SEARCH

SEARCH BY CITATION

Keywords:

  • GTPase;
  • signaling;
  • Rit;
  • Rin;
  • activation;
  • wing venation;
  • eye development

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The mammalian Rit and Rin proteins, along with the Drosophila homologue RIC, comprise a distinct and evolutionarily conserved subfamily of Ras-related small GTP-binding proteins. Unlike other Ras superfamily members, these proteins lack a signal for prenylation, contain a conserved but distinct effector domain, and, in the case of Rin and RIC, contain calmodulin-binding domains. To address the physiological role of this Ras subfamily in vivo, activated forms of the Drosophila Ric gene were introduced into flies. Expression of activated RIC proteins altered the development of well-characterized adult structures, including wing veins and photoreceptors of the compound eye. The effects of activated RIC could be mitigated by a reduction in dosage of several genes in the Drosophila Ras cascade, including Son of sevenless (Sos), Dsor (MEK), rolled (MAPK), and Ras itself. On the other hand, reduction of calmodulin exacerbated the defects caused by activated RIC, thus providing the first functional evidence for interaction of these molecules. We conclude that the activation of the Ras cascade may be an important in vivo requisite to the transduction of signals through RIC and that the binding of calmodulin to RIC may negatively regulate this small GTPase. Developmental Dynamics 232:817–826, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The Ras subfamily of small GTPases consists of the classic Ras (H-, K-, and N-), R-Ras, Ral, Rheb, TC21/R-Ras2, M-Ras/R-Ras3, and the newest members of the family Rit, Rin, and RIC (Reuther and Der, 2000). As molecular switches, the Ras-related GTPases respond to external signals by exchanging GTP for bound GDP and the GTP-bound active proteins trigger intracellular signaling cascades through their interaction with a variety of target effector proteins (Campbell et al., 1998; Takai et al., 2001). Although many genetic and biochemical studies have defined a critical role for the classic Ras proteins in the regulation of cell growth and differentiation, less is known about the other members of the Ras subfamily (Reuther and Der, 2000). This is particularly true for the mammalian Rit (Ras-like protein in all tissues), Rin (Ras-like protein in neurons), and Drosophila RIC (Ras-related protein which interacts with calmodulin) proteins whose cellular functions are only beginning to be characterized (Lee et al., 1996; Wes et al., 1996; Shao et al., 1999; Spencer et al., 2002a, b; Hynds et al., 2003). Rit, Rin, and RIC constitute a distinct subgroup of the Ras subfamily, sharing approximately 70% identity amongst each other. These proteins share more than 50% sequence identity with Ras, including highly conserved core effector domains, and their GTP binding and hydrolysis activities have been confirmed (Shao et al., 1999). Unlike Ras, the C-termini of Rit, Rin, and RIC lack a typical prenylation motif required for the association of Ras proteins with the plasma membrane. However, all three proteins contain a series of basic amino acids at the C-terminus (Lee et al., 1996; Wes et al., 1996), although its significance in subcellular localization is not clear. Perhaps more importantly, this region serves as a calmodulin-binding domain for both RIC and Rin (Lee et al., 1996; Wes et al., 1996), and recent studies suggest that calmodulin association is necessary for Rin function (Hoshino and Nakamura, 2003).

Recent work has uncovered the ability of Rit and Rin to regulate cell growth, transformation, differentiation, and activity of several signaling pathways used by other Ras family proteins in vitro (Rusyn et al., 2000; Spencer et al., 2002a, b; Hynds et al., 2003). These studies demonstrate that Rit signals to Ras-responsive elements and transforms NIH3T3 cells to tumorigenicity but fails to activate the ERK, JNK, p38, or PI3-kinase/Akt kinases, indicating that Rit regulates growth control by effector pathways distinct from other transforming members of the Ras family, at least in vitro (Rusyn et al., 2000). These biological activities may result, at least in part, from activation of the RalGEF/Ral signaling cascade (Shao and Andres, 2000). In PC6 and SH-SY5Y cells, activation of Rit stimulates differentiation and neurite outgrowth (Spencer et al., 2002a; Hynds et al., 2003). This effect requires the activity of mitogen-activated protein kinase kinase 1 (MEK1). However, whereas Rit and Rin interact with the same putative effectors in the two-hybrid system (Shao et al., 1999), constitutively active Rin does not demonstrate these same biological effects (Rusyn et al., 2000), raising the possibility that there may be a requirement for concomitant calcium/calmodulin (CaM) signaling. Indeed, recent studies suggest that Rin is involved in calcium/CaM-mediated neuronal signaling pathways (Hoshino and Nakamura, 2003). Rin activation is regulated by growth factor-dependent signaling in neuronal cells, suggesting that Rin may regulate signaling cascades within the mature nervous system (Spencer et al., 2002b).

