Technical Comments

Comment on "Nervy Links Protein Kinase A to Plexin-Mediated Semaphorin Repulsion"

See allHide authors and affiliations

Science  22 Jul 2005:
Vol. 309, Issue 5734, pp. 558
DOI: 10.1126/science.1108719

Terman and Kolodkin (1) recently suggested that Nervy, a Drosophila homolog of the mammalian myeloid translocation gene (MTG) family, functions at the plasma membrane as a protein kinase (PKA) anchoring protein (AKAP). Their findings were surprising because mammalian MTGs localize to nuclei and function as transcriptional corepressors by recruiting histone deacetylases (24). The antiserum Terman and Kolodkin used to detect endogenous Drosophila Nervy in immunoblot analysis and to localize Nervy to the plasma membrane of neurons by immunostaining (1) was not raised against Nervy but was an antibody to ETO (MTG8) developed in the Heibert laboratory against the C terminus of ETO/MTG8 (marketed through EMD Biosciences as anti-ETO Ab-1). Although the human ETO/MTG family members are well conserved in the C-terminal region, this antibody modestly recognized MTG16 and only weakly recognized the third MTG family member, Mtgr1, which is 61% identical in the antigenic region (Fig. 1A).

Fig. 1.

Anti-ETO Ab-1 does not recognize Nervy. Gal4-tagged forms of ETO/MTG8, MTG16, and Mtgr1 (the mammalian homologs of Nervy), a dual-epitope-tagged (Gal4 and Myc) form of Nervy (Gal4-Nvy), and Myc-tagged Nervy (Myc-Nvy) were expressed from the CMV immediate early promoter in Cos-7 cells. Equal amounts of protein were analyzed by SDS–polyacrylamide gel electrophoresis and immunoblotting with (A) anti-ETO Ab-1, (B) anti-Gal4, and (C) anti-Myc (monoclonal antibody 9E10). The right-hand panel of (A) shows an immunoblot analysis of whole-cell lysates from human erythroleukemia (HEL) cells and Drosophila embryo extracts probed with anti-ETO Ab-1. Although endogenous ETO/MTG8 was detected, endogenous Nervy was not detected. In (D), the same extracts used in (A) to (C) were rerun and probed with anti-Nvy, which detected Gal4-Nvy and Myc-Nvy, but not the mammalian family members. See the lower panel of fig. S1 for the matching anti-ETO Ab-1 blot.

To determine whether anti-ETO Ab-1 recognizes the less homologous Drosophila Nervy (which is only 29% homologous in this region), we obtained a Myc-epitope tagged form of nervy (Myc-nvy) from Terman and Kolodkin. DNA sequence analysis confirmed the identity of the cDNA. We also created a second form of Nervy containing both the Myc tag and the GAL4 DNA binding domain (Gal4-nvy). Using the cytomegalovirus (CMV) immediate early promoter, Gal4-Nvy and Myc-Nvy were expressed at very high levels in Cos-7 cells. In numerous experiments with several antibody concentrations, anti-ETO Ab-1 failed to detect either form of Nervy, even when expressed at very high levels and under conditions in which the ETO/MTG8 signal was rapidly overexposed (Fig. 1, A and B, and fig. S1). In contrast, antibodies directed to the epitope tags or to Nervy confirmed robust expression of GAL-Nvy and Myc-Nvy (Fig. 1, C and D). Although anti-ETO Ab-1 was obtained from a single rabbit, two bleeds of antiserum were supplied to EMD Biosciences. We tested both bleeds and found that neither recognized Nervy. Upon long exposures, however, several background bands were observed, with one migrating in the 75-kD range (fig. S1). Thus, anti-ETO Ab-1 did not recognize Nervy in immunoblot experiments, which suggests that the band observed previously (1) was a background band that was serendipitously reduced in the nervy mutant.

We also compared immunolocalization results in Drosophila embryos obtained with antibodies raised against Drosophila Nervy and anti-ETO Ab-1 (Fig. 2). In wild-type embryos, anti-ETO Ab-1 diffusely labeled the longitudinal axon tracts (Fig. 2A, arrow) and several scattered cells (Fig. 2A, arrowhead), a pattern similar to that previously reported (1). However, an identical pattern was observed in nervy loss-of-function mutant embryos [nervyPDFKG1 (1)] (Fig. 2B), indicating that it is nonspecific staining. In contrast, an extensively characterized anti-Nvy (Fig. 1D and fig. S2) labeled the nuclei of a subset of central nervous system (CNS) and peripheral nervous system (PNS) cells (Fig. 2, C to E). The pattern of Nervy protein matched the nervy mRNA pattern (5) and, in contrast to anti-ETO, anti-Nvy produced no detectable staining in nervy mutant embryos (Fig. 2D). Further, in contrast to the previous study (1), we did not observe any maternal nervy expression by in situ hybridization (5) or by immunostaining (Fig. 2E). Anti-Nvy also failed to label axons, which were labeled by anti-Fasciclin II (FasII) (fig. S2E). Although these results do not exclude the possibility that there is some Nervy in axons, they demonstrate that Nervy is predominantly, and perhaps exclusively, a nuclear protein and that anti-ETO could not be used to detect Nervy in Drosophila embryos (Fig. 2).

Fig. 2.

A comparison of Drosophila embryos stained with anti-ETO Ab-1 and anti-Nvy (9). All panels show ventral views of fluorescently stained Drosophila embryos; anterior is to the left. Left-hand panels show entire embryos; middle and right-hand panels show higher magnification views of the CNS. All images are confocal micrographs. (A) nvyPDFKG1/Cyo-GFP or (B) nvyPDFKG1/nvyPDFKG1 stage 16 embryos stained with anti-ETO Ab-1 (green) and anti-Elav (blue). Anti-ETO Ab-1 diffusely labeled the longitudinal axon tracts (arrow) and several scattered nonneuronal cells (arrowhead) in both nvy+ (A) and nvy mutant (B) embryos. Anti-Elav labeled CNS nuclei in both embryos. The right-hand panels show only the anti-ETO channel. (C) nvyPDFKG1/Cyo-GFP or (D) nvyPDFKG1/nvyPDFKG1 stage 16 embryos stained with anti-Nvy (green) and anti-Elav (blue). In nvy+ embryos (C), anti-Nvy labeled a subset of nuclei in the CNS and PNS. In nvy mutant embryos (D), no staining was observed with anti-Nvy. The right-hand panels show only the anti-Nvy channel. (E) Wild-type (Oregon-R) stage 14 embryo stained with anti-Nvy (green), anti-Elav (blue), and anti-FasII (red). Anti-FasII labeled axons and anti-Nvy labeled nuclei; no colocalization was observed.

Although ETO/MTG8 and MTG16 have also been suggested to be AKAPs (68), these studies found the proteins in the Golgi, not the plasma membrane. The Golgi localization is surprising because MTGs lack membrane translocation signal sequences. The antisera used in some of these experiments (6, 8) detected several proteins approximately 40 kD larger than the predicted or observed molecular weights of MTG8 or MTG16 (3), which raises the possibility that the antisera may cross-react with a Golgi protein or that these proteins correspond to novel isoforms of MTG family members.

In summary, the results described here are consistent with previous findings showing that members of