SelR reverses Mical-mediated oxidation of actin to regulate F-actin dynamics.

Actin's polymerization properties are markedly altered by oxidation of its conserved Met 44 residue. Mediating this effect is a specific oxidation-reduction (redox) enzyme, Mical, that works with Semaphorin repulsive guidance cues and selectively oxidizes Met 44. We now find that this actin-regulatory process is reversible. Employing a genetic approach, we identified a specific methionine sulfoxide reductase (MsrB) enzyme SelR that opposes Mical redox activity and Semaphorin-Plexin repulsion to direct multiple actin-dependent cellular behaviours in vivo. SelR specifically catalyses the reduction of the R isomer of methionine sulfoxide (methionine-R-sulfoxide) to methionine, and we found that SelR directly reduced Mical-oxidized actin, restoring its normal polymerization properties. These results indicate that Mical oxidizes actin stereospecifically to generate actin Met-44-R-sulfoxide (actin(Met(R)O-44)), and also implicate the interconversion of specific Met/Met(R)O residues as a precise means to modulate protein function. Our results therefore uncover a specific reversible redox actin regulatory system that controls cell and developmental biology.

Identifying the factors that shape the actin cytoskeleton, the basic building blocks of cellular form and function, is a critical biomedical goal [1,2]. Interestingly, actin is susceptible to post-translational modification of its amino acid residues but the physiological importance of these covalent modifications is still poorly understood [3]. Recently, we found that actin's polymerization properties are altered by specific oxidation of its conserved methionine (Met)-44 residue on the pointed-end of actin subunits [4]. These observations raise issues of the susceptibility of this residue to pathological modification [3], but we have also identified a specific oxidation-reduction (Redox) enzyme, Mical, that selectively oxidizes Met-44 to disassemble actin filaments (F-actin) and impair actin polymerization [4,5]. Our results reveal that Mical uses F-actin as a direct substrate, employing an oxidation-dependent post-translational mechanism to regulate filament dynamics [4].

MICAL family proteins, which include one Drosophila Mical and three mammalian MICALs, regulate numerous cellular events in different tissues including morphology, motility, navigation, exocytosis, and survival (Reviewed in [6-9]). At least some of these effects occur through MICALs ability to regulate actin cytoskeletal organization [4,5,10-12]. Interestingly, 2 MICALs also directly link one of the largest families of extracellular guidance cues, the Semaphorins and their Plexin cell surface receptors, to changes in the actin cytoskeleton [5,13]. Semaphorins are the largest family of repulsive guidance cues [14,15] and have been characterized for their ability to disassemble F-actin and “collapse” the actin cytoskeleton of multiple different cell types [6,16]. MICALs directly bind to the Semaphorin receptor Plexin through their C-termini [13,17] and employ their actin-binding/regulatory Redox domain to mediate the destabilizing effects of Semaphorins/Plexins on the actin cytoskeleton [5]. These effects include a loss of F-actin, the decreased ability to polymerize new F-actin, a decrease in the number of F-actin bundles, and the regulation of F-actin-rich filopodia/branches [6].

We now find a specific methionine sulfoxide enzyme SelR/MsrB that selectively reverses this Mical-mediated oxidation of actin. SelR counteracts Mical in vivo to direct multiple actin-dependent cellular processes including axon guidance, synaptogenesis, muscle organization, and mechanosensory development. SelR also neutralizes Semaphorin/Plexin repulsion. Thus, Mical and SelR comprise a reversible Redox cellular signaling system that orchestrates proper cytoskeletal-mediated physiology.

RESULTS

SelR Counteracts Mical-mediated F-actin Alterations In Vivo

Mical directs the organization of actin in a number of different cell types [4,5,10-12] including within developing bristle processes, which are akin to mammalian mechanotransducing inner ear hair cells that detect sound [18,19]. Bristles have also long served as a simple, single cell model for characterizing actin dependent events in vivo [5,20,21]. Raising the levels of Mical specifically in bristle cells using the GAL4-UAS [22] system (bristle-specific GAL4/UAS:Mical) results in F-actin disassembly and bristle branching (compare ) that is dependent on Mical's Redox activity and the Met-44 residue of actin [4,5]. Thus, to better characterize Mical-mediated F-actin alterations, we have initiated a large-scale genetic screen to look for enhancers and suppressors of Mical-mediated bristle branching. One of the mutations that we identified in our genetic screen, the transposable element mutation EY22443, strongly suppressed Mical-induced actin-dependent bristle branching (). Molecular analysis revealed that the EY22443 transposable element mutation was situated within the Drosophila SelR gene (). SelR codes for a methionine sulfoxide reductase (MsrB) family enzyme, that has been characterized for its ability to reduce oxidized methionine residues [23]. In light of our observations that Mical oxidizes methionine residues on actin [4], we wondered if SelR might play a role in modulating Mical's effects on actin.