The fruit fly, Drosophila melanogaster, provides a system to investigate the roles of small GTP-binding proteins during development. In particular, genetic dissection of compound eye development was pivotal to the understanding of the complex events involved in Ras-mediated signal transduction (reviewed in Wassarman et al., 1995). Activation of Ras and its downstream effectors is essential to a plethora of developmental phenomena that give rise to the ordered arrays of cells in the eye (Halfar et al., 2001). Because of the ease of genetic manipulation and the extensive understanding of the signaling involved, perturbation of eye development has been a widely used system to dissect signaling pathways in vivo (Thomas and Wassarman, 1999). In this study, we characterized the developmental defects caused by aberrant activation of Drosophila RIC, the probable orthologue of Rit and Rin. Mutations analogous to those that constitutively activate other Ras proteins (Shao et al., 1999; Takai et al., 2001) also activated RIC. When expressed in the developing wing and eye discs, activated RIC caused ectopic production of wing vein and aberrant photoreceptor differentiation. These effects were similar to those observed for activated Ras. Indeed, heterozygous reduction of genes from the Ras pathway suppressed the phenotypes caused by activated RIC. Furthermore, the reduction of calmodulin enhanced the effects of activated RIC. Taken together, these data suggest a mechanism whereby RIC signals either in parallel with or upstream of Ras to produce developmental effects, whereas calmodulin acts to restrict RIC activity.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

RIC Can be Activated by Amino Acid Substitution

The high similarity of RIC to other Ras family GTPases provided a basis for generating potential activating mutations for the Ric gene. Two cognates of activated Ras and one of dominant-negative Ras were generated by site-directed mutagenesis, resulting in the amino acid substitutions G68V, Q117L, and S73N, respectively (Fig. 1). To establish that these cognates act in the predicted manner, RicQ117L, the stronger activating mutation in Rit, was tested in a well-characterized cell culture assay. Activated forms of Rit have been demonstrated to promote differentiation and neurite outgrowth in PC6 cells (Spencer et al., 2002a). A fusion of green fluorescent protein (GFP) with the wild-type form of RIC was able to stimulate a modest proportion of cells to undergo differentiation (Fig. 2). On the other hand, expression of a GFP-RicQ117L fusion protein strongly induced differentiation and neurite formation in this cell line (Fig. 2). Expression of GFP alone was unable to induce any change in these cells. We conclude that the Q to L activating mutation seen in mammalian Rit is also activating for Drosophila RIC.

thumbnail image

Figure 1. Activating mutations in Rit, Ras, and RIC. Alignments of the human Rit and Drosophila Ras85D and RIC protein sequences are shown. The positions of activating substitutions relevant to this study (RasG12V, RasG13Q, and RasQ61L, and equivalents) are indicated with asterisks. The position of a dominant-negative substitution is indicated with “x”. Note the equivalent positions have different numbers due to the amino terminal extensions on Rit and RIC. The canonical G protein domains are overlined and labeled.

Download figure to PowerPoint

thumbnail image

Figure 2. RIC induces neurite outgrowth in PC6 cells. A,B: PC6 cells were transiently transfected with expression vectors encoding green fluorescent protein (GFP) -tagged wild-type RIC, GFP-RICQ117L (A), or GFP alone (B), allowed to grow for 48 hr, fixed with 3.7% (v/v) formaldehyde, and examined by epifluorescence microscopy to identify transfected, GFP-expressing cells. Photomicrographs are typical of fields from experiments performed in triplicate. C: Quantification of the effect of exogenous protein expression on PC6 cell neurite outgrowth. GFP-expressing cells bearing neurites exceeding two cell body diameters were recorded as a percentage of the total number of transfected cells. Approximately 600–800 cells/condition were counted, and data are the mean ± SD values of triplicate experiments.

