that Lztr1-associated embryonic lethality can be rescued by administering a MEK1/2 inhibitor to pregnant females, indicating that the phenotype is largely dependent on MAPK hyperactivation (Steklov et al., 2018). In fact, in other murine NS models, excessive activation of MAPK during embryonic development results in lethality, as seen in the KrasV14I model (Hernández-Porras et al., 2014). Interestingly, this phenotype appears to be more prominent in certain mouse background strains, such as C57BL/6 N, and more permissive in 129Sv (Araki et al., 2004; Hernández-Porras et al., 2015). Since our Lztr1 mutant mice are in the C57BL/6 N strain, we hypothesized that mouse background affects embryonic lethality. Backcrossing our mice to 129Sv females yielded heterozygous Lztr1 mutant mice in a 50% mixed background. After one backcross (referred to as F1), using a heterozygous x heterozygous breeding scheme, we obtained 1/55 (~1.8%) homozygous viable mice. At F2, we obtained 3/45 (~6.7%) homozygous viable mice. At F3, we obtained 3/34 (~8.8%) homozygous viable mice (Figure 3e). In contrast, mice in the 129Sv pure background did not yield any homozygous viable mice, suggesting that a combination of strain-specific genetic modifiers is likely required to tolerate Lztr1 deletion. In Raf1 D486N NS mice, a potential gene modifier was mapped at chromosome 8 of the 129Sv strain (Wu et al., 2012); however, we hypothesize that in the context of Lztr1, a negative gene modifier is also likely to be located in the Y chromosome of the 129Sv strain, given that we do not yield any homozygous viable mice after Y-chromosome fixing.

The limited number of viable homozygous Lztr1 knockout mice prevented us from undertaking quantitative phenotyping; however, when compared to wild type littermates, Lztr1 knockout mice appeared smaller, displayed characteristic dysmorphic facial features (round snout, hypertelorism, low set of ears, and proptosis), and exhibited cardiomegaly, consistent with other mouse models of NS (Figure 3f). In addition, Lztr1 knockout mice exhibited increased levels of Rit1 in all the tissues that we analyzed, while RAS levels remained mostly unchanged (Figure 3g).

Next, we sought to investigate the cause of embryonic lethality in Lztr1 knockout mice. We and others had previously shown that many embryos survive to late developmental stages (i.e. embryonic day (E)17.5–19.5), therefore, we harvested embryos on day E18.5. Most embryos display extensive hemorrhages (Figure 4a) and, consistent with this observation, conditional deletion of Lztr1 in blood vessels has been previously shown to cause vascular leakage (Sewduth et al., 2020). Since many NS alleles that cause lethality in mice have been related to cardiovascular dysfunction, we analyzed the heart phenotype of E18.5 embryos. Valve leaflets and endocardial cushions were normal, in contrast to the previous Ptpn11 NS-associated alleles (Araki et al., 2004). In Lztr1 knockout embryos, the ventricular myocardial wall showed defects that are compatible with ventricular noncompaction cardiomyopathy, similar to that seen in the Ptpn11 Q79R transgenic mouse model (Nakamura et al., 2007). Detailed analysis of compact and trabecular myocardial thickness showed significant differences in mutant embryos (Figure 4b). In addition, we observed interventricular septal defects in ~20% mutant hearts. Although the heart phenotype is striking, it is likely that embryonic lethality results from a combination of vascular and cardiac defects during development.

Figure 4 with 1 supplement see all
Lethality in Lztr1 knockout mutants as a result of Rit1-dependent cardiovascular defects.

(a) Gross morphology of Lztr1+/+ (wild type) and Lztr1-/- (knockout) embryos at E13.5 and E18.5. Note the presence of extensive hemorrhage in knockout embryos (yellow arrows). (b) Histological characterization of E18.5 hearts shows highly penetrant defects in ventricular wall thickness at both outflow tract (OFT) and atrioventricular canal (AVC) levels, as well as septal defects (arrows) in ~20% Lztr1-/- embryos. Quantification of ventricular wall thickness in both CM (yellow) and TM (green) and CM/TM index are shown in the graphs (n = 5). Rv: right ventricle; lv: left ventricle; ivs: interventricular septum; CM: compact myocardium; TM: trabecular myocardium. p Values were calculated using Student’s t-test. (c) Immunoblot analysis of different tissues isolated from Rit1+/+ (wild type), Rit1+/- (heterozygous), and Rit1-/- (knockout) adult mice. (d) Percentage of Lztr1 genotypes obtained upon weaning (21 days of age) the offspring of Lztr1 heterozygous mutant intercrosses in either a Rit1+/+ (n = 19) or Rit1-/- (n = 85) background. All these mice were maintained in a C57BL/6 N background. (e) Histological analysis of the heart of Lztr1; Rit1 double knockout E18.5 embryos (n = 3). (f) Immunoblot analysis of lysates isolated from primary MEF with the indicated genotypes. (g) Primary MEF derived from wild type, Lztr1 knockout, Rit1 knockout, and Lztr1; Rit1 double knockout were starved overnight and stimulated with 10% FBS during the indicated times. Protein lysates were immunoblotted as indicated. Immunoblots represent a representative result from two biological replicates. (h) Quantitative PCR analysis of Spry2 and Dusp6 mRNA levels in primary MEF with the indicated genotypes stimulated for 1 hr with 10% FBS (n = 3 biological replicates). p Values were calculated using Student’s t-test. p values: * (p < 0.05); ** (p < 0.01); *** (p < 0.005).

Lztr1 embryonic lethality provides a unique model to undertake a genetic epistatic rescue experiment to assess which RAS GTPase is the critical downstream substrate of the CRL3LZTR1 complex. We have previously shown that at the biochemical level CRL3LZTR1 complex preferentially binds and promotes degradation of RIT1. Therefore, we generated Rit1 knockout mice to test whether Rit1 depletion can rescue Lztr1-mediated embryonic lethality. To avoid potential heterosis, we generated Rit1 knockouts in the C57BL6/N background; mice were born at expected Mendelian ratios, were fertile, and did not exhibit any visible phenotype, similar to a previous Rit1 null mouse strain (Cai et al., 2011). We confirmed the elimination of Rit1 by immunoblot in different tissues, including brain, lung, and heart (Figure 4c). Next, we crossed Rit1 knockout mice with Lztr1 heterozygous mice to obtain both Lztr1-/+/Rit1+/+ and Lztr1-/+/Rit1-/- progeny. As previously shown, a heterozygous breeding scheme with Lztr1-/+/Rit1+/+ mice did not yield any viable Lztr1 knockout mice (Figures 3e and 4d; Sewduth et al., 2020). In contrast, when Lztr1-/+/Rit1-/- breeders were used, we obtained 13/85 (~15.3%) mice that were double knockout (DKO) for both Lztr1 and Rit1 (Figure 4d). This result shows that Rit1 deletion can rescue the embryonic lethality caused by Lztr1 deletion in mice, indicating that Rit1 is the critical substrate of the CRL3LZTR1 complex during embryonic development. The resulting DKO mice appeared normal, were fertile, and were absent of any detectable phenotype that resembled other NS mouse models, as assessed by size, heart weight, and cranial morphology (Figure 4—figure supplement 1). In addition, Lztr1/Rit1 DKO embryos harvested at E18.5 showed a complete rescue of the cardiac phenotype, which were indistinguishable from Rit1 KO embryos (Figure 4e, Figure 4—figure supplement 1e). To assess whether the Rit1 dependency established by Lztr1/Rit1 DKO mice is correlated to a rescue of MAPK pathway hyperactivation, we isolated primary mouse embryonic fibroblasts (MEF) from wild type, Lztr1 KO, Rit1 KO, and Lztr1/Rit1 DKO embryos (Figure 4f). We then subjected these MEFs to FBS stimulation and observed a noticeable decrease in MAPK signaling in cells devoid of both LZTR1 and RIT1 compared to LZTR1 KO cells, both by immunoblot and using the two well-characterized MAPK-regulated transcriptional targets Spry2 and Dusp6 (Figure 4g–h). This indicated that, in our MEF cells, MAPK pathway hyperactivation is mediated by RIT1 protein stabilization in the absence of CRL3LZTR1 regulation.

Many human RAS proteins are highly conserved in less complex organisms and a common ancestor Ras protein has been described in distant relatives such as yeast and slime mold, among other Eukarya (Cox and Der, 2010). Phylogenetic analyses of certain regulators of RAS function, such as GAPs and GEFs, highlight the prominence of early RAS activity regulation and its robust coevolution; with ancestral genomes harboring multiple GAP and GEF genes for a single RAS GTPase and a notable expansion of GAP genes congruent with the expansion of RAS GTPases in eukaryotes (van Dam et al., 2011; van Dam et al., 2009). The high percentage of GTP-loaded RIT1 in mammalian cells indicates the potential absence or evolutionary loss of GAP-mediated regulation and highlights the importance of alternative RIT1 regulatory mechanisms, such as protein turnover. Here, we show that in contrast to the highly conserved protein KRAS, both RIT1 and LZTR1 proteins are less conserved in lower organisms. Despite this, LZTR1 orthologs of invertebrate model organisms retain preferential binding toward their respective RIT1, but not RAS, orthologs. Moreover, fruit flies exhibit strong conservation of the LZTR1-mediated protein degradation mechanism with specificity toward RIT1, as we had previously described with human orthologs, but minimal activity toward other fruit fly RAS GTPases. These data indicate a functional co-evolution between LZTR1 and RIT1.