The EY22443 transposable element mutation situated in SelR contains a UAS promoter (), thereby suggesting that this mutation might be abnormally inducing SelR expression to suppress GAL4/UAS:Mical-dependent bristle branching. To test this hypothesis, we generated transgenic flies expressing SelR directly under the UAS promoter. Consistent with our results with EY22443 () and another UAS-containing mutation within SelR, EP3340 (), multiple transgenic lines revealed that raising the levels of SelR specifically in bristles strongly suppressed Mical-induced bristle branching and even generated normal appearing bristles (). Moreover, elevating the levels of SelR in a wild-type background generated abnormally bent bristles that resembled Mical−/− mutant bristles (; [5]); and these effects of SelR were genetically enhanced by decreasing the levels of Mical (). Further analysis revealed that SelR localized with Mical at the tips of bristles and suppressed Mical-mediated F-actin disassembly and reorganization (). Therefore, SelR counteracts the effects of Mical on actin reorganization in vivo.

SelR Restores the Polymerization of Mical-treated Actin

To better understand the role of SelR in counteracting Mical-mediated actin reorganization, we purified recombinant Drosophila SelR protein (). Using in vitro actin biochemical and imaging assays, we previously observed that purified Mical protein in the presence of its coenzyme NADPH disrupts actin polymerization and induces F-actin disassembly (; [4,5]). Strikingly, we found that purified SelR protein rescued the ability of Mical-treated actin to polymerize (). This Mical/SelR-treated actin re-polymerized to an extent that was indistinguishable from normal untreated actin (). Moreover, while Mical-treated actin failed to polymerize even after removal of Mical and NADPH (; [4]), SelR induced the polymerization of this purified Mical-treated actin in a dosage-dependent manner (). Thus, SelR restores the polymerization properties of Mical-treated actin.

SelR converts methionine sulfoxide (MetO) to methionine [23,24], requiring a redox active cysteine (Cys124) residue (; [25]) and also utilizing reducing agents to cycle back to its reduced form (; [24,25]). In some cases methionine oxidation is also reversed by general reducing agents [26], so we wondered if Mical-treated actin was specifically reversed by SelR. In contrast to SelR, neither chemical reducing agents such as DTT ( [buffer only contains DTT]; ) nor other reducing enzymes including thioredoxins/thioredoxin reductases altered Mical-mediated effects on actin in vitro ( or in vivo (). Furthermore, SelR did not restore the normal polymerization properties of other oxidized forms of actin (e.g., H2O2-treated actin; ), indicating that SelR selectively affects Mical- modified actin. Mutating SelR's critical catalytic cysteine (Cys124) to generate an enzymatically dead SelR (SelRC124S; ; [25]), abolished SelR's effects on Mical-treated actin in vitro () and in vivo (). Moreover, consistent with such a role for SelR's reductase activity in counteracting Mical's oxidative effects on actin, elevating the levels of wild-type SelR not only phenocopied the in vivo effects of disrupting Mical's monooxygenase (Redox) domain (), but it also rescued the severe bristle/F-actin alterations that result from hyperactive Mical Redox signaling (; MicalredoxCH; [5]). Thus, SelR specifically employs its catalytic activity to restore Mical-treated actin polymerization and counteract the in vivo effects of Mical.

SelR Reverses Mical-mediated ActinMet-44 Oxidation

In many organisms, including Drosophila and mammals, two main types of methionine sulfoxide reductases have been identified: SelR (MsrB family proteins) and Drosophila Eip71CD (MsrA) (; [27]). Interestingly, SelR and MsrA/Eip71CD are both methionine sulfoxide reductases, but they do not exhibit similarity in their sequence, domain organization, or substrate specificity ( [27]). In particular, methionine has a unique oxidation pattern in that two stereoisomers can be produced by oxidation [27]. SelR/MsrB family proteins catalyze the reduction of the R-isomer of methionine sulfoxide (methionine-R-sulfoxide; ) to methionine, while MsrA/Eip71CD catalyzes the reduction of the S-isomer of methionine sulfoxide (methionine-S-sulfoxide; ) to methionine [23,25,27]. Therefore, to further test the specificity of SelR in restoring the polymerization properties of Mical-treated actin, we purified recombinant MsrA/Eip71CD protein (; [25]). Unlike SelR, MsrA/Eip71CD did not restore the polymerization properties of Mical-treated actin in vitro (), nor did it counteract Mical-mediated actin reorganization/bristle branching in vivo (). These results further reveal that Mical-treated actin polymerization is specifically restored by SelR. Moreover, in light of the isomer-specific nature of the methionine sulfoxide enzymes SelR and MsrA/Eip71CD, these results also indicate that Mical oxidizes actin in a stereo-specific manner.