Download figure to PowerPoint

Each of the RIC mutants, as well as the wild-type gene, was introduced into flies under control of the GAL4-UAS binary expression system (Brand and Perrimon, 1993). The RIC proteins were expressed using an array of GAL4 insertion lines that drive transcription of UAS constructs in various tissues. Strong effects on wing vein patterning were seen with both activated forms of RIC. An extra wing vein was produced in the intervein spaces at locations in the wing consistent for each GAL4 driver (Fig. 3). Expression of either activated RIC using the hairy-GAL4 (h-GAL) driver caused the appearance of an ectopic vein in the distal part of the wing between the anterior margin and L2 in all animals. Most also showed an ectopic vein near the posterior crossvein. As expected, when expressed using an engrailed-GAL4 (e16E) driver that is expressed only in the posterior compartment, an ectopic vein was restricted to the posterior of the wing (not shown). Several other GAL4 driver lines that express strongly in the wing imaginal discs also resulted in the production of ectopic veins in these UAS-activated RIC lines (Fig. 3, and data not shown). Because GAL4-driven expression increases with temperature (van Roessel and Brand, 2000), misexpressing transgenic animals were assayed at 19°C to 29°C. As anticipated, phenotypes were more severe at higher temperatures (Fig. 3). Expression of wild-type Ric showed only mild phenotypes, even in combination with the strongest GAL4 drivers and at 29°C (not shown). Control animals carrying only the UAS-Ric constructs or the GAL4 drivers alone did not show any developmental defects. Like the negative controls, expression of the dominant negative cognate Ric allele failed to cause visible phenotypes in any tissues with any of the tested GAL4 drivers. Furthermore, expression of a UAS construct for dsRNA from Ric also showed no visible effects with any of the drivers. Although it is possible that both constructs fail to knock out RIC activity, failure to see a phenotype seems more likely to indicate that loss of RIC activity does not result in a visible phenotype.

thumbnail image

Figure 3. Ectopic wing vein induced by activated RIC. A: Without a GAL4 driver, U-RicG68V (A) and U-RicQ117L (not shown) do not affect wing vein patterning. B,D,E,G,H: Ectopic wing vein material (arrows) is produced by expression of either activated RicG68V (D,G,H) or activated RicQ117L (B,E) when driven by various GAL4 driver lines. C,F,I: Each of the GAL4 drivers expresses broadly across the wing blade of the imaginal disc, as revealed by β-galactosidase–staining discs from animals also carrying a UAS-lacZ transgene. D,E: With a given driver line, such as hairy-GAL4 (D,E), effects for activated RicQ117L (E) are more severe than for activated RicG68V (D); as with other GAL4-driven assays, defects are more severe at higher temperatures (compare G with H).

Download figure to PowerPoint

In addition to wing vein defects, expression of RicQ117L using GAL-T59 or ey-GAL, both of which express broadly across the developing eye discs (Bonini et al., 1997; Hazelett et al., 1998), resulted in a consistent roughening of the adult eye (Fig. 4). Using ey-GAL, this phenotype was fully penetrant at 29°C. Frequently, but less consistently, growth of ectopic eye and head cuticle material was observed. Expression of RicG68V in the developing eye had little detectable affect on adult eye morphology (not shown). This finding is likely to be due to differential potency of the two activated Ric alleles. Consistent with this explanation, although both activated forms affected vein formation in a similar manner, the amount of ectopic vein material was always greater in animals carrying the U-RicQ117L activating mutation. The greater potency of U-RicQ117L is consistent with the in vitro activity of human RitQ79L (Rusyn et al., 2000). To examine whether there is normally RIC activity in the developing eyes and wings, in situ hybridization to RIC RNA was performed. No signal was detected in imaginal discs above background, suggesting that there is little or no expression of RIC in those tissues (not shown).

thumbnail image

Figure 4. Eye defects induced by activated RIC. A–D: Without a GAL4 driver, U-RicQ117L (A) does not perturb eye morphology, but when expressed in the eye under the control of various GAL4 lines (B–D), eyes are roughened, reduced, and/or severely misshapen. As with wing defects, the eye defects are more severe at higher temperatures (C compared with D). E: Strong defects caused by activated RIC are comparable to those seen with activated RasG13Q at a much lower temperature.

Download figure to PowerPoint

Of interest, the phenotypes resulting from the expression of an activated form of the closest Drosophila Ras homologue, Ras85D, were similar in wings (de Celis, 2003) and eyes (Dominguez et al., 1998) to those shown here for RIC. As seen in Figure 4, the expression of an ey-GAL4 driven Ras85DG13Q, an activating mutation in the G1 domain, created substantial eye disorganization, even at low temperature where GAL4 is much less active. As noted above, a similar but not identical G1 activating mutation in Ric, RicG68V, had no phenotype in the eye. The more potent RicQ117L caused disorganizations on the scale of those seen for Ras85DG13Q but required much higher temperatures. Thus, eye morphology was much more sensitive to perturbation by activated Ras85D than by activated RIC. Furthermore, h-GAL4 driving U-Ras85DG13Q was invariably lethal at any temperature, suggesting that activated Ras is also more potent than activated RIC in other tissues.