We also demonstrate that, in contrast to previous reports, LZTR1 is haplosufficient in mice (Steklov et al., 2018). However, homozygous knockout mutations are not tolerated due to embryonic lethality caused by aberrant cardiovascular development concomitant with severe peripheral hemorrhaging and ventricular septal defects, which can be rescued by Rit1 germline deletion. Interestingly, Sewduth et al., demonstrate that pharmacological inhibition of the RAF/MEK/ERK MAPK or AKT pathway was unable to fully rescue the embryonic lethality and vascular defects of mice with conditional Lztr1 KO in blood vessels (Sewduth et al., 2020). Given that our Lztr1/Rit1 DKO mice did not exhibit cardiovascular abnormalities during embryogenesis, excessive accumulation of Rit1 protein in Lztr1 KO cells likely results in the hyperactivation or dysregulation of additional Rit1 effector pathways that contribute to Lztr1-/- embryonic lethality (Cuevas-Navarro et al., 2021; Meyer Zum Büschenfelde et al., 2018; Vichas et al., 2021).

While other RAS GTPases have cognate GAPs and GEFs that regulate their activity, RIT1 GAPs and GEFs remain to be identified. Supported by the robust functional and evolutionary relationship between LZTR1 and RIT1 described here, we posit that LZTR1 may have co-evolved with RIT1 as an alternative mechanism to regulate RIT1 GTPase activity.

Materials and methods

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Escherichia coli)BL21(DE3)NEBC2527H
Genetic reagent (D. melanogaster)Lztr11This paperCG3711/Lztr1 knockout
Genetic reagent (D. melanogaster)Lztr12This paperCG3711/Lztr1 knockout
Genetic reagent (D. melanogaster)RasHAThis paperTransgene of HA-Ras at attP-86Fb landing site
Genetic reagent (D. melanogaster)RicHAThis paperTransgene of HA-Ric at attP-86Fb landing site
Genetic reagent (M. musculus)Lztr1-/-EUCOMMLztr1tm1a(EUCOMM)Wtsi; RRID: IMSR_EM:06794Lztr1 knockout
Genetic reagent (M. musculus)Rit1-/-This paperRit1 knockout
Genetic reagent (M. musculus)Lztr1-/-;Rit1-/-This paperLztr1 and Rit1 double knockout
Genetic reagent (M. musculus)129S1/SvlmjJackson Laboratories002448; RRID:IMSR_JAX:002448
Recombinant DNA reagentpcDNA3-DEST-Flag-SOScat (H. sapiens)This paperVector to express Flag-SOS1 (residues 564–1049) in mammalian cells.
Recombinant DNA reagentpcDNA3-DEST-HA-LZTR1 (H. sapiens)This paperVector to express HA-LZTR1 in mammalian cells.
Recombinant DNA reagentpcDNA3-DEST-HA-LZTR1 (M. musculus)This paperVector to express HA-LZTR1 in mammalian cells.
Recombinant DNA reagentpcDNA3-DEST-HA-LZTR1 (D. rerio)This paperVector to express HA-LZTR1 in mammalian cells.
Recombinant DNA reagentpcDNA3-DEST-HA-LZTR1 (D. melanogaster)This paperVector to express HA-LZTR1 in mammalian cells.
Recombinant DNA reagentpGEX-6P-DEST-KRAS (H. sapiens)This paperVector to express GST-KRAS in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-RIT1 (H. sapiens)This paperVector to express GST-RIT1 in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-KRAS (M. musculus)This paperVector to express GST-KRAS in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-RIT1 (M. musculus)This paperVector to express GST-RIT1 in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-KRAS (D. rerio)This paperVector to express GST-KRAS in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-RIT1 (D. rerio)This paperVector to express GST-RIT1 in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-RAS (D. melanogaster)This paperVector to express GST-KRAS in E. coli cells.
Recombinant DNA reagentpGEX-6P-DEST-RIC (D. melanogaster)This paperVector to express GST-RIT1 in E. coli cells.
AntibodyAnti-HA (Rabbit monoclonal)Cell Signalling TechnologyCat#: 3724; RRID: AB_1549585WB (1:3,000)
AntibodyAnti-Flag (Rabbit monoclonal)Cell Signalling TechnologyCat#: 14793; RRID: AB_2572291WB (1:3,000)
AntibodyAnti-p-ERK1/2 (Rabbit monoclonal)Cell Signalling TechnologyCat#: 4370; RRID: AB_2315112WB (1:1,000)
AntibodyAnti-ERK1/2 (Rabbit monoclonal)Cell Signalling TechnologyCat#: 4696; RRID: AB_390780WB (1:2,000)
AntibodyAnti-p-MEK1/2 (Rabbit monoclonal)Cell Signalling TechnologyCat#: 9154; RRID: AB_2138017WB (1:1,000)
AntibodyAnti-MEK1/2 (Mouse monoclonal)Cell Signalling TechnologyCat#: 4694; RRID: AB_10695868WB (1:1,000)
AntibodyAnti-RIT1 (Rabbit polyclonal)AbcamCat#: ab53720; RRID: AB_882379WB (1:1,000)
AntibodyAnti-b-Actin (Mouse monoclonal)Sigma-AldrichCat#: A2228; RRID: AB_476697WB (1:10,000)
AntibodyAnti-a-Tubulin (Mouse monoclonal)Sigma-AldrichCat#: T6199; RRID: AB_477583WB (1:10,000)
AntibodyAnti-KRAS (Mouse monoclonal)Sigma-AldrichCat#: WH0003845M1; RRID: AB_1842235WB (1:500)
AntibodyAnti-Ras (Rabbit monoclonal)Cell Signalling TechnologyCat#: 4370; RRID: AB_2910195WB (1:1,000)
AntibodyAnti-NRAS (Mouse monoclonal)Santa Cruz BiotechnologyCat#: sc-31; RRID: AB_628041WB (1:1,000)
AntibodyAnti-HRAS (Rabbit polyclonal)Santa Cruz BiotechnologyCat#: sc-520; RRID: AB_631670WB (1:500)
AntibodyAnti-LZTR1 (Mouse monoclonal)Santa Cruz BiotechnologyCat#: sc-390166X; RRID: AB_2910196WB (1:1,000)
Sequence-based reagentDusp6_FThis paperPCR primersTCCTATCTCGGATCACTGGAG
Sequence-based reagentDusp6_RThis paperPCR primersGCTGATACCTGCCAAGCAAT
Sequence-based reagentSpry2_FThis paperPCR primersCATCGCTGGAAGAAGAGGAT
Sequence-based reagentSpry2_RThis paperPCR primersCATCAGGTCTTGGCAGTGT
Sequence-based reagentTbp_FThis paperPCR primersCCTTGTACCCTTCACCAATGAC
Sequence-based reagentTbp_RThis paperPCR primersACAGCCAACATTCACGGTAGA

DNA constructs

Request a detailed protocol

Human, mouse (Mus musculus), zebrafish (Danio rerio), and fruit fly (Drosophila melanogaster) RIT1 and KRAS orthologs were synthesized as E. coli codon-optimized gene blocks (IDT) and cloned into pDONR221 using Gateway BP cloning. SOScat was subcloned from human SOS1 (residues 564–1049) into pDONR221. Gateway LR reaction was used to generate a pGEX-6P destination vector for bacterial expression. Human, mouse (Mus musculus), zebrafish (Danio rerio), and fruit fly (Drosophila melanogaster) LZTR1 orthologs were synthesized as H. sapiens codon-optimized gene blocks (IDT) and cloned into pDONR221. Gateway LR reaction was used to generate either pcDNA3-HA or pcDNA3-Flag-tagged destination vectors for mammalian expression. All plasmids were verified by Sanger sequencing.

Recombinant proteins

Request a detailed protocol

GST-tagged recombinant proteins were expressed in BL21 (DE3) E. coli cells and expression was induced with 0.2 mM IPTG for 14–16 hr at 18 °C. Cells were lysed by sonication in 50 mM Tris-HCl (pH 8.0), 300 mM NaCl, 5% glycerol, 1 mM DTT. Proteins were immobilized on Glutathione Sepharose 4B beads (Cytiva Life Sciences), washed extensively, and stored as a 50% glycerol bead suspension at −20 °C. HA- or Flag-tagged recombinant proteins were expressed in HEK-293T cells by transient transfection and collected in lysates after 24 hr.

Immunoblot

Request a detailed protocol

Protein samples were prepared from frozen mouse tissues using RIPA buffer and a Dounce tissue homogenizer. Drosophila samples were prepared from either frozen L3 instar larvae, adults, or isolated fly heads and were lysed in RIPA lysis buffer with a Dounce homogenizer. Protein lysates were cleared by centrifugation. For immunoblot detection, samples were separated by SDS-PAGE (NuPAGE) and transferred onto nitrocellulose membranes using iBlot2. Membranes were blocked using 2.5% skimmed milk in TBS-T buffer for 1 hr at room temperature and incubated with appropriate primary antibodies overnight. Detection was performed using HRP-linked secondary antibodies and developed with Amersham ECL (Cytiva Life Sciences) and X-ray films.

Antibodies used in this work were: HA (Cell Signaling Technologies, Cat #3724; 1:3,000), β-Actin (Sigma-Aldrich, Cat #A2228; 1:10,000), ɑ-Tubulin (Sigma-Aldrich, Cat #T6199; 1:10,000), LZTR1 (Santa Cruz Biotechnology, Cat #sc-390166X; 1:1,000), RIT1 (Abcam, Cat #ab53720; 1:1,000), p-ERK (Cell Signaling Technologies, Cat #4370; 1:1000), ERK1/2 (Cell Signaling Technologies, Cat #4696; 1:2000), p-MEK (Cell Signaling Technologies, Cat #9154; 1:1000), MEK1/2 (Cell Signaling Technologies, Cat #4694; 1:1000), panRAS (Cell Signaling Technologies, Cat # 67648; 1:1000), NRAS (Santa Cruz Biotechnology, Cat #sc-31; 1:1,000), HRAS (Santa Cruz Biotechnology, Cat #sc-520; 1:500), and KRAS (Sigma-Aldrich, Cat # WH0003845M1; 1:500).

Mice

The Lztr1 allele (Lztr1tm1a(EUCOMM)Wtsi) was previously described (Castel et al., 2019). Briefly, frozen sperm was obtained from the Knockout Mouse Project and IVF was performed on C57BL/6NTac eggs. Heterozygous mice were maintained in C57BL/6NTac congenic background. Every six months, new C57BL/6NTac males (Taconic) were introduced to our colony to avoid genetic drift. Homozygous knockout Lztr1 embryos (E18.5) were obtained from timed pregnancies using a heterozygous x heterozygous breeding scheme. 129S1/Svlmj mice were purchased at the Jackson Laboratories. Mixed C56BL/6 N-129S1 mice were obtained by crossing Lztr1 heterozygous male C56BL/6 N mice with 129S1 females for the first 5 backcrossings. Then,129S1 males were introduced for Y-chromosome fixing. Knockout Lztr1 mice were considered to be in a pure 129S1 background after at least eight backcrosses. Rit1 knockout mice were generated by Cyagen using CRISP/Cas9-mediated large homology arm recombination in C57BL/6NTac fertilized eggs. Briefly, a cassette containing a loxP-STOP-loxP was engineered to be located at the first coding exon of Rit1 and was microinjected with appropriate gRNA and Cas9 mRNA into zygotes and transferred into the oviducts of pseudopregnant females. Mice with germline transmission of the Rit1 allele were used as founders to establish the Rit1 knockout homozygous colony. This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols (#AN165444 and #AN179937) of the University of California San Francisco.

Mouse phenotyping

Request a detailed protocol

Noonan syndrome-like phenotypes in mice were analyzed as previously reported (Castel et al., 2019). Briefly, mice were weaned at 3 weeks of age, weighted at 4 weeks, skull CT was performed at 8 weeks and then mice were euthanized for heart measurements. Sample size was calculated based on anticipated values from our previous phenotyping experiments using a Rit1M90I Noonan syndrome mouse model (Castel et al., 2019). For skull morphometry, sample size was determined to be at least n = 5 to achieve an alpha of 0.05 with a power of 80%. For body weight, sample size was determined to be at least n = 13, and for organ-to-body weight measurements, sample size was determined to be at least n = 6, following the same statistical parameters described above.

For E18.5 embryos, trunks were fixed overnight in 10% buffered formalin, dehydrated with ethanol, and embedded in paraffin. For cardiac analysis, the whole trunk was sectioned transversely at 6 μm and H&E staining was performed following standard protocols at the Histology Core Facility of the National Center for Cardiovascular Research (CNIC). Ventricular wall measurements were obtained from H&E-stained sections. The length of at least 4 lines from a minimum of 8 sections obtained from 3 different embryos was measured with Ruler tool (NDP.View 2 Software). Statistical analyses were carried out using Prism 7 (GraphPad). Data are presented as mean ± SEM unless stated otherwise. Statistical significance was determined by performing 2-tailed, unpaired Student’s t tests when comparing 2 groups. p Values of less than 0.05 are considered significant. At least three independent dissections were performed to obtain E18.5 embryos.

RT-qPCR

Request a detailed protocol

Total RNA from MEFs was isolated using the RNeasy kit (Qiagen) according to the manufacturer’s instructions. cDNA was obtained by reverse transcription (RT) of 1 µg RNA using qScript XLT cDNA SuperMix (QuantaBio; 95161). Ten ng of cDNA was diluted in nuclease-free water and ran in technical triplicates using PowerUp SYBR Green Master Mix (Applied Biosystems) on a QuantStudio 5 (Thermo Fisher Scientific). Tbp (TATA-box binding protein) was used as an endogenous control.

Primer sequences were:

  • Tbp Fw: CCTTGTACCCTTCACCAATGAC; Tbp Rv: ACAGCCAACATTCACGGTAGA

  • Spry2 Fw: CATCGCTGGAAGAAGAGGAT; Spry2 Rv: CATCAGGTCTTGGCAGTGT

  • Dusp6 Fw: TCCTATCTCGGATCACTGGAG; Dusp6 Rv:GCTGATACCTGCCAAGCAAT

Drosophila strains and crosses

Request a detailed protocol

Flies were maintained on standard food medium under a 12:12 light:dark cycle at 25 °C. The Lztr11 allele was generated by CRISPR-Cas9-mediated mutagenesis using transgenic nanos-Cas9 (Kondo and Ueda, 2013; Ren et al., 2013) and guide RNA (gRNA) lines. Briefly, a double gRNA expression construct directed against the following protospacer sequences, 5´- GAAGCAAGCACACAGTGG-3’ and 5´-GATGCGATGTTTGTATTCGG-3’ (both corresponding to the upper strand), was generated in vector pBFv-U6.2B (Kondo and Ueda, 2013) and inserted via ΦC31 integrase-mediated transformation at the attP40 landing site (Bischof et al., 2007). The resulting line was then crossed to nanos-Cas9-expressing flies to isolate Lztr1 mutations, including Lztr11. Lztr12 was obtained via CRISPR-Cas9 engineering by GenetiVision Corporation and verified by PCR amplification and sequencing. HA-tagged Ric- and Ras-expressing constructs were ordered as custom genes from GenScript and transferred into the pattB vector (Bischof et al., 2013). Both contain the corresponding Ric and Ras genomic sequences including their respective 5’ and 3’ regulatory regions, with the HA coding sequence inserted after the initiation codon. Transgenic lines for each construct were established by ΦC31-mediated integration at the attP-86Fb landing site. Both homozygous lines are viable and fertile and were placed in the Lztr12 mutant background by standard crosses.

Quantification of protein abundance by mass spectrometry

Request a detailed protocol

Drosophila head lysates (~300 μg in 100 μL per sample) were precipitated with acetone by adding 400 μL of ice-cold acetone, vortexing the lysates briefly, incubating the lysates at –20 °C for 1 hr, centrifuging at 15,000 x g for 10 min and decanting the supernatants. The protein pellets were then dissolved in 20 μL of 50 mM Tris-HCl buffer, pH 8.0 containing 8 M urea with vortexing. The samples were reduced by adding 1.2 μL of 100 mm dithiothreitol and vortexing for 30 min at room temperature (RT), followed by alkylation with added 3.5 μL of 100 mM iodoacetamide and gentle vortexing for 30 min in the dark for 30 min at RT. One μL from each sample was used to quantify protein amounts using BCA protein assay kit (Thermo Fisher Scientific, USA) according to the manufacturer’s protocol. The samples were diluted by adding 90 μL of 50 mM Tris-HCl buffer, pH 8.0 containing 1 mM CaCl2 and digested with 6 μg trypsin overnight at 37 °C with agitation. The samples were then acidified by adding 1 μL of formic acid, and the digested peptides were desalted using Sep-Pak C18 Classic Cartridges (Waters, USA) using the manufacturer’s protocol.