Mical oxidizes actin on its Met-44 and Met-47 residues, although it is the oxidation of the Met-44 residue through which Mical induces F-actin disassembly [4]. Thus, we wondered if SelR directly reverses Mical-mediated oxidation of actin. Previously, we determined the conditions to purify Mical-treated actin, which is polymerization impaired and exhibits a mass increase of two oxygens (32 Daltons) [4]. SelR, but not the enzymatically dead SelRC124S protein, restored the polymerization properties of purified Mical-treated actin (), an effect that was maintained even after removal of SelR (). Subjecting both purified Mical/SelR-treated and Mical/SelRC124S-treated actin to mass spectrometry revealed that SelR, but not the enzymatically dead SelRC124S protein, eliminated the Mical-catalyzed two oxygen (32 Dalton) mass increase on actin ().

Mical's ability to effect actin in vitro and in vivo is dependent on the presence of the Methionine (M) 44 residue of actin [4]. To further examine a physiological role for SelR in reducing Mical-mediated oxidation of Met-44, we turned to in vivo assays. We first noted that overexpression of either a non-Mical oxidizable Met44Leu (M44L) version of actin [4] or wild-type SelR generated the same effects: suppression of Mical-mediated actin/bristle morphology and Mical loss-of-function-like defects (; [4]). Furthermore, we found that actinM44L worked in combination with SelR to generate Mical loss-of-function-like bristle defects (). Moreover, actinM44L prevented the enhanced Mical-mediated bristle branching/actin reorganization that occured with expression of the reductase dead SelRC124S (). Thus, SelR reverses Mical-mediated oxidation of actin, including using its catalytic activity to directly reduce Mical-induced MetO-44 actin to Met-44 actin () – and these observations with purified proteins are supported by our in vivo genetic assays.

The Mical/SelR System Regulates Actin Organization in Multiple Cell Types

In addition to bristle cells, Mical regulates the organization of actin in multiple other cell types including mammalian cells in vitro and muscles in vivo [4-6,10,11]. Thus, we wondered if SelR could also counteract the effects of Mical on actin in these other cellular systems. Our initial examination revealed that as in bristle cells, SelR rescued Mical-dependent changes in morphology and actin organization in cultured cells (). Further examination revealed that overexpression of SelR in muscles in vivo phenocopied the muscle actin defects found in Mical mutants (; [10]). Moreover, SelR could even rescue the lethality and changes in actin organization associated with overexpression of Mical in muscles ( as well as the lethality that results when Mical is broadly expressed using an actin promoter ().

Drosophila SelR, like Mical, is broadly expressed (; [5,10,13,25,28-31]) and thus to better examine these Mical-SelR interactions and their physiological effects on actin, we characterized SelR mutants (). Strikingly, loss of SelR generated bristle and muscle defects that resembled overexpression of Mical (). Moreover, loss of SelR specifically enhanced Mical-mediated effects on actin organization/bristle morphology () and phenocopied overexpression of the SelRC124S reductase mutant protein (). Thus, SelR, like Mical, plays both important and selective roles in regulating actin organization in vivo in different cell types. Likewise, an equilibrium between Mical and SelR activities underlies normal actin-directed cell biology.

SelR Neutralizes Semaphorin/Plexin/Mical Repulsive Signaling

Besides its Redox region that Mical uses to oxidize actin, Mical has several other domains and protein interaction motifs including a region that interacts with the cytoplasmic portion of Plexin (; [13,17]). Plexins are receptors for Semaphorin guidance cues and play critical roles in regulating multiple actin-dependent events in vivo [6,32,33]. Semaphorins/Plexins signal through Mical to induce changes in bristle morphology and F-actin disassembly [5], so we wondered if SelR also counteracted the effects of Semaphorin/Plexin/Mical signaling. Employing loss and gain-of-function genetics in the bristle system, we found that similar to our results with Mical, SelR counteracted Semaphorin/Plexin effects on actin-dependent bristle morphology (). Next, we turned to in vivo axon guidance assays, where Semaphorins/Plexins have been characterized as repulsive axon guidance molecules [15] and were first linked to MICAL family proteins [13]. Interestingly, one of the SelR mutants that we found in our screen (EP3340, ) recently emerged from a genetic screen as an uncharacterized regulator of axon guidance [34]. Employing our SelR transgenic lines, we found that overexpression of SelR generated axon guidance and synaptogenic defects that phenocopy Mical mutants (; [5,10,13]). Furthermore, SelR mutants generated axon guidance defects that phenocopy increased Semaphorin/Plexin/Mical-mediated repulsive axon guidance (; [5,35-37]). Moreover, increasing the levels of SelR rescued these Semaphorin/Plexin/Mical-triggered repulsive axon guidance defects (). Thus, SelR also plays critical roles in axon guidance and synaptogenesis and counteracts the effects of Semaphorin/Plexin/Mical repulsive signaling in vivo.

DISCUSSION

Our results reveal that Mical-mediated actin alterations – a selective means to post-translationally regulate F-actin dynamics and cellular behaviors – are reversible. This Micalcatalyzed reaction is directly reversed by a specific methionine sulfoxide reductase enzyme, SelR/MsrB, which we also find selectively controls actin-dependent cellular events in vivo and regulates specific neuronal, muscular, and mechanosensory developmental processes. We also find that SelR counteracts Semaphorins, which are one of the largest families of extracellular guidance cues and play a critical role in the formation and function of multiple tissues [6,32]. Thus, our results demonstrate an important role for these methionine sulfoxide reductases – enzymes thought to function primarily in the repair of oxidatively “damaged” methionine residues [24,38] – in modulating normal signaling events. Moreover, our genetic data, which reveals that SelR and Mical loss and gain-of-function phenotypes are opposite in appearance, indicate that SelR has a specific, primary, and regulated role in counteracting Mical during development.

The Mical substrate Met-44 residue of actin is conserved in all actin family members from yeast to humans [4] and a dominant (hetereozygous) mutation in the Met-44 residue (M44T) of skeletal muscle actin underlies a human musculoskeletal disease associated with actin accumulation and aggregation (nemaline myopathy [39]). This Met-44 mutant version of human skeletal muscle actin would be predicted to prevent Mical from having effects on skeletal muscle actin – and generally phenocopies both Mical mutants and SelR muscle overexpression. However, the Met-44 residue is well-conserved and is at a subunit interface in filaments [4,40,41] and thus mutating it may influence F-actin organization for reasons other than that it is non oxidizable. It should be noted however, that our previous results indicate that Met-44 mutant actin (M44L) appears to polymerize normally in vitro and in vivo, but is resistant to Mical-mediated F-actin disassembly [4].

It is also interesting to note the differences in the cellular localization we see between SelR and different forms of Mical. For example, in bristles, SelR shows overlapping localization with Mical, but is more broadly distributed than full-length Mical, which strongly localizes to bristle tips (; [5]). The broader cellular localization of SelR is similar to that seen when the hyperactive MicalredoxCH is expressed in bristles and other cells (; [5]). One of the differences between full-length Mical and the hyperactive MicalredoxCH is the presence of the Plexin-interacting region (; [13]). Our results indicate that full-length Mical is susceptible to regulation by Plexin, whereas the MicalredoxCH protein (which does not have the Plexin-interacting region) is not regulated by Plexin [5] (see also [11,17]). Interestingly, the MICALs express multiple different transcripts, including versions that may be similar to MicalredoxCH 6. Thus, there may be roles for both endogenous Sema/Plexin-regulated and perhaps, non Sema/Plexin-regulated forms of Mical (which appear to be more generally localized in cells). In any case, it should be noted that we find that SelR rescues both the lethality and F-actin defects associated with overexpression of either full-length Mical or MicalredoxCH. Likewise, we find that SelR counteracts Semaphorin/Plexin effects in vivo.

Our results herein, coupled with our previous observations [4], also indicate that unlike diffusible oxidants that induce random protein modifications [24,38,42], Mical-mediated oxidation is substrate, residue, and stereo-specific. Our results indicate that Mical oxidizes the methioinine-44 residue of actin stereospecifically to generate actin methionine-44-R-sulfoxide (actinMet() to alter F-actin dynamics. These observations contend that the enzyme-driven interconversion of specific Met/Met(R)O residues, similar to the reversible phosphorylation of specific serine, threonine, and tyrosine residues [43], provides a selective means to precisely modulate protein function. Moreover, in contrast to a view that oxidation simply plays a destructive role in cell health and protein function, our results indicate that the site specific and reversible oxidation of proteins is critical for proper cellular physiology. Thus, together, our results uncover a specific reversible Redox cellular signaling system that dynamically regulates multiple actin cytoskeletal-mediated events and controls Semaphorin/Plexin repulsion.