Mutations in the Ras/Raf Pathway Suppress Activated RIC

To determine the signaling pathways used by activated RIC, animals expressing activated RIC in wings (h-GAL4, U-RicG68V) or eyes (ey-GAL4, U-RicQ117L) were crossed to flies carrying candidate interacting mutations. Although this approach only allows the detection of genes that interact as heterozygous, having only a 50% reduction in the candidate gene activity, several interacting loci were identified. The most potent suppressors of activated RIC were mutations of genes in the Ras signaling cascade. Although there was some variability among animals of a given genotype, consistent reduction of either the ectopic wing vein or rough eye phenotype was detectable for mutations in Son of sevenless (Sos), rolled-Map kinase (rl), Map kinase kinase (Dsor1), and Ras itself (Ras85D; Figs. 5, 6). Furthermore, a mutation in the Drosophila Src homologue Src42A, enhanced the defects caused by activated RIC (Fig. 5D). This finding is consistent with previous reports that Src42A is antagonistic to Ras signaling in flies (Lu and Li, 1999).

thumbnail image

Figure 5. Interaction of Ras and cam with activated RIC in wings. All wings shown are from animals heterozygous for h-GAL, U-RicG68V. A: Control animals that are also heterozygous for the CyO balancer produce a modest amount of ectopic vein (arrows). Creases in the wing (arrowheads) are due to the CyO and were not scored as ectopic vein. B–H: The remaining panels show representative wings from animals heterozygous for h-GAL, U-RicG68V and heterozygous for a modifier mutation. B,C: Reduction of cam from two different alleles enhances the production of ectopic wing vein. D: Similarly, reduction of Src42A, a known antagonist of the fly Ras pathway, results in enhancement of vein. E–H: Heterozygosity for a loss of Ras85D (E) or other members of the Ras pathway (F–H) reduces the quantity of ectopic vein to varying degrees (also see Table 1).

Download figure to PowerPoint

thumbnail image

Figure 6. Interaction of Ras and cam with activated RIC in eyes. All eyes shown are from animals heterozygous for ey-GAL, U-RicQ117L. A: Control animals that are also heterozygous for the CyO balancer have moderately roughened eyes. B,C: As with wing defects, additional heterozygosity for mutations in calmodulin enhance the defects in eye organization and shape. D–G: Also consistent with the wing defects, reduction of Ras pathway activity suppresses the eye roughening phenotype, although to different extents than seen for wings.

Download figure to PowerPoint

Qualitative examination indicated that the degree to which Ras pathway mutations suppressed the activated RIC phenotype varied somewhat between tissues and with different mutations. To provide some quantitative assessment of relative suppression, the amount of ectopic wing vein was determined as a fraction of the total wing area for at least 10 wings of each genotype. As a comparison, ectopic wing veins of siblings from the same vial that did not carry the modifier mutation were also quantified. Although this method may be insensitive to small changes, it takes into account variability in expressivity for each genotype and gauges the relative strength of interaction for each of the mutations. The quantitation confirmed the observation that Ras85D, MEK, and Sos mutations significantly suppress RIC-mediated ectopic wing vein (Table 1).

Table 1. Quantitation of Ectopic Wing Vein Modifier Mutationsa
MutationControlMutantChange
  • a

    All flies examined were heterozygous for h-GAL, U-RicG68V. Flies also carrying a mutation in a putative modifier locus were compared with their control siblings carrying a balancer chromosome. Wings were photographed, and the amount of ectopic wing vein was determined as a proportion of the total wing area. The change in percentage ectopic vein in the experimental animals relative to the control is expressed as a percentage change in overall ectopic vein material. Statistically significant differences are marked

  • *

    P ≤ 0.05;

  • **

    P ≤ 0.0001.

Ras85DelB/ +0.343 ± 0.039 (n = 13)0.202 ± 0.025 (n = 10)[DOWNWARDS ARROW]41%*
MEKLH110/ +0.613 ± 0.019 (n = 14)0.244 ± 0.025 (n = 13)[DOWNWARDS ARROW]60%**
SosX122/ +0.620 ± 0.045 (n = 10)0.228 ± 0.046 (n = 10)[DOWNWARDS ARROW]63%**
rl10a/ +0.388 ± 0.041 (n = 10)0.286 ± 0.062 (n = 10)[DOWNWARDS ARROW]26%
cam8t/ +0.427 ± 0.049 (n = 12)0.986 ± 0.118 (n = 10)[UPWARDS ARROW]230%**
camn339/ +0.572 ± 0.052 (n = 10)1.399 ± 0.107 (n = 10)[UPWARDS ARROW]244%**
cn bw sp/ +0.582 ± 0.050 (n = 10)0.802 ± 0.097 (n = 12)[UPWARDS ARROW]38%