For LC-MS/MS analysis, digested and desalted peptides were suspended in 0.1% formic acid, and ~0.5 ug was injected per sample on the EasySpray 50 cm C18 column (ES903, Thermo Fisher Scientific, USA) using Acquity UPLC M-Class System (Waters, USA) on line with Orbitrap Fusion Lumos Tribrid Mass Spectrometer (Thermo Fisher Scientific, USA). The column was held at 45 °C, and a 185 min low-pH reversed phase chromatography linear gradient was performed using a conventional two-buffer system (buffer A: 0.1% formic acid in MS-grade water; buffer B: 0.1% formic acid in acetonitrile) by increasing buffer B concentration from 3.5% to 30% over 185 min, followed by a 2 min wash to 50%, with a constant flow rate of 300 nL/min. The mass spectrometer was operated in positive ionization mode with the spray voltage of 2.5 KV, ion transfer tube temperature of 275 °C, RF lens at 30%, and internal calibration set to Easy-IC. MS1 spectra were collected in the Orbitrap in profile mode at 120 K resolution, 375–1500 m/z mass range, 50ms maximum injection time (IT), and automatic gain control (AGC) target of 4.0e5. Precursor ions charged 2 + to 7 + with MS1 intensity above the threshold of 2.0e4 were isolated in the Quadruple using 1.6 Th isolation window, fragmented using higher-energy collisional dissociation (HCD) with 30% collision energy, and detected in the Orbitrap in Centroid mode at 30 K resolution, scan range set to Auto Normal, 100ms maximum IT, and 5.0e4 AGC target. The maximum cycle time was set to 3 s and dynamic exclusion to 30 s with +/-10 ppm tolerance.

MS.raw data files were converted to peak lists using PAVA in-house script and searched with Protein Prospector (v6.0.0) (Chalkley et al., 2008) against Drosophila UniProtKB FASTA-formatted database, which included proteins and their splicing isoforms downloaded on 2019-04-19 (uniprot.org) and a corresponding random-concatenated database of decoy peptides added with Protein Prospector. Instrument was set to "ESI-Q-high-res." Enzyme was set to Trypsin. Up to 2 missed cleavages were allowed and up to 2 post-translational modifications per peptide. "Carbamidomethyl (C)" was set as a constant modification, and default variable modifications were allowed: Acetyl (Protein N-term), Acetyl +Oxidation (Protein N-term M), Gln- > pyro Glu (N-term Q), Met-loss (Protein N-term M), Met-loss +Acetyl (Protein N-term M), Oxidation (M), and Oxidation (P). Precursor charges were set to 2 + to 5+, MS1 mass tolerance to 10 ppm, and MS2 mass tolerance to 20 ppm. FDR filters for peptides were set to 1% and for proteins to 5%. The search results were exported from Protein prospector as tab-delimited text and as.blib spectral library, which was imported into Skyline (v20) along with.raw files for quantification using MS1 filtering (Schilling et al., 2012). The identified peaks were detected by training built-in mProphet model against corresponding random-sequence decoy peptides (Reiter et al., 2011), peaks with an FDR < 1% were integrated. Built-in MSstats (v 3.13) (Choi et al., 2014) plugin in Skyline was used to normalize runs by median centering log2 precursor intensities, to calculate aggregate protein abundances, and to estimate their statistical significance.

Data availability

All data generated or analysed during this study are included in the manuscript and supporting file; Source Data files have been provided for all Figures.

References

    1. Rauen KA
    (2013) The RASopathies
    Annual Review of Genomics and Human Genetics 14:355–369.
    https://doi.org/10.1146/annurev-genom-091212-153523

Decision letter

  1. Alice Berger
    Reviewing Editor; Fred Hutchinson Cancer Research Center, United States
  2. Jonathan A Cooper
    Senior Editor; Fred Hutchinson Cancer Research Center, United States
  3. Alice Berger
    Reviewer
  4. Daniel Abankwa
    Reviewer; University of Luxembourg, Luxembourg

Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Cross-species analysis of LZTR1 loss-of-function mutants demonstrates dependency to RIT1 orthologs" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Alice Berger as Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Jonathan Cooper as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Daniel Abankwa (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) The pulldown experiments in Figure 1b should state the number of times the experiments were each repeated. In addition, a positive control for Ras pulldown should be included to ensure that the GST-Ras fusions are properly folded and capable of interacting with known effectors or other binding proteins.

2) Figure 2a-b or accompanying text should state the number of flies analyzed of each genotype.

3) Figure 3a-d should state the number of mice of each genotype analyzed.

4) Please provide quantification of the rescue phenotype shown in Figure 4e across a population of mice of each genotype.

5) The weakest conclusion in the papers is that Lztr1 inactivation functions through upregulation of MAPK signaling and that this phenotype is reversed by Rit1 inactivation (Figure 4g). The difference in MAPK signaling between Lztr1-/-/Rit1-/- cells and Lztr1-/- cells is not clear. This claim needs to be bolstered by additional evidence. How many repeats were performed, this information is not in the legend? Importantly, data in Figure 4g show that LZTR1 levels are serum-induced and then remain increased long term (24h). Thus, LZTR1 levels peak later than pMEK levels. Surprisingly, the modulation of LZTR1 is not reflected in the RIT1 levels in these blots. Here it would be important to see how RAS (ideally Hras and Kras) levels are modulated. All claims need quantification in particular those relating to pERK differences, which are not as obvious as pMEK differences. Could it be that in this particular background RIT1-levels are relatively uncoupled from LZTR regulation (Figure 4g)? Why?

Reviewer #1 (Recommendations for the authors):

Experimental considerations:

1) The pulldown experiments in Figure 1b should state the number of times the experiments were each repeated. In addition, a positive control for Ras pulldown should be included to ensure that the GST-Ras fusions are properly folded and capable of interacting with known effectors or other binding proteins.

2) Figure 2a-b or accompanying text should state the number of flies analyzed of each genotype.

3) Figure 3a-d should state the number of mice of each genotype analyzed.

4) Please provide quantification of the rescue phenotype shown in Figure 4e across a population of mice of each genotype.

5) The value of the gNOMAD analysis is unclear. Please explain what the screenshot is showing, and/or summarize the data in a more clear way.

6) Please explicitly state the background of both parents in the cross shown in Figure 4d. Is this cross on a pure C57Bl/6 background?

7) The authors have the unique ability to determine RAS and RIT1 levels in cells/tissues from Noonan Syndrome patients with LZTR1 mutation (as shown in Castel et al., Science 2019). To what extent are Ras protein levels altered in those samples? Including this data would enhance the impact of their work.

8) The weakest conclusion in the papers is that Lztr1 inactivation functions through upregulation of MAPK signaling and that this phenotype is reversed by Rit1 inactivation (Figure 4g). This reviewer has trouble seeing the "noticeable" difference in MAPK signaling between Lztr1-/-/Rit1-/- cells and Lztr1-/- cells. This claim needs to be bolstered by additional evidence.

Reviewer #3 (Recommendations for the authors):

1.

a) Figure 1b: Reciprocal pull-down experiments i.e. pulling down with LZTR1, could further strengthen these data. Based on the interaction proteomics data from Steklov et al., it could have been more relevant to examine Hras or Nras in Figure 1b. However, Steklov et al., found also Kras as ubiquitination target of LZTR1. In the current study, Kras may have been used due to its species-wide conservation. However, the above-mentioned background in regards to the other Ras isoforms should be stated or discussed.

b) The abstract states a bit vaguely, 'biochemical dependency', which may be better expressed as 'preferred interaction with RIT1 than RAS orthologs' or similar.

c) Given the constitutive GTP-bound state of RIT1 it may not be surprising that its activity level is regulated 11-fold via LZTR1. While this compares to a mere 1.5-fold regulation of canonical RAS levels in Drosophila, the latter needs to be multiplied by the probably 100 to 1000 fold-change in affinity e.g. for the RBD of the effector Raf, resulting in a total of 150 or 1500-fold activity change (PMID: 7852367). These activity ranges should be the actual values that are considered in order to assess the end-result of the regulation by LZTR1. However, canonical Ras activity will always depend on the coincidence of high local GEF and low GAP activity, while RIT1 might be active 'globally', a constellation particularly advantageous across long distances (e.g. in axons). This RAS-family activity consteallation could be discussed/ considered in their data interpretation.

2.

a) In mice the RIT1 tissue expression levels vary greatly (Figure 3g). Does this inversely correlate with LZTR1 levels in these tissues? If not, it could be assumed that an additional tissue specific modifier of the LZTR1/RIT1 output is highly relevant. In this context, do the RIT1 expression levels somehow correlate with the tissue-specific severity of NS-like phenotypes that are observed in LZTR1 -/- mice?

b) RIT1 and RIT2 are known to be highly expressed during brain development and were implicated in particular in neurite outgrowth (in Colicelli J 2004, PMID: 15367757). The increasing relevance for brain development (or in general complexity also with respect to vasculature) somewhat correlates with the evolutionary pattern the authors describe, namely that a higher divergence for RIT1 and LZTR1 orthologs is seen in less complex organisms. Neurite outgrowth, vascular outgrowth and developmental cell migration appear to often utilise similar molecular machineries, which would all impact on characteristic NS phenotypes. Can the authors therefore speculate on the specific relevance of RIT1 for the observed embryonal bleeding phenotypes (Figure 4a) in LZTR1 ko mice? The point is that it is quite possible that a specific relevance of RIT1 for vascularisation is sufficient to cause the lethal phenotype.

Given the discrepancy with the results reported by Steklov et al., and the cross-breeding data presented here, it is also plausible that additional modifiers are strongly involved in this complex process. In this context, the authors may first wish to consider the increased phenotypic relevance of the LZTR1/RIT1 coupling from fly (hardly any wing vein defects) to mouse (embryonal lethality)? This could be speculated on in the discussion.

c) The rescue in the double-knockout mice may be mostly an outcome of the rescue of this particular vascularization phenotype. It is not entirely clear, whether all LZTR1-/- associated defects are restored, data are mentioned but not shown (L. 264). The authors should provide the data or change their statement. They are encouraged furthermore to discuss the above points and eventually adjust their concluding statements.

d) It could have been interesting to examine the RIT1 and RAS levels in the heterozygous LZTR1 knockout animals from Steklov and see, whether they are increased; yet it is understood this is too much extra work. However, Steklov find Hras and Nras as main interactors of LZTR1, while here typically pan-Ras (unclear in Figure 3g) or Kras (Figure 1b) were considered. Data in Figure 3g need to be quantified, so that the (pan-)Ras effect is clearer. In the end, the fact that Ras is mildly up-modulated across all tissues may be more consistent with the multi-system defects observed in RASopathies.

One may wish to modify the statement that RIT1 is 'the' critical substrate of LZTR1 (L. 261), but rather 'an important'. Furthermore, it should be emphasized that only a 'partial' rescue (L.260) is observed

3.

a) Data in Figure 4g need to be quantified and sufficiently reproduced (i.e. in some repeats). How many repeats were performed, this information is not in the legend? Importantly, data in Figure 4g show that LZTR1 levels are serum-induced and then remain increased long term (24h). Thus, LZTR1 levels peak later than pMEK levels. Surprisingly, the modulation of LZTR1 is not reflected in the RIT1 levels in these blots. Here it would be important to see how RAS (ideally Hras and Kras) levels are modulated. All claims need quantification in particular those relating to pERK differences, which are not as obvious as pMEK differences. Could it be that in this particular background RIT1-levels are relatively uncoupled from LZTR regulation (Figure 4g)? Why?

b) If RIT1 is the major target responsible for the LZTR1 ko phenotype then how can a MEKi rescue the lethal phenotype as reported by Steklov et al.? Does RIT1 also bind to canonical RAS effectors? How are LZTR1 and RIT1 levels modulated with a MEKi? While such MEKi experiments during pregnancy are too laborious, it could be possible to test effects using MEFs such as in Figure 4g.

https://doi.org/10.7554/eLife.76495.sa1

Author response

Essential revisions:

1) The pulldown experiments in Figure 1b should state the number of times the experiments were each repeated. In addition, a positive control for Ras pulldown should be included to ensure that the GST-Ras fusions are properly folded and capable of interacting with known effectors or other binding proteins.

We have now indicated in the figure legend that the pulldown experiment in Figure 1b was repeated 3 times using two different batches of purified GST-tagged protein. Regarding the positive control for Ras interaction, this is a challenging experiment because the known effectors of Ras only bind in the GTP-bound conformation. In Figure 1b, the GST-fused recombinant Ras GTPases used for the pulldown assays are in the GDP-bound conformation, because we had previously described that only the RIT1 GDP-bound conformer binds LZTR1 (Castel et al., 2019). Therefore, interaction of RIT1 to LZTR1 cannot be compared to a classic effector (i.e. Raf RBD). Bacterial expression of Ras proteins has been extensively shown to result in enzymatically competent, properly folded, and GDP-bound GTPases (Lacal et al., 1984; Fang et al., 2016). In fact, we routinely use these recombinant proteins for nucleotide exchange, an additional proof that these enzymes are properly folded.

Nevertheless, to try to address this important comment by the reviewers, we have decided to test the interaction with a catalytic fragment of SOS1 (SOScat; residues 550 to 1050) that is sufficient for Ras-specific nucleotide exchange activity. SOScat contains the Cdc25 domain, which was previously demonstrated to bind RAS, distort the nucleotide binding site, and promote the release of either GDP or GTP (Boriack-Sjodin et al., 1998; Margarit et al., 2003). Given the highly conserved nature of the Cdc25 domain, we used the human protein to assess the structural integrity of all the RAS orthologs. As expected, FLAG-tagged SOScat was able to bind all the recombinant RAS, but not RIT1, orthologs. This data has now been included as Figure 1 – supplement 1.

2) Figure 2a-b or accompanying text should state the number of flies analyzed of each genotype.

We have now indicated in the figure legend that we scored 209 flies in this experiment.

3) Figure 3a-d should state the number of mice of each genotype analyzed.

We have now indicated in the figure legend the number of mice analyzed in each experiment.

4) Please provide quantification of the rescue phenotype shown in Figure 4e across a population of mice of each genotype.

We have now included the quantification of compact and trabecular myocardial thickness in Lztr1-/- and Lztr1-/;Rit1-/- embryos shown in Figure 4e.

In addition, based on the comments from multiple reviewers, we have now included an analysis of Noonan syndrome-like traits in the Lztr1-/-;Rit1-/- double knockout adult mice (Figure 4-supplement 1). In these mice, we measured body weight, craniofacial dysmorphia by micoCT, and heart and spleen weight. As previously indicated, these mice did not display any noticeable phenotype.

5) The weakest conclusion in the papers is that Lztr1 inactivation functions through upregulation of MAPK signaling and that this phenotype is reversed by Rit1 inactivation (Figure 4g). The difference in MAPK signaling between Lztr1-/-/Rit1-/- cells and Lztr1-/- cells is not clear. This claim needs to be bolstered by additional evidence. How many repeats were performed, this information is not in the legend? Importantly, data in Figure 4g show that LZTR1 levels are serum-induced and then remain increased long term (24h). Thus, LZTR1 levels peak later than pMEK levels. Surprisingly, the modulation of LZTR1 is not reflected in the RIT1 levels in these blots. Here it would be important to see how RAS (ideally Hras and Kras) levels are modulated. All claims need quantification in particular those relating to pERK differences, which are not as obvious as pMEK differences. Could it be that in this particular background RIT1-levels are relatively uncoupled from LZTR regulation (Figure 4g)? Why?

This is a very important point. We agree that the differences in MAPK are mild; this is generally the case in Noonan syndrome alleles in which the dysregulation of the MAPK pathway is mild and differences can only be seen upon stimulation with growth factors. Due to this small increase in MAPK activation in the Lztr1-/- cells, it is also difficult to appreciate the rescue in the Lztr1-/-/Rit1-/- cells. Given that immunoblot is a semiquantitative assay, we have now undertaken quantitative PCR of known and well-characterized MAPK target genes (Spry2 and Dusp6) (Ekerot et al., 2008; Ozaki et al., 2001; Wagle et al., 2018). Using this assay, we can conclude that Lztr1/-/Rit1-/- cells rescue the increased MAPK activity in response to growth factor stimulation (Figure 4h). We have included this new data and included the number of repeats in the figure legend for both immunoblot and qPCR as requested.

We have previously observed changes in LZTR1 protein levels upon serum stimulation in other cell types and we are currently following up on this interesting observation. However, these changes in LZTR1 levels are not associated with changes in RIT1 protein levels. This is consistent with our unpublished observations that low levels of LZTR1 do not necessarily result in higher RIT1 levels (as assessed by knocking down LZTR1 with different degrees of efficacy), while complete knockout of LZTR1 results in RIT1 accumulation. We think this is explained by the fact that LZTR1-mediated degradation of RIT1 is extremely efficient and even in the presence of low levels of LZTR1, there is still proficient RIT1 proteolysis. These observations are in line with the haplosufficiency seen in mice and humans for NS phenotypes. Moreover, it is likely that RIT1 ubiquitination by CRL3LZTR1 requires a molecular trigger that is yet to be discovered; therefore, levels of LZTR1 are not necessarily a good readout of RIT1 levels.

Reviewer #1 (Recommendations for the authors):

Experimental considerations:

1) The pulldown experiments in Figure 1b should state the number of times the experiments were each repeated. In addition, a positive control for Ras pulldown should be included to ensure that the GST-Ras fusions are properly folded and capable of interacting with known effectors or other binding proteins.

We have now included the number of times the Figure 1b experiment has been repeated. We have also included the data from an additional repeat in which we have included a positive control. Note that the binding between LZTR1 and RIT1 is GDP-dependent, so these recombinant proteins are not expected to bind to known effectors such as Raf kinases (GTP-dependent interaction). Therefore, we have used SOS1 catalytic domain (SOScat), which is highly conserved and can bind GDP-bound RAS. GST-RAS recombinant proteins are well-documented to fold properly and are purified in their GDP-bound form.

2) Figure 2a-b or accompanying text should state the number of flies analyzed of each genotype.

We have now included the number of flies analyzed in these experiments.

3) Figure 3a-d should state the number of mice of each genotype analyzed.

We have now included the number of mice analyzed in these experiments.

4) Please provide quantification of the rescue phenotype shown in Figure 4e across a population of mice of each genotype.

We have now included the quantification for the cardiovascular rescue.

5) The value of the gNOMAD analysis is unclear. Please explain what the screenshot is showing, and/or summarize the data in a more clear way.

We have clarified the results from the gNOMAD analysis in the main text and have removed the supplementary figure containing the screenshot to avoid confusion (we came to the conclusion that the screenshot does not add any additional valuable information).

6) Please explicitly state the background of both parents in the cross shown in Figure 4d. Is this cross on a pure C57Bl/6 background?

Yes, this experiment was carried out in the same background (C57BL/6N). We have now clarified this in the figure legend.

7) The authors have the unique ability to determine RAS and RIT1 levels in cells/tissues from Noonan Syndrome patients with LZTR1 mutation (as shown in Castel et al., Science 2019). To what extent are Ras protein levels altered in those samples? Including this data would enhance the impact of their work.

This is a great point. We have analyzed these samples in the past and found that RAS levels inconsistently change between the different patient samples. This is also the case in other fibroblast lines derived from additional Noonan syndrome families (for which we do not have matched parents and/or unaffected sibling samples). However, because these variants (SNV, splice, early stop codon) are not necessarily equivalent to the null mutations characterized in our mouse and fruit fly models, we think it is better to not include this data within the manuscript as it can lead to confusion.

8) The weakest conclusion in the papers is that Lztr1 inactivation functions through upregulation of MAPK signaling and that this phenotype is reversed by Rit1 inactivation (Figure 4g). This reviewer has trouble seeing the "noticeable" difference in MAPK signaling between Lztr1-/-/Rit1-/- cells and Lztr1-/- cells. This claim needs to be bolstered by additional evidence.

We agree that the changes in MAPK signaling are small and difficult to notice when analyzed by immunoblot. This is very common given that the Noonan syndrome alleles weakly activate the signaling and western blot is not a very quantitative technique. To address this point, we have now undertaken transcriptional analysis by quantitative PCR of two well-known MAPK-regulated genes, Spry2 and Dusp6. We now show in a more quantitative manner that deletion of Rit1 decreases the degree of MAPK activation in Lztr1-/- MEFs (new Figure 4h).

Reviewer #3 (Recommendations for the authors):

1.

a) Figure 1b: Reciprocal pull-down experiments i.e. pulling down with LZTR1, could further strengthen these data. Based on the interaction proteomics data from Steklov et al., it could have been more relevant to examine Hras or Nras in Figure 1b. However, Steklov et al., found also Kras as ubiquitination target of LZTR1. In the current study, Kras may have been used due to its species-wide conservation. However, the above-mentioned background in regards to the other Ras isoforms should be stated or discussed.

We thank the reviewer for these suggestions. As the reviewer points out, the reason we decided to use KRAS for our analysis and in vitro experiments is because Drosophila Ras1 (the only RAS gene in this organism) is more similar to human KRAS than to NRAS or HRAS. In our previous publication (Castel et al., 2019 Science), using pulldown assays, we did not see measurable LZTR1 binding to any of the classical RAS proteins KRAS, NRAS or HRAS.

Regarding reciprocal pull downs, we have attempted this in the past, but have not been very successful. We hypothesize that because the N-terminal tag is in the Kelch repeats (domain required for RIT1 interaction), antibodies used to pull down LZTR1 compete with RIT1 interaction.

b) The abstract states a bit vaguely, 'biochemical dependency', which may be better expressed as 'preferred interaction with RIT1 than RAS orthologs' or similar.

We have now changed this statement as recommended by the reviewer.

c) Given the constitutive GTP-bound state of RIT1 it may not be surprising that its activity level is regulated 11-fold via LZTR1. While this compares to a mere 1.5-fold regulation of canonical RAS levels in Drosophila, the latter needs to be multiplied by the probably 100 to 1000 fold-change in affinity e.g. for the RBD of the effector Raf, resulting in a total of 150 or 1500-fold activity change (PMID: 7852367). These activity ranges should be the actual values that are considered in order to assess the end-result of the regulation by LZTR1. However, canonical Ras activity will always depend on the coincidence of high local GEF and low GAP activity, while RIT1 might be active 'globally', a constellation particularly advantageous across long distances (e.g. in axons). This RAS-family activity consteallation could be discussed/ considered in their data interpretation.

This is an important point and we would like to clarify. The interaction between LZTR1 and RIT1 is GDPdependent, as we had previously described (Castel et al., 2019 Science). Recombinant RIT1 produced in bacteria is found in the GDP-loaded state; it is in mammalian cells where we have previously found high GTP loading of RIT1 (even upon growth factor starvation). We attribute this observation to the selective degradation of the GDP bound form by CRL3LZTR1.

2.

a) In mice the RIT1 tissue expression levels vary greatly (Figure 3g). Does this inversely correlate with LZTR1 levels in these tissues? If not, it could be assumed that an additional tissue specific modifier of the LZTR1/RIT1 output is highly relevant. In this context, do the RIT1 expression levels somehow correlate with the tissue-specific severity of NS-like phenotypes that are observed in LZTR1 -/- mice?

This is a very interesting question, but unfortunately, we are unable to measure the levels in a panel of mouse tissues. While the RIT1 antibody works well for this application, in our hands, the LZTR1 antibody immunoreacts nonspecifically in many tissues, which makes it impossible to assess the levels.

b) RIT1 and RIT2 are known to be highly expressed during brain development and were implicated in particular in neurite outgrowth (in Colicelli J 2004, PMID: 15367757). The increasing relevance for brain development (or in general complexity also with respect to vasculature) somewhat correlates with the evolutionary pattern the authors describe, namely that a higher divergence for RIT1 and LZTR1 orthologs is seen in less complex organisms. Neurite outgrowth, vascular outgrowth and developmental cell migration appear to often utilise similar molecular machineries, which would all impact on characteristic NS phenotypes. Can the authors therefore speculate on the specific relevance of RIT1 for the observed embryonal bleeding phenotypes (Figure 4a) in LZTR1 ko mice? The point is that it is quite possible that a specific relevance of RIT1 for vascularisation is sufficient to cause the lethal phenotype.

Given the discrepancy with the results reported by Steklov et al., and the cross-breeding data presented here, it is also plausible that additional modifiers are strongly involved in this complex process. In this context, the authors may first wish to consider the increased phenotypic relevance of the LZTR1/RIT1 coupling from fly (hardly any wing vein defects) to mouse (embryonal lethality)? This could be speculated on in the discussion.

This is an interesting point. (Meyer Zum Büschenfelde et al., 2018) have demonstrated that RIT1 can regulate cell migration and actin dynamics though an association with the Rho family GTPases, CDC42 and Rac1. These GTPases engage with cellular developmental programs (such as planar cell polarity) that coordinate the migration and organization of tissues during development, including the formation of vascular networks and neurite migration. One can speculate in higher organisms, upregulation of RIT1 protein levels during development may have profound effects on the organization of certain tissues, such as the vasculature, potentially through MAPK-independent mechanisms, such as the CDC41/Rac1 signaling node.

The difference in the phenotypic relevance of LZTR1/RIT1 between mice and flies could be attributed to differences in the spatial and temporal expression of RIT1 or LZTR1 (as well as the expression and/or existence of different downstream effector pathways regulated by RIT1) during the development of these two organisms.

c) The rescue in the double-knockout mice may be mostly an outcome of the rescue of this particular vascularization phenotype. It is not entirely clear, whether all LZTR1-/- associated defects are restored, data are mentioned but not shown (L. 264). The authors should provide the data or change their statement. They are encouraged furthermore to discuss the above points and eventually adjust their concluding statements.

This is an important point. We think that the rescue is due to the improvement of both the cardiac and the vascular phenotype. As shown in Figure 4e, in E18.5 embryos the cardiac phenotype is completely rescued. Unfortunately, we have not quantified the vascular phenotype, but we did not find any double knockout embryo with hemorrhages.

To further characterize the rescued phenotype of the double knockout animals, we have now included the data of our morphometric analysis of adult animals for NS-like phenotype. We did not find any noticeable differences in body weight, skull morphology, or heart and spleen weight (Figure 4 – supplement 1). In the future, it would be interesting to further phenotype these mice (especially when aged) to see if there are any other phenotypes that could be linked to LZTR1 loss independently of RIT1.

d) It could have been interesting to examine the RIT1 and RAS levels in the heterozygous LZTR1 knockout animals from Steklov and see, whether they are increased; yet it is understood this is too much extra work. However, Steklov find Hras and Nras as main interactors of LZTR1, while here typically pan-Ras (unclear in Figure 3g) or Kras (Figure 1b) were considered. Data in Figure 3g need to be quantified, so that the (pan-)Ras effect is clearer. In the end, the fact that Ras is mildly up-modulated across all tissues may be more consistent with the multi-system defects observed in RASopathies.

One may wish to modify the statement that RIT1 is 'the' critical substrate of LZTR1 (L. 261), but rather 'an important'. Furthermore, it should be emphasized that only a 'partial' rescue (L.260) is observed

We have now included immunoblot analysis of RAS and RIT1 proteins in liver and lung protein extracts from Lztr1 heterozygous mice (C57BL6/N background). These mice were originated in the Knockout Mouse Project (KOMP) and are the same line used by (Steklov et al., 2018). As seen in these tissues, there is no consistent upregulation of any RAS protein in the Lztr1 heterozygous mice. Unfortunately, we were unable to assess other tissues, because the LZTR1 antibody shows extensive immunoreactivity in mouse lysates. This data has now included as Figure 3 – supplement 1e.

3.

a) Data in Figure 4g need to be quantified and sufficiently reproduced (i.e. in some repeats). How many repeats were performed, this information is not in the legend? Importantly, data in Figure 4g show that LZTR1 levels are serum-induced and then remain increased long term (24h). Thus, LZTR1 levels peak later than pMEK levels. Surprisingly, the modulation of LZTR1 is not reflected in the RIT1 levels in these blots. Here it would be important to see how RAS (ideally Hras and Kras) levels are modulated. All claims need quantification in particular those relating to pERK differences, which are not as obvious as pMEK differences. Could it be that in this particular background RIT1-levels are relatively uncoupled from LZTR regulation (Figure 4g)? Why?

We have now included in the figure legend the number of repeats for experiment 4g.

As the reviewer has noticed, we have previously observed that LZTR1 levels can be modulated by FBS and we are currently following up on this interesting observation. These changes in LZTR1 protein levels are not associated with changes in RIT1 levels. This is consistent with our unpublished observations that low levels of LZTR1 do not necessarily result in higher RIT1 levels (as assessed by knocking down LZTR1 with different degrees of efficacy), while complete knockout of LZTR1 results in RIT1 accumulation. We think this is explained by the fact that LZTR1-mediated degradation of RIT1 is extremely efficient and even in the presence of low levels of LZTR1, there is still proficient RIT1 proteolysis. These observations are in line with the haplosufficiency seen in mice and humans for Noonan syndrome-related phenotypes. Moreover, it is likely that RIT1 ubiquitination by CRL3LZTR1 requires a molecular trigger that is yet to be discovered; therefore, levels of LZTR1 are not necessarily a good readout of RIT1 protein levels.

Because western blot is a semiquantitative technique, we have now included the results from our experiments in these cells using qPCR of well-known transcriptional targets of the MAPK (Dusp6 and Spry2). We now clearly show that double knockout cells decrease the levels of MAPK activity seen in LZTR1 knockout cells (Figure 4e).

b) If RIT1 is the major target responsible for the LZTR1 ko phenotype then how can a MEKi rescue the lethal phenotype as reported by Steklov et al.? Does RIT1 also bind to canonical RAS effectors? How are LZTR1 and RIT1 levels modulated with a MEKi? While such MEKi experiments during pregnancy are too laborious, it could be possible to test effects using MEFs such as in Figure 4g.

RIT1 has been shown by us and others to activate MAPK (Berger et al., 2014; Castel et al., 2019; Fang et al., 2016, p. 1; Lo et al., 2021; Van et al., 2020, p. 1). In fact, RIT1 is one of the most common mutations in Noonan syndrome, which is a disorder driven by increased MAPK response (Aoki et al., 2013). Consistent with this, a few case reports have shown that the cardiac phenotype of RIT1 mutant Noonan syndrome ameliorated upon pharmacological treatment with a MEK1/2 allosteric inhibitor (Andelfinger et al., 2019). A plausible explanation of Steklov’s data is that the embryonic phenotype of LZTR1 knockout mice is due to RIT1-driven MAPK activation. This would be consistent with the fact that in our hands, we rescue this phenotype with the RIT1 knockout allele.

We are currently working on a project that explores the mechanism by which RIT1 can activate the MAPK; our data so far indicates that RIT1 can bind directly Raf (with low affinity), but is likely to still require RAS activity to activate MAPK.

References

Andelfinger G, Marquis C, Raboisson M-J, Théoret Y, Waldmüller S, Wiegand G, Gelb BD, Zenker M, Delrue M-A, Hofbeck M. 2019. Hypertrophic Cardiomyopathy in Noonan Syndrome Treated by MEK-Inhibition. Journal of the American College of Cardiology 73:2237–2239. doi:10.1016/j.jacc.2019.01.066

Aoki Y, Niihori T, Banjo T, Okamoto N, Mizuno S, Kurosawa K, Ogata T, Takada F, Yano M, Ando T, Hoshika T, Barnett C, Ohashi H, Kawame H, Hasegawa T, Okutani T, Nagashima Tatsuo, Hasegawa S, Funayama R, Nagashima Takeshi, Nakayama K, Inoue S-I, Watanabe Y, Ogura T, Matsubara Y. 2013. Gain-of-function mutations in RIT1 cause Noonan syndrome, a RAS/MAPK pathway syndrome. Am J Hum Genet 93:173–180. doi:10.1016/j.ajhg.2013.05.021

Berger AH, Imielinski M, Duke F, Wala J, Kaplan N, Shi G-X, Andres DA, Meyerson M. 2014. Oncogenic RIT1 mutations in lung adenocarcinoma. Oncogene 33:4418–4423. doi:10.1038/onc.2013.581

Boriack-Sjodin PA, Margarit SM, Bar-Sagi D, Kuriyan J. 1998. The structural basis of the activation of Ras by Sos. Nature 394:337–343. doi:10.1038/28548

Castel P, Cheng A, Cuevas-Navarro A, Everman DB, Papageorge AG, Simanshu DK, Tankka A, Galeas J, Urisman A, McCormick F. 2019. RIT1 oncoproteins escape LZTR1-mediated proteolysis. Science 363:1226–1230.

doi:10.1126/science.aav1444

Ekerot M, Stavridis MP, Delavaine L, Mitchell MP, Staples C, Owens DM, Keenan ID, Dickinson RJ, Storey KG, Keyse SM. 2008. Negative-feedback regulation of FGF signalling by DUSP6/MKP-3 is driven by ERK1/2 and mediated by Ets factor binding to a conserved site within the DUSP6/MKP-3 gene promoter. Biochem J 412:287–298. doi:10.1042/BJ20071512

Fang Z, Marshall CB, Yin JC, Mazhab-Jafari MT, Gasmi-Seabrook GMC, Smith MJ, Nishikawa T, Xu Y, Neel BG, Ikura M. 2016. Biochemical Classification of Disease-associated Mutants of RAS-like Protein Expressed in Many Tissues (RIT1). J Biol Chem 291:15641–15652. doi:10.1074/jbc.M116.714196

Lacal JC, Santos E, Notario V, Barbacid M, Yamazaki S, Kung H, Seamans C, McAndrew S, Crowl R. 1984. Expression of normal and transforming H-ras genes in Escherichia coli and purification of their encoded p21 proteins. Proc Natl Acad Sci U S A 81:5305–5309. doi:10.1073/pnas.81.17.5305

Lo A, Holmes K, Kamlapurkar S, Mundt F, Moorthi S, Fung I, Fereshetian S, Watson J, Carr SA, Mertins P, Berger AH. 2021. Multiomic characterization of oncogenic signaling mediated by wild-type and mutant RIT1. Sci Signal 14:eabc4520. doi:10.1126/scisignal.abc4520

Margarit SM, Sondermann H, Hall BE, Nagar B, Hoelz A, Pirruccello M, Bar-Sagi D, Kuriyan J. 2003. Structural evidence for feedback activation by Ras.GTP of the Ras-specific nucleotide exchange factor SOS. Cell 112:685–695.

doi:10.1016/s0092-8674(03)00149-1

Meyer Zum Büschenfelde U, Brandenstein LI, von Elsner L, Flato K, Holling T, Zenker M, Rosenberger G, Kutsche K. 2018. RIT1 controls actin dynamics via complex formation with RAC1/CDC42 and PAK1. PLoS Genet 14:e1007370. doi:10.1371/journal.pgen.1007370

Ozaki K, Kadomoto R, Asato K, Tanimura S, Itoh N, Kohno M. 2001. ERK pathway positively regulates the expression of Sprouty genes. Biochem Biophys Res Commun 285:1084–1088. doi:10.1006/bbrc.2001.5295

Steklov M, Pandolfi S, Baietti MF, Batiuk A, Carai P, Najm P, Zhang M, Jang H, Renzi F, Cai Y, Abbasi Asbagh L, Pastor T, De Troyer M, Simicek M, Radaelli E, Brems H, Legius E, Tavernier J, Gevaert K, Impens F, Messiaen L, Nussinov R, Heymans S, Eyckerman S, Sablina AA. 2018. Mutations in LZTR1 drive human disease by dysregulating RAS ubiquitination. Science 362:1177–1182. doi:10.1126/science.aap7607

Van R, Cuevas-Navarro A, Castel P, McCormick F. 2020. The molecular functions of RIT1 and its contribution to human disease. Biochem J 477:2755–2770. doi:10.1042/BCJ20200442

Wagle M-C, Kirouac D, Klijn C, Liu B, Mahajan S, Junttila M, Moffat J, Merchant M, Huw L, Wongchenko M, Okrah K, Srinivasan S, Mounir Z, Sumiyoshi T, Haverty PM, Yauch RL, Yan Y, Kabbarah O, Hampton G, Amler L, Ramanujan S, Lackner MR, Huang S-MA. 2018. A transcriptional MAPK Pathway Activity Score (MPAS) is a clinically relevant biomarker in multiple cancer types. NPJ Precis Oncol 2:7. doi:10.1038/s41698-018-0051-4

https://doi.org/10.7554/eLife.76495.sa2

Article and author information

Author details

  1. Antonio Cuevas-Navarro

    Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, United States
    Contribution
    Data curation, Investigation, Methodology, Visualization, Writing – original draft, Writing – review and editing
    Competing interests
    No competing interests declared
  2. Laura Rodriguez-Muñoz

    Institute for Molecular Biology of Barcelona, Consejo Superior de Investigaciones Científicas, Barcelona, Spain
    Contribution
    Investigation, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
  3. Joaquim Grego-Bessa

    Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain
    Contribution
    Investigation, Methodology, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0938-2346
  4. Alice Cheng

    Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, United States
    Contribution
    Investigation, Writing – review and editing
    Competing interests
    No competing interests declared
  5. Katherine A Rauen

    1. UC Davis MIND Institute, University of California Davis, Sacramento, United States
    2. Department of Pediatrics, University of California Davis, Sacramento, United States
    Contribution
    Investigation, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
  6. Anatoly Urisman

    Department of Anatomic Pathology, University of California San Francisco, San Francisco, United States
    Contribution
    Formal analysis, Investigation, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8364-5303
  7. Frank McCormick

    Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, United States
    Contribution
    Funding acquisition, Supervision, Writing – review and editing
    Competing interests
    is a consultant for Ideaya Biosciences, Kura Oncology, Leidos Biomedical Research, Pfizer, Daiichi Sankyo, Amgen, PMV Pharma, OPNA-IO, and Quanta Therapeutics and has received research grants from Boehringer-Ingelheim and is a consultant for and cofounder of BridgeBio Pharma
  8. Gerardo Jimenez

    1. Institute for Molecular Biology of Barcelona, Consejo Superior de Investigaciones Científicas, Barcelona, Spain
    2. Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain
    Contribution
    Funding acquisition, Investigation, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
  9. Pau Castel

    Department of Biochemistry and Molecular Pharmacology, New York University Grossman School of Medicine, New York, United States
    Contribution
    Conceptualization, Funding acquisition, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review and editing
    For correspondence
    pau.castel@nyulangone.org
    Competing interests
    PC is a founder and advisory board of Venthera
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4972-4347

Funding

National Cancer Institute (F31CA265066)

  • Antonio Cuevas-Navarro

National Cancer Institute (R35CA197709)

  • Frank McCormick

National Cancer Institute (R00CA245122)

  • Pau Castel

DOD CDMRP Neurofibromatosis Research Program (W81XWH-20-1-0391)

  • Pau Castel

Comunidad de Madrid (Programa “Atracción de Talento”)

  • Joaquim Grego-Bessa

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Tony Huynh and Juan Antonio Camara Serrano for help with microCT imaging and Stephanie Mo for feedback on the manuscript. JG-B was funded by Programa “Atracción de Talento” de la Comunidad de Madrid. This work was supported by the NCI (1F31CA265066 to AC-N), (R35CA197709 to FM), (K99CA245122 to PC) and the Department of Defense Neurofibromatosis Research Program (W81XWH-20-1-0391 to PC). We thank the UCSF Mass Spectrometry Facility and A L Burlingame for providing MS instrumentation support for this project (funded by the NIH grants P41GM103481 and S10OD016229). GJ and LR-M were funded by grants from the Spanish Government (BFU2017-87244-P, PID2020-119248GB-I00 and Predoctoral contract BES-2015–071486).

Ethics

This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols (#AN165444 and #AN179937) of the University of California San Francisco.

Senior Editor

  1. Jonathan A Cooper, Fred Hutchinson Cancer Research Center, United States

Reviewing Editor

  1. Alice Berger, Fred Hutchinson Cancer Research Center, United States

Reviewers

  1. Alice Berger
  2. Daniel Abankwa, University of Luxembourg, Luxembourg

Publication history

  1. Received: December 17, 2021
  2. Preprint posted: January 5, 2022 (view preprint)
  3. Accepted: April 22, 2022
  4. Accepted Manuscript published: April 25, 2022 (version 1)
  5. Version of Record published: May 4, 2022 (version 2)

Copyright

© 2022, Cuevas-Navarro et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,005
    Page views
  • 174
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Antonio Cuevas-Navarro
  2. Laura Rodriguez-Muñoz
  3. Joaquim Grego-Bessa
  4. Alice Cheng
  5. Katherine A Rauen
  6. Anatoly Urisman
  7. Frank McCormick
  8. Gerardo Jimenez
  9. Pau Castel
(2022)
Cross-species analysis of LZTR1 loss-of-function mutants demonstrates dependency to RIT1 orthologs
eLife 11:e76495.
https://doi.org/10.7554/eLife.76495

Further reading

    1. Cancer Biology
    2. Computational and Systems Biology
    Erika K Ramos, Chia-Feng Tsai ... Huiping Liu
    Research Article

    Tumor-initiating cells with reprogramming plasticity or stem-progenitor cell properties (stemness) are thought to be essential for cancer development and metastatic regeneration in many cancers; however, elucidation of the underlying molecular network and pathways remains demanding. Combining machine learning and experimental investigation, here we report CD81, a tetraspanin transmembrane protein known to be enriched in extracellular vesicles (EVs), as a newly identified driver of breast cancer stemness and metastasis. Using protein structure modeling and interface prediction-guided mutagenesis, we demonstrate that membrane CD81 interacts with CD44 through their extracellular regions in promoting tumor cell cluster formation and lung metastasis of triple negative breast cancer (TNBC) in human and mouse models. In-depth global and phosphoproteomic analyses of tumor cells deficient with CD81 or CD44 unveils endocytosis-related pathway alterations, leading to further identification of a quality-keeping role of CD44 and CD81 in EV secretion as well as in EV-associated stemness-promoting function. CD81 is co-expressed along with CD44 in human circulating tumor cells (CTCs) and enriched in clustered CTCs that promote cancer stemness and metastasis, supporting the clinical significance of CD81 in association with patient outcomes. Our study highlights machine learning as a powerful tool in facilitating the molecular understanding of new molecular targets in regulating stemness and metastasis of TNBC.

    1. Biochemistry and Chemical Biology
    2. Cancer Biology
    Adi Amar-Schwartz, Vered Ben Hur ... Rotem Karni
    Research Article

    The mTORC1 substrate, S6 Kinase 1 (S6K1), is involved in the regulation of cell growth, ribosome biogenesis, glucose homeostasis, and adipogenesis. Accumulating evidence has suggested a role for mTORC1 signaling in the DNA damage response. This is mostly based on the findings that mTORC1 inhibitors sensitized cells to DNA damage. However, a direct role of the mTORC1-S6K1 signaling pathway in DNA repair and the mechanism by which this signaling pathway regulates DNA repair is unknown. In this study, we discovered a novel role for S6K1 in regulating DNA repair through the coordinated regulation of the cell cycle, homologous recombination (HR) DNA repair (HRR) and mismatch DNA repair (MMR) mechanisms. Here, we show that S6K1 orchestrates DNA repair by phosphorylation of Cdk1 at serine 39, causing G2/M cell cycle arrest enabling homologous recombination and by phosphorylation of MSH6 at serine 309, enhancing MMR. Moreover, breast cancer cells harboring RPS6KB1 gene amplification show increased resistance to several DNA damaging agents and S6K1 expression is associated with poor survival of breast cancer patients treated with chemotherapy. Our findings reveal an unexpected function of S6K1 in the DNA repair pathway, serving as a tumorigenic barrier by safeguarding genomic stability.