Lysines K117 and K147 play conserved roles in Ras activation from Drosophila to mammals

Abstract Ras signaling plays an important role in growth, proliferation, and developmental patterning. Maintaining appropriate levels of Ras signaling is important to establish patterning in development and to prevent diseases such as cancer in mature organisms. The Ras protein is represented by Ras85D in Drosophila and by HRas, NRas, and KRas in mammals. In the past dozen years, multiple reports have characterized both inhibitory and activating ubiquitination events regulating Ras proteins. Inhibitory Ras ubiquitination mediated by Rabex-5 or Lztr1 is highly conserved between flies and mammals. Activating ubiquitination events at K117 and K147 have been reported in mammalian HRas, NRas, and KRas, but it is unclear if these activating roles of K117 and K147 are conserved in flies. Addressing a potential conserved role for these lysines in Drosophila Ras activation requires phenotypes strong enough to assess suppression. Therefore, we utilized oncogenic Ras, RasG12V, which biases Ras to the GTP-loaded active conformation. We created double mutants RasG12V,K117R and RasG12V,K147R and triple mutant RasG12V,K117R,K147R to prevent lysine-specific post-translational modification of K117, K147, or both, respectively. We compared their phenotypes to RasG12V in the wing to reveal the roles of these lysines. Although RasG12V,K147R did not show compelling or quantifiable differences from RasG12V, RasG12V,K117R showed visible and quantifiable suppression compared to RasG12V, and triple mutant RasG12V,K117R,K147R showed dramatic suppression compared to RasG12V and increased suppression compared to RasG12V,K117R. These data are consistent with highly conserved roles for K117 and K147 in Ras activation from flies to mammals.


Introduction
Ras signaling is important in development and disease.The KRas, HRas, and NRas genes in mammals are represented by a single Ras gene in Drosophila (referred to in the literature as Ras1, Ras85D, and here referred to as Ras).The E3 ubiquitin ligase Rabex-5 inhibits Drosophila Ras and mammalian HRas and NRas by promoting their mono-and di-ubiquitination (Jura et al. 2006;Yan et al. 2009Yan et al. , 2010;;Xu et al. 2010;Washington et al. 2020).The Cul3-Lztr1 ubiquitin ligase also inhibits Ras by ubiquitination in both flies and mammals (Steklov et al. 2018;Bigenzahn et al. 2018).Thus, multiple means of inhibitory Ras ubiquitination are highly conserved between Drosophila and mammals.In addition to inhibitory ubiquitination, mammalian KRas, HRas, and NRas are all reported to be ubiquitinated at lysines K117 and K147 (Akimov et al. 2018;Filipčík et al. 2017;Wagner et al. 2012;Sasaki et al. 2011;Baker et al. 2013;Udeshi et al. 2013;Yoshino et al. 2019;Mertins et al. 2013).Ubiquitination events at K117 and K147 increase Ras activation by distinct mechanisms.K117 mono-ubiquitination accelerates nucleotide exchange whereas K147 ubiquitination increases GTP-loading and interaction with downstream effectors (Sasaki et al. 2011;Baker et al. 2013;Hobbs et al. 2013;Filipčík et al. 2017;Wagner et al. 2012).
We report here that individual arginine substitution in K117 in Ras G12V , an activated form of Ras, results in phenotypic suppression of Ras G12V phenotypes.Although individual arginine substitution in K147 in Ras G12V did not produce obvious phenotypic suppression of Ras G12V phenotypes, concurrent mutation of both K117 and K147 in Ras G12V results in even greater suppression than individual mutation in K117.These data are consistent with a conserved role for both of these lysines in Ras activation between flies and mammals.

Reproducibility
The reported work represents reproducible experiments that reflect a minimum of 3 well-controlled, independent trials performed by at least 2 different lab members.Many experiments reported here were repeated in excess of 8 times.

Statistical analysis
Wings were measured using ImageJ software.Raw measurements in pixels and normalized measurements appearing in graphs are in Supplementary File 1. Wing size comparisons and categorical analysis were analyzed using GraphPad Prism software.Specifically, 1-way ANOVA analysis with multiple comparisons  or unpaired t-tests (Fig. 2f and g) was used to assess statistical significance of changes in wing size.G3, 2023, 13(11)

Investigation
Chi-square analysis and Fisher's exact tests were applied as appropriate using contingency tables for categorical scoring of wing phenotypes (Fig. 1g and n), categorical analysis of pupal lethality (Fig. 2a and aʹ), and categorical analysis of anterior dorsocentral (aDC) bristle number (Fig. 3g and h).P values are listed in Supplementary File 1.

Ras constructs and Drosophila transgenic lines
UAS Ras G12V,K117R , UAS Ras G12V,K147R , and UAS Ras G12V,K117R,K147R  were cloned into pUAST-attB using the EcoRI and NotI restriction sites.FLAG and His6 sequences MDYKDDDDKRGSHHHHHHALE were added immediately after the EcoRI site (corresponding to the N-terminal nucleotide sequence GAATTCATGGATTACAA GGATGACGACGATAAGAGAGGATCGCATCACCATCACCATCACG CGCTCGAG, EcoRI site underlined) as we did previously with UAS Ras G12V (Washington et al. 2020).An additional stop codon and a NotI site were added immediately after the original stop codon (corresponding to the sequence TAATAAGCGGCCGC, original stop codon underlined).The plasmids were sent to BestGene for injection and generation of transgenic lines at the attp40 locus.Lines were balanced with second chromosome balancers, and then each individual balanced line was allowed to homozygose.Homozygous lines were then maintained as true-breeding homozygous stocks.BestGene confirmed insertion at attp40, and we sequenced genomic DNA to confirm the sequence of each insert.Importantly, UAS Ras G12V,K117R , UAS Ras G12V,K147R , and UAS Ras G12V,K117R,K147R differ in sequence from UAS Ras G12V only at K117 and/or K147.These transgenes are listed in Supplementary Table 1, and protein sequences are listed in Supplementary Table 2.

Drosophila experiments
Crosses were performed at the indicated temperatures on standard Drosophila medium as in our previous work (Yan et al. 2009;Yan et al. 2010;Washington et al. 2020).In each experimental trial, crosses were set up on food from the same batch, and all vials were kept in close proximity in the same box and experienced the same environment; if there were any environmental variables (such as variations in batches of food or slight fluctuations in temperature when incubator doors were opened), all vials in each trial experienced them simultaneously.We cannot rule out slight Fig. 1.Mutations in lysines K117 and/or K147 suppress Drosophila Ras G12V wing phenotypes at 21°C.Experiments were performed at 21°C; male wings are shown in a-g; female wings are shown in h-n.a, h) Control c765-gal4/+ wing.b, bʹ, i, and iʹ) Expressing Ras G12V in the wing using c765-gal4 causes visible patterning phenotypes including differentiation of extra wing vein material.The degree of mispatterning varies from wings of reduced size from multiple abnormalities including ectopic wing veins, crumpling, and blisters (b, i) to normal to overgrown wings with ectopic wing veins (bʹ, iʹ).The range of sizes is evident when graphing wing size (shown later in f and m) or categorizing severity (shown later in g and n).c, j) Expressing Ras G12V,K117R in the wing using c765-gal4 results in significant suppression of the patterning abnormalities although some wing vein disruption is still obvious.Notably, there is little variability in wing size (shown later in f and m) in contrast to Ras G12V wings.d, d′, dʺ, k, k′, and kʺ) Expressing Ras G12V,K147R in the wing using c765-gal4 results in a range of phenotypes such as wings resembling Ras G12V wings including small wings with abnormalities (d, k) and normal to overgrown wings with abnormalities (dʹ, kʹ) but also a number of wings in which the patterning abnormalities are significantly suppressed (dʺ, kʺ).Wing sizes are variable (shown later in f and m).e, l) Expressing Ras G12V,K117R,K147R in the wing using c765-gal4 results in significant suppression of the patterning abnormalities.As with Ras G12V,K117R wings, there is little variability in wing size (shown in f and m) in contrast to Ras G12V wings.f, m) Wing sizes for wings in a-e and h-l were measured and graphed.g, n) Control wings and wings expressing Ras G12V or Ras G12V,K117R,K147R were scored in the categories of "normal" (no abnormalities-white, dotted in left bar and top portion of right bar), "mild" (some ectopic vein material, mostly where the longitudinal veins meet the wing margin-very light gray, bottom portion of right bar), "moderate" (moderate ectopic wings such as image in iʹ-medium gray, top portion of middle bar), "severe" (severe ectopic wing veins similar to the wing shown in dʹ-dark gray, middle portion of middle bar), or "crumpled" (crumpled and/or blistered wings such as the wings shown in b, d, i, and k-black, bottom portion of middle bar).Categorical scoring highlights that there is no overlap in phenotypes between Ras G12V and Ras G12V,K117R,K147R wings.For f and m, ns indicates not significant; *, **, ***, and **** indicate statistically significant; for P values, see Supplementary File 1.For statistical analysis of categorical scoring and P values for g and n, see Supplementary File 1. Genotypes in this figure are as follows: w; c765-gal4/+ (a, h; left-most genotype in graphs in f, g, m, and n), w; UAS Ras G12V /+; c765-gal4/+ (b, bʹ, i, and iʹ; second genotype in graphs in f, g, m, and n), w; UAS Ras G12V,K117R /+; c765-gal4/+ (c, j; third genotype in graphs in f and m), w; UAS Ras G12V,K147R /+; c765-gal4/+ (d, d′, dʺ, k, k′, and kʺ; fourth genotype in graphs in f and m), w; UAS Ras G12V,K117R,K147R /+; c765-gal4/+ (e, l; right-most genotype in graphs in f, g, m, and n).
environmental fluctuations between trials or slight differences in batches of Drosophila food that may contribute to variations between trials.

Image analysis and processing
Adult wings and other structures were photographed using a Nikon DS-Fi3 microscope camera and saved as TIFF files.Raw wing images were converted to grayscale using Adobe Photoshop.Brightness and contrast of wing images were adjusted using Adobe Photoshop to maximize clarity; adjustments were applied to the entire images.Genotypes are summarized in the figure legends, and identifiers are annotated in Supplementary Table 1.

Pupal lethality
For experiments in Fig. 2a and aʹ, pupal cases were scored as empty (reflecting flies surviving to adulthood) or dead (in which dead pupae remained in the pupal case) and counted 18 days after egg-laying.Although activated Ras transgene expression can cause a delay beyond the 10-day developmental period, this is long enough to ensure all delayed pupae have eclosed.Moreover, dead pupae/pharate adults that have not eclosed are easily distinguished from live pupae given their black and shriveled appearance.Percent pupal lethality was then graphed in Fig. 2a and aʹ.Categorical analyses using chi-square and Fisher's exact test were applied as described in the statistical analysis section.

Results and discussion
To address conservation of the role of K117 and K147 between mammalian Ras isoforms and Drosophila Ras, we utilized a Ras G12V construct we created previously (Washington et al. 2020), similar versions of which have been widely used throughout the field for decades (Karim and Rubin 1998).The G12V substitution mutation biases Ras to the active GTP-loaded conformation, but Ras G12V still undergoes a low rate of GTP hydrolysis and nucleotide exchange (Hunter et al. 2015).This makes Ras G12V an excellent background in which to interrogate the reported mammalian roles of K117 and K147 ubiquitination in Ras activation through phenotype severity.We created arginine substitution mutations (to preserve positive charge but prevent conjugation to ubiquitin) in the G12V background to assess the contribution of K117 and K147 to Ras activation: double mutants Ras G12V, K117R and Ras G12V,K147R and triple mutant Ras G12V,K117R, K147R .All transgenes carry the same tags and were inserted at the same attp40 location in the genome to rule out position insertion effects.If neither K117 nor K147 contributes to Ras activation, these double and triple mutants will phenocopy Ras G12V .If only 1 lysine contributes to Ras activation, then that specific double /+; c765-gal4/+ (fourth genotype in graphs in a and aʹ), w; UAS Ras G12V,K117R,K147R /+; c765-gal4/+ (c, cʹ, e, and eʹ; right-most genotype in graphs in a, aʹ, f, and g).mutant should show phenotypic suppression compared to a Ras G12V control and will phenocopy the suppression seen in the triple mutant.If both K117 and K147 contribute to Ras activation, then the triple mutant should show suppression compared to a Ras G12V control and greater suppression than that shown by each individual double mutant.
To assess Ras overexpression phenotypes in the wing, we utilized gal4 driver c765-gal4, which drives expression generally across the wing (Guillen et al. 1995;de Celis et al. 1996) and is typically used as a pan-wing driver but also drives expression in other tissues including generalized expression in the thorax (Gomez-Skarmeta et al. 1996;Yang et al. 2012), in leg discs (Azpiazu and Morata 2002), and in the brain (Rodan et al. 2002).Driving Ras G12V in the wing using c765-gal4 at 21°C results in differentiation of extra and ectopic wing vein material and thickened veins (Fig. 1b, bʹ, i, and iʹ).Ras can promote proliferation and overgrowth, and some wings reach normal size or increase in size (Fig. 1bʹ and iʹ); however, the differentiation of extra wing veins can happen at the expense of wing material thus reducing overall wing size (Fig. 1b and i) compared to a control wing (Fig. 1a and h).The variety of wing sizes is reflected in scatter plots in Fig. 1f and m, and categorical scoring of variable wing phenotypes is shown in Fig. 1g and n.In addition, these wings can have other abnormalities including blisters and folds that cause the wing to crumple resulting in a reduction in apparent wing size.Driving Ras G12V,K117R in the wing with c765-gal4 at 21°C resulted in obvious suppression of the variable wing size and wing differentiation phenotypes including a statistically significant difference in wing size (Fig. 1c and j and scatter plots in Fig. 1f and m; Supplementary File 1) compared to Ras G12V wings (Fig. 1b, bʹ, i, and iʹ; Supplementary File 1).There was a reproducible trend of increased wing size compared to control wings (Fig. 1f and  m), but this was typically not statistically significant (Supplementary File 1).The suppression of wing vein phenotypes might have allowed a Ras-induced overgrowth phenotype to be observed more frequently resulting in this trend; alternatively, this lysine might be involved in distinguishing between downstream The bristle phenotypes are decreased in Ras G12V,K117R,K147R flies.g, h) As shown in a, in wild-type flies, there are 2 aDC macrochaetes, 1 on each side.We scored the number of bristles at the aDC positions for each of these genotypes at 21°C.Flies expressing Ras G12V showed 1-3 bristles at each aDC position.Graphs showing relative percent in each category are shown in g for males and in h for females.Categorical analysis of the categories of 1, 2, or 3 bristles at each aDC position and using chi-square or Fisher's exact tests indicated statistically significant difference between genotypes (see Supplementary File 1 for P values).This was suppressed in Ras G12V,K117R,K147R flies.gʹ, hʹ) Tables summarizing the average number of bristles and SEM at each aDC position for each genotype for (gʹ) males and (hʹ) females.ANOVA analysis indicates a statistically significant increase in the average number of bristles for Ras G12V compared to controls and statistically significant suppression in bristle number for Ras G12V,K117R,K147R compared to Ras G12V (see Supplementary File 1 for P values).i, k) Control c765-gal4/+ thorax.j, l) Expressing Ras G12V,K117R, K147R using c76ap5-gal4 causes bristle phenotypes (arrows) on the thorax similar to Ras G12V phenotypes at 21°C shown in b and c.Genotypes in this Figure are: w; c765-gal4/+ (a, d, i, and k; left-most 2 bars in graphs in g and h), w; UAS Ras G12V /+; c765-gal4/+ (b, e; middle 2 bars in graphs in g and h), w; UAS Ras G12V,K117R,K147R /+; c765-gal4/+ (c, f, j, and l; right-most two bars in graphs in g and h).biological outputs of wing patterning and growth.Driving Ras G12V, K147R with c765-gal4 at 21°C resulted in a highly variable phenotype ranging from wings similar to Ras G12V wings (Fig. 1d, dʹ, k, and kʹ; Supplementary File 1) to individual wings showing some suppression of both the wing vein phenotypes and the reduced wing size phenotypes (Fig. 1dʺ and kʺ), but there was no statistically significant difference in wing size in the population of Ras G12V,K147R wings compared to Ras G12V wings (scatter plot shown in Fig. 1f and m).The variable phenotype of Ras G12V,K147R is difficult to interpret in comparison to Ras G12V .Therefore, we also utilized triple mutant Ras G12V,K117R,K147R .Driving Ras G12V,K117R,K147R with c765-gal4 at 21°C led to more dramatic suppression of the wing vein phenotypes than either double mutant (Fig. 1e and l) and suppression of the variable size phenotypes (Fig. 1f and m) including statistically significantly different wing size compared to Ras G12V,K147R double mutant wings and Ras G12V wings (Supplementary File 1).Categorical scoring of phenotypic severity in Ras G12V,K117R,K147R wings highlights this suppression compared to Ras G12V wings (Fig. 1g and n; Supplementary File 1).As with the Ras G12V,K117R double mutant, there was a reproducible trend of increased wing size compared to control wings (Fig. 1f and m).In some trials, this was statistically significant (example shown in Fig. 1f for males; Supplementary File 1) but not always (example shown in Fig. 1m for females; Supplementary File 1).The increased suppression of the wing vein phenotypes in the triple mutant Ras G12V,K117R,K147R compared to the Ras G12V,K117R double mutant is consistent with the K147R mutation contributing to the suppression.
Driving Ras G12V with c765-gal4 at 25°C results in lethality.Driving Ras G12V,K117R with c765-gal4 at 25°C led to some escapers surviving to adulthood in some trials (Supplementary File 1).In most cases, this was not statistically significant survival compared to Ras G12V controls.The majority of trials resulted in no survival to adulthood of these flies (Fig. 2a and aʹ).Driving Ras G12V, K147R with c765-gal4 at 25°C resulted in lethality as with Ras G12V (Fig. 2a and aʹ).Driving triple mutant Ras G12V,K117R,K147R with c765-gal4 at 25°C resulted in dramatic and statistically significant suppression of lethality compared to Ras G12V controls (Supplementary File 1).We saw variability in the extent of survival but reproducibly observed at least a third of pupae survived to adulthood (Fig. 2a) in some trials, whereas in others, more than 75% of pupae surviving to adulthood (Fig. 2aʹ).Most Ras G12V, K117R,K147R -expressing wings showed clear Ras gain-of-function phenotypes (Fig. 2c, cʹ, e, and eʹ; scatter plot shown in Fig. 2f and  g) compared to control flies (Fig. 2b and d; Supplementary File 1) and consistent with Ras G12V phenotypes from 21°C (Fig. 1).Because the Gal4/UAS system is temperature responsive, having to increase the temperature to 25° to elicit a phenotype in Ras G12V,K117R,K147R similar to Ras G12V at 21°C further supports the differences in severities between these mutants.In addition, the dramatic suppression of lethality for Ras G12V,K117R,K147R despite no significant suppression of lethality for Ras G12V,K117R and Ras G12V,K147R suggests that both K117R and K147R mutations contribute to the suppression of lethality.
While c765-gal4 is often used for its ability to drive expression in the wing, it is important to note that this driver directs expression in other tissues including generally in the thorax (Gomez-Skarmeta et al. 1996;Yang et al. 2012), in the leg discs (Azpiazu and Morata 2002), in the mushroom body, in the fan-shaped body, in the ellipsoid body of the adult brain (Rodan et al. 2002), and possibly in other tissues.Lethality might result from broader expression than in the wing.In fact, even at 21°C, we see bristle phenotypes on the thorax of Ras G12V -expressing flies (Fig. 3b  and e) compared to control flies (Fig. 3a and d).These phenotypes are somewhat suppressed in triple mutant Ras G12V,K117R,K147R (Fig. 3c and f).To highlight this suppression, we counted the number of bristles at the position of the aDC bristles.Control flies (Fig. 3a and d; Supplementary File 1) typically have 1 aDC bristle on the left and 1 on the right.In Ras G12V -expressing flies, we counted 1-3 bristles at this position (although we have not addressed if this was due to bristle duplication or ectopic bristles in the region).We performed categorical analysis (Fig. 3g and h; Supplementary File 1) and ANOVA analysis (Fig. 3gʹ and hʹ; Supplementary File 1) both of which showed statistically significant differences in Ras G12V compared to controls and statistically significant suppression in Ras G12V,K117R,K147R compared to Ras G12V .Interestingly, those Ras G12V,K117R,K147R -expressing flies that survived to adulthood at 25°C exhibited similar bristle phenotypes (Fig. 3j and l) to Ras G12V flies from 21°C (Fig. 3b and e) not seen in control flies (Fig. 3i and k).These bristle phenotypes are consistent with c765-gal4-induced expression in the thorax reported previously (Gomez-Skarmeta et al. 1996;Yang et al. 2012) and Ras G12V and bristle phenotypes on the thorax reported for apterous-gal4 (ap-gal4) (Culí et al. 2001), which expresses in the pattern of the apterous gene including in the third instar dorsal wing, tarsal segment 4, brain, thorax, and elsewhere (Guarner et al. 2014;Kim et al. 2004;Cohen et al. 1992;Moreno and Morata 1999;Azpiazu and Morata 2002;Paul et al. 2013).
As noted earlier, if both K117 and K147 contribute to Ras activation, then triple mutant Ras G12V,K117R,K147R would show suppressed phenotypes compared to Ras G12V and greater suppression than either Ras G12V,K117R or Ras G12V,K147R alone.Taken together, the different degrees of suppression of Ras G12V phenotypes in Figs. 1 and 2 by Ras G12V,K117R and Ras G12V,K117R, K147R are consistent with both lysines K117 and K147 playing a role in the status of Ras activation in Drosophila as seen in mammals.Arginine substitution preserves the positive charge of lysine while preventing modifications such as ubiquitination and acetylation.It is formally possible these substitution mutants are inactivating by affecting Ras structure or its ability to adopt an active conformation, and we have not ruled out such structural defects from K117R and K147R mutants in Drosophila Ras.However, work in mammalian systems showed that K147R in mammalian KRas maintained an active conformation capable of binding the RBD (Sasaki et al. 2011).In addition, K117R mutations are found in HRas in the rasopathy Costello syndrome and in KRas in colorectal cancer (Kerr et al. 2006;Haigis 2017), thus exhibiting gain-of-function phenotypes.Moreover, although triple mutant Ras G12V,K117R,K147R showed suppressed phenotypes compared to Ras G12V , it exhibited clear Ras gain-of-function wing (Figs. 1 and 2) and bristle (Fig. 3) phenotypes at 25°C confirming that it can adopt an active Ras configuration.
An alternate explanation for the activating role of K117 and K147 is modification at these sites.K117 is a site of activating ubiquitination in mammals, and K147 undergoes both activating ubiquitination and activating acetylation events (Akimov et al. 2018;Filipčík et al. 2017;Wagner et al. 2012;Sasaki et al. 2011;Baker et al. 2013;Udeshi et al. 2013;Yoshino et al. 2019;Mertins et al. 2013;Knyphausen et al. 2016;Song et al. 2016).Our data are consistent with a highly conserved role for K117 and K147 modification in activating Ras; we speculate that modification of these lysines was a mechanism for controlling Ras activity in a common ancestor between flies and mammals.The allele phenotypes in this work are consistent with but do not distinguish between activating ubiquitination and acetylation.Future work will be required to explore the modifications at each lysine to elucidate the specific biological roles of ubiquitination and/or acetylation at these sites and to further understand the activating nature of K117R single mutations vs the loss of function we observe here for K117R mutation in the G12V background.We speculate that K117 ubiquitination has a much greater effect on activity than K117 arginine substitution.The E3 ubiquitin ligases that place the ubiquitin at K117 and K147 would play important roles in Ras signaling dynamics; the Ras G12V,K117R and Ras G12V,K147R transgenes would be useful tools in future work to identify and evaluate these E3 enzymes and explore their role in development and disease.

Data availability
Drosophila strains used in this work (listed in Supplementary Table 1, protein sequences listed in Supplementary Table 2; cloning sites and nucleotide sequence of N-and C-termini listed in Materials and methods section) are available upon request.Raw data, normalized data for graphs in Figs.1-3, and P values are listed in Supplementary File 1.The authors affirm that all data necessary for interpreting the data and drawing conclusions are present within the article text, the figures, tables, and Supplementary File 1.
Supplemental material available at G3 online.

Fig. 2 .
Fig. 2. Concurrent mutation in K117 and K147 suppresses Drosophila Ras G12V lethality at 25°C.Experiments were performed at 25°C; male wings are shown in b-cʹ and f; female wings are shown in d-eʹ and g).a, aʹ)The lethality of flies pupating on the vial walls was calculated by counting the number of empty pupal cases and dead pupae.In every trial, all or almost all c765-gal4/+ flies survive to adulthood whereas Ras G12V showed 100% pupal lethality.We observed variability from trial to trial in terms of overall percent lethality for the Ras G12V,K117R (see Supplementary File 1) and Ras G12V,K117R,K147R genotypes but the trend was the same.a) Graph summarizing the percent pupal lethality for the indicated genotypes.Example of a trial with lower level survival of Ras G12V,K117R,K147R flies.ns indicates not significant by Fisher's exact test; **** indicates statistically significant in both chi-square and Fisher's exact tests; for P values, see Supplementary File 1. aʹ) Graph summarizing the percent pupal lethality for the indicated genotypes.Example of a trial with high survival of Ras G12V,K117R,K147R flies.ns indicates not significant by Fisher's exact test; **** indicates statistically significant in both chi-square and Fisher's exact tests; for P values, see Supplementary File 1. b, d) Control c765-gal4/+ wing.c, cʹ, e, and eʹ) Expressing Ras G12V,K117R,K147R in the wing using c765-gal4 results in a range of phenotypes including wings resembling Ras G12V wings from 21°C including small wings with abnormalities (c, e) but also a number of wings in which the patterning abnormalities are less obvious (cʹ, eʹ).Wing sizes are variable (shown in f and g).f, g) Wing sizes of wings from b-eʹ were measured and graphed in scatter plots.f) Graph summarizing the relative apparent male wing area from experiments shown in b-cʹ.g) Graph summarizing the relative apparent female wing area from experiments shown in d-eʹ.For data in f and g, ns indicates not significant; * and *** indicate statistically significant; for P values, see Supplementary File 1. Genotypes in this figure are: w; c765-gal4/+ (b, d; left-most genotype in graphs in a, aʹ, f, and g), w; UAS Ras G12V /+; c765-gal4/+ (second genotype in graphs in a and aʹ), w; UAS Ras G12V,K117R /+; c765-gal4/+ (third genotype in graphs in a and aʹ), w; UAS RasG12V,   K147R

Fig. 3 .
Fig. 3. Concurrent mutation in K117 and K147 suppresses Drosophila Ras G12V bristle phenotypes at 21°C.Experiments in a-f were performed at 21°C; experiments in i-l were performed at 25°C.a, d) Control c765-gal4/+ thorax.aDC bristles are marked with white arrowheads in aʹ.b, and e) Expressing Ras G12V using c765-gal4 causes visible bristle phenotypes (arrows) on the thorax including extra bristles and disrupted bristle patterning.Extra bristles at the position of the aDC bristles are noted with white arrows.c, f)The bristle phenotypes are decreased in Ras G12V,K117R,K147R flies.g, h) As shown in a, in wild-type flies, there are 2 aDC macrochaetes, 1 on each side.We scored the number of bristles at the aDC positions for each of these genotypes at 21°C.Flies expressing Ras G12V showed 1-3 bristles at each aDC position.Graphs showing relative percent in each category are shown in g for males and in h for females.Categorical analysis of the categories of 1, 2, or 3 bristles at each aDC position and using chi-square or Fisher's exact tests indicated statistically significant difference between genotypes (see Supplementary File 1 for P values).This was suppressed in Ras G12V,K117R,K147R flies.gʹ, hʹ) Tables summarizing the average number of bristles and SEM at each aDC position for each genotype for (gʹ) males and (hʹ) females.ANOVA analysis indicates a statistically significant increase in the average number of bristles for Ras G12V compared to controls and statistically significant suppression in bristle number for Ras G12V,K117R,K147R compared to Ras G12V (see Supplementary File 1 for P values).i, k) Control c765-gal4/+ thorax.j, l) Expressing Ras G12V,K117R, K147R using c76ap5-gal4 causes bristle phenotypes (arrows) on the thorax similar to Ras G12V phenotypes at 21°C shown in b and c.Genotypes in this Figureare: w; c765-gal4/+ (a, d, i, and k; left-most 2 bars in graphs in g and h), w; UAS Ras G12V /+; c765-gal4/+ (b, e; middle 2 bars in graphs in g and h), w; UAS Ras G12V,K117R,K147R /+; c765-gal4/+ (c, f, j, and l; right-most two bars in graphs in g and h).