Mutations in calmodulin Enhance Activated RIC Phenotypes

Because RIC and Rin have been demonstrated to bind calmodulin (CaM; Lee et al., 1996; Wes et al., 1996) and CaM appears to regulate Rin function (Hoshino and Nakamura, 2003), the effect of reduced CaM activity on activated RIC phenotypes was tested. As above, flies carrying h-GAL4, U-RicG68V, or ey-GAL4, U-RicQ117L were crossed to calmodulin mutants. Wings or eyes of activated RIC-expressing flies carrying an allele of cam were compared with those carrying the CyO balancer. Heterozygous reduction of cam activity dramatically enhanced the defects caused by activated RIC. Enhancement was consistent for both the wing and eye assays but was more dramatic for the wing vein phenotype (Figs. 5, 6). Quantitative assessment shows a greater than twofold increase in the ectopic wing vein, further attesting to the potent interaction of activated RIC with cam (Table 1). No modification of the phenotypes from expression of U-Ras85DG13Q was detected with similar reduction of CaM activity (not shown). These data may indicate that the normal function of CaM binding to RIC is to regulate its activity.

RIC Activation Alters Photoreceptor Differentiation in the Developing Eye Disc

Ras pathway activity is pleiotropic in development of the compound eye (Freeman, 1996; Dominguez et al., 1998). Consequently, varying the level or duration of Ras pathway activation leads to different defects in ommatidial morphologies or fates (Halfar et al., 2001). To examine the eye defects in activated RIC animals during development, we dissected and analyzed late third-instar larval imaginal discs from wild-type and transgenic animals expressing activated RicQ117L. At the morphological level, there were clear aberrations in the mutant animals. The eye discs from activated RIC larvae mostly showed a mild curvature, resulting in a recognizable but distorted morphogenetic furrow (Fig. 7). Eye discs from activated Ras85DG13Q animals were more variable, but all were aberrant. Morphologies ranged from greatly reduced eye discs (not shown) to discs with multiple lobes (Fig. 7). The morphogenetic furrow was not detectable for any of the activated Ras85D discs. These phenotypes were never seen in the wild-type control animals.

thumbnail image

Figure 7. Effects of activated RIC and Ras on eye imaginal discs. A–C: Differentiation of photoreceptors in the late third-instar developing eye imaginal disc is shown for animals expressing activated RIC or Ras. A: In wild-type (WT) eye discs, behind the morphogenetic furrow (arrows), the neuronal marker ELAV is detected in regularly spaced clusters of cells that give rise to ommatidial photoreceptors. B: When activated RIC is expressed at 29°C, the imaginal discs are often distorted and show irregularities and gaps in ELAV-positive photoreceptor clusters (arrows). C: For activated Ras at 19°C, discs are dramatically misshapen, having either additional lobes (arrowhead) or missing eye disc material (not shown). These discs often also show irregularly spaced neuronal clusters, or clusters with too many or too few photoreceptors. D–F: Activation of Drosophila ERK, Rolled, is visualized with antibody to diphospho-ERK in developing eye discs. D: In wild-type discs, activated Rolled is detected weakly in developing photoreceptors behind the morphogenetic furrow and strongly in a line of photoreceptors at the furrow. E: In animals expressing U-RicQ117L under the control of the eye-specific driver GMR-GAL4 at 29°C, there is similar or slightly more detectable diphospho-ERK behind the furrow. F: There is dramatically more staining at 19°C in animals expressing U-Ras. Staining of cells at the morphogenetic furrow in D–F is from endogenous activation and serves as an internal control for development of the staining reaction to normalize levels between samples.

Download figure to PowerPoint

To examine the cellular cause of the morphological defects, various molecular markers were used. The late third instar larval imaginal discs from animals of the three genotypes were stained with anti–phosphohistone H3 (PH3) and anti-ELAV. Anti-PH3 marks mitotic cells, and therefore provides an assessment of relative rates of proliferation. There was no difference between the three genotypes in numbers of cells staining with anti-PH3 (not shown). However, staining discs with anti-ELAV, a marker of neurons, including photoreceptors, revealed that both activated RIC and Ras85D caused anomalies in photoreceptor development (Fig. 7B,C). Wild-type eye discs have a very regular hexagonal array of photoreceptor clusters that differentiate behind the morphogenetic furrow (Fig. 7A). However, the eye discs from activated RIC and Ras animals had disorganized arrays with obvious gaps between photoreceptor clusters. Furthermore, many of the clusters contained either too many or too few photoreceptors. To determine whether the gaps were due to apoptotic cell death, in situ end-labeling was performed to detect fragmented DNA. There was no difference in amount of in situ end-labeling between wild-type and activated RIC and Ras85D discs (not shown). We conclude that the observed defects in eye morphology for both activated GTPases are likely due to altered specification of the cells in the eye disc and not due to changes in proliferation or cell death.

Activated RIC Does Not Strongly Stimulate ERK Activation

Despite the high degree of similarity between the mammalian Ras and Rit proteins, activation of Rit does not necessarily stimulate the ERK pathway, the major target of Ras signaling. In fibroblasts, Rit does not detectably stimulate ERK, whereas Rit in neurons induces substantial ERK activation (Rusyn et al., 2000; Spencer et al., 2002a). To determine the relationship between these GTPases and ERK in Drosophila, activation of the fly ERK, Rolled, was examined in animals expressing activated RIC or Ras. As shown above, expression of either activated GTPase resulted in misshapen eyes. Using diphospho-ERK antisera, dramatic activation was detected in eye disks from larvae expressing U-Ras85DG13Q (Fig. 7F). However, expression of activated RIC resulted in minimal ectopic activation of ERK that was nearly indistinguishable from control disks (Fig. 7D,E). However, because heterozygous mutations in rolled partially suppress the rough eye phenotype of U-RicQ117L in the eye (Fig. 6), this finding suggests that mild activation of ERK signaling by U-RicQ117L contributes to but is not the only pathway important in generating the observed rough eye phenotype.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The small GTPases mediate cellular signals that are essential for a diversity of processes (Takai et al., 2001). This large group of regulatory proteins shares striking similarities to each other, although each has unique features. The Ras subfamily that is composed of Rit and Rin in vertebrates and RIC in Drosophila is particularly interesting because of three unusual features. First, they all lack a recognizable lipidation signal to direct membrane localization. Despite the absence of any known lipidation motifs, Rit and Rin are membrane localized, suggesting the presence of novel domains that mediate association with the membrane (Lee et al., 1996). Second, Rin and RIC bind calmodulin in vitro, raising the possibility that these GTPases are regulated by Ca2+ levels in vivo (Lee et al., 1996; Wes et al., 1996; Hoshino and Nakamura, 2003). Third, whereas Rit shares most core effector domain residues with Ras and transforms NIH3T3 cells, Rit uses novel effector pathways to regulate proliferation and transformation (Rusyn et al., 2000).

Here, we reported the developmental and genetic consequences of the activation of Drosophila RIC. Two separate amino acid substitutions that cause constitutive activation of Ras also activate RIC. Although both of these mutations result in phenotypes, RicQ117L is more potent in generating eye and wing defects than RicG68V. These findings are consistent with previous studies using the vertebrate Rit protein (Rusyn et al., 2000). The developing Drosophila wing and eye discs are particularly sensitive to RIC activation. In wings, activated RIC drives the production of excess vein tissue and, in the eye, results in the misspecification of photoreceptor cells. Both of these phenotypes are reminiscent of those described for activated Ras85D. Furthermore, activity of the Ras pathway is essential for production of the activated RIC phenotypes, as reduction of genetic dosage of several Ras pathway genes suppresses those defects. Intriguingly, a functional consequence of interaction of RIC with CaM can also be detected in this genetic assay. Reduction of CaM enhances the activated RIC phenotypes, thus providing in vivo evidence for a role for CaM in regulating RIC activity.

Ras Signaling Mediates Response to RIC

Activated forms of the mammalian Rit and Rin proteins can promote NIH3T3 cell transformation, suggesting a potential oncogenic function for this family of GTPases (Rusyn et al., 2000). However, the mechanisms by which Rit and Rin transduce proliferative signals have remained elusive. Activated Rit fails to stimulate the ERK, JNK, p38 MAPK, or PI3K pathways in COS cells (Rusyn et al., 2000). However, both Rit and Rin cooperate with Raf to promote morphologically transformed foci. Furthermore, activated Rit, but not Rin, can stimulate neurite outgrowth in PC6 cells in a manner that is dependent on Raf and MEK (Spencer et al., 2002a, b). In this work, we find that activated RIC shares a similar relationship with the Ras/Raf pathway. The eye and wing phenotypes resulting from activated RIC depended on activity of the Ras pathway. One possible explanation for this relationship is simply that activated RIC can partially substitute for Ras85D and directly activate its downstream targets. This model seems unlikely because mutations in Ras85D or in Sos, an upstream component of the pathway, also suppress the RIC phenotypes. If an activated form of RIC could directly substitute for Ras85D, reduction of Ras85D or a molecule required for its activation should not affect the RIC phenotypes. It is more likely that activated RIC either signals in parallel with the Ras pathway or stimulates the Ras pathway at some upstream point, perhaps by indirectly promoting ligand production. It is also of interest to note that both activated mammalian Rit and Drosophila RIC can induce robust differentiation and neurite outgrowth in PC6 cells (Spencer et al., 2002a, b), whereas mammalian Rin only modestly stimulates neurite outgrowth (Hoshino and Nakamura, 2003). Perhaps this finding suggests that RIC is more similar functionally to Rit than Rin, despite the presence of calmodulin binding domains in RIC and Rin, but not Rit.

Calmodulin May Be a Negative Regulator of RIC Activity

Calmodulin is a key transducer of intracellular calcium signals. CaM binds to and regulates a large number of diverse cellular proteins. The interaction of CaM with these proteins is regulated by binding of Ca2+ to CaM. Consequently, CaM confers Ca2+ sensitivity to signaling through various molecules, including protein kinases, cyclic nucleotides, cytoskeletal proteins, and ion channels. Through these interactions, CaM plays a role in a vast array of cellular processes that broadly include cell growth, proliferation, movement, and sensory transduction (reviewed in Beckingham et al., 1998; Chin and Means, 2000). CaM and other Ca2+sensor proteins are particularly important in neurons and their target muscles because of the fundamental role Ca2+plays in synaptic development and transmission, neuronal survival, and axon outgrowth and pathfinding (Burgoyne and Weiss, 2001). It is because of this centrality of Ca2+signaling in neurons that RIC was first identified as a potentially important molecule in neuronal development and/or sensory signaling (Wes et al., 1996).

Here, we demonstrate that the reduction of CaM enhances the developmental defects arising from the ectopic activation of RIC. Previous developmental analysis of mutations in CaM uncovered an array of defects caused by reduction or loss of CaM activity (Nelson et al., 1997). Of interest, one of the phenotypes caused by reduction of CaM is the production of ectopic wing vein in a pattern that is strikingly similar to that described above for activated RIC. The similarity in phenotypes and the observed genetic interaction suggests two alternative models for the normal roles of these proteins and the mechanism of developmental defects. The first model is that the ectopic expression of activated RIC results in the depletion of CaM in the developing wing disc, thereby phenocopying a mutation in CaM. Predictably, this finding would result in an enhancement of the phenotype when CaM activity is further reduced by the introduction of a mutant allele. Consistent with this model, recent reports have suggested that CaM is present in limiting quantities in some cells (reviewed in Persechini and Stemmer, 2002). A second model consistent with the data is that CaM normally acts as a negative regulator of RIC signaling. Thus, the reduction of CaM activity by introduction of a mutant allele would increase RIC activity, enhancing the phenotypes caused by ectopic activation. Furthermore, the current data suggest that the wing vein phenotype observed in CaM mutants may be due to a loss of regulation of endogenous RIC activity. Although both models are valid, we favor the latter model for two reasons. First, reduction of CaM activity resulted in several phenotypes in addition to ectopic wing vein, including melanotic scabs, failure of head eversion, and behavioral defects (Nelson et al., 1997). None of these defects was observed with UAS-activated RIC, despite testing with a large bank (>50) of GAL4 drivers with expression in many tissues. Second, ectopic expression of normal RIC did not lead to phenotypes. If RIC merely sequesters CaM, then activation state should not be important for the observed phenotypes. Thus, it seems likely that RIC is one of the multitude of proteins that is regulated by direct interaction with CaM. Interestingly, reports indicate that another small GTPase, Gem, is also negatively regulated by CaM (Fischer et al., 1996). CaM association was shown to inhibit GTP binding (Fischer et al., 1996) and disrupt association with downstream effector proteins (Beguin et al., 2001), thus restricting the functions of Gem. Additional experiments will determine whether a similar regulatory mechanism is used to govern RIC activity.

The genetic data reported here provide the first in vivo analysis of the biological functions of the newly discovered mammalian Rit and Rin and the Drosophila RIC. Consistent with previous studies of Rit and Rin in human cells, we find that RIC acts synergistically with Ras signaling. In the developing eye and wing, we show that RIC can coordinate with Ras signaling to regulate proper development of these tissues. Furthermore, this work substantiates a physiological role for the previously reported binding of RIC to CaM (Wes et al., 1996). Interaction of these proteins inhibits RIC function, conferring sensitivity to intracellular calcium levels on this signaling pathway. The ability of activated RIC to alter Drosophila development, particularly in the eye, will permit a more extensive genetic investigation of this interesting signaling molecule. We anticipate that these genetic tools will be invaluable in the dissection of the physiology and characterization of the signaling pathways regulated by this unique subfamily of small GTPases.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Mammalian Cell Culture

Mutant or wild-type RIC was subcloned into the pEGFP-C1 vector to generate the GFP-tagged mutant or wild-type RIC mammalian expression vector. The constructs were verified by DNA sequence analysis.

PC6 is a subline of PC-12 cells that produces neurites in response to nerve growth factor but grows, as well, as isolated cells rather than in clumps (Spencer et al., 2002a, b; the generous gift of Thomas Vanaman, University of Kentucky, Lexington, KY). The cells were cultured in Dulbecco's modified Eagle's medium (Sigma) containing 10% (v/v) heat-inactivated horse serum, 5% (v/v) fetal bovine serum (Gibco), and 50 μg/ml gentamicin at 37°C in a humidified atmosphere of 5% CO2.

Neurite Outgrowth Studies

PC6 cells were seeded at 5 × 104 cells on dishes precoated with 25 μg/ml poly-L-lysine (Sigma) and transiently transfected using Effectene (Qiagen) with one of the following plasmids: pEGFP-C1, pEGFP-Ric, or pEGFP-RicQ117L. Protein expression was examined by epifluorescence microscopy for the GFP fusion proteins. Cells were scored positive for neurite outgrowth if one or more neurites exceeded two cell body diameters in length (Spencer et al., 2002a, b).

Drosophila Constructs and Transgenics

Plasmids for P-element mediated germline transformation were constructed by cloning either the original full-length Ric cDNA (Wes et al., 1996) or a clone bearing a site-directed mutation into the binary expression vector pUAST (Brand et al., 1994), creating pU-Ric, pU-RicG68V, and pU-RicQ117L. These transposon constructs were introduced into flies using standard methods (Spradling, 1986).

Fly Strains and Genetic Analysis

Transgenic flies with inducible expression of Ric alleles were crossed with a bank of GAL4 driver lines know to produce phenotypes in a variety of tissues (Harrison et al., 1995) and examined for lethality or visible phenotypes. For modifier interaction analyses, two balanced lines bearing phenotypes in wings (h-GAL4, U-RicG68V) or the eyes (ey-GAL4, U-RicQ117L) were generated by recombination. These balanced recombinants were crossed with candidate interacting mutations, also carried over a balancer. Candidate mutations were obtained from the Bloomington Stock Center. Relevant interacting mutations are described in FlyBase (http://flybase.net). Comparisons were made of the phenotypes from flies carrying the GAL-Ric chromosome and the balancer vs. the GAL-Ric chromosome and the candidate mutation. Rearing temperatures for crosses are noted in the text.

Wing Analysis

Quantitative analyses of ectopic wing vein was performed by removing the wings and mounting in Hoyer's mountant (Ashburner, 1989), as previously described (Harrison et al., 1995). Whole wings were digitally photographed (Spot Camera, Diagnostic Instruments) with a ×4 objective (Nikon E800), using the same optical conditions for all samples. For each wing image, the total area of ectopic wing vein and the total area of the wing were determined with Scion Image software by selecting the appropriate area by using a digitizing pad (Wacom). The ratio of ectopic wing vein area to total wing area was calculated for at least 10 wings of each genotype. The control wings for each mutant genotype were obtained from sibling animals in the same vial that carried the appropriate balancer chromosome rather than the mutation of interest. The proportion of ectopic wing vein for the mutant of interest was compared with the control using the Student's t-test to evaluate the significance of the observed differences.

Antibody Staining

Third-instar larvae were dissected in phosphate buffered saline (PBS), fixed in 4% methanol-free formaldehyde (Sigma-Aldrich) in PBS for 20 min, then incubated with antibodies as previously described (McGregor et al., 2002). Rat anti-ELAV (DSHB) was used at 1:25 and Cy2 anti-rat (Jackson ImmunoResearch) was used at 1:200. dpERK staining was performed by dissecting third-instar larvae in PBT (PBS plus 0.1% Tween20) and fixing for 30 min in 3.7% formaldehyde, then 15 min in 100% methanol. Fixed discs were incubated in rabbit anti-dpERK (Cell Signaling Technology) at 1:100, followed by biotinylated anti-rabbit (1:500). Antibody was detected by using the horseradish peroxidase (HRP) VectaStain Elite kit (Vector Labs). Eye imaginal discs were dissected, mounted, and imaged with a Leica TCS/NT laser scanning confocal microscope as previously described (McGregor et al., 2002) for fluorescently labeled discs or imaged on a Nikon E800 microscope and photographed with a Diagnostic Instruments Spot RT Slider camera for HRP labeled discs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank K. Beckingham, A. Michelson, and the Bloomington Stock Center for Drosophila strains. We also thank R. Barlett for the excellent technical assistance. D.A.H. was funded by the National Science Foundation, D.A.A. was funded by a Public Health Service grant from the National Institute of Neurological Diseases and Stroke and the Kentucky Lung Cancer Research Fund, and C.M. was funded by the National Eye Institute.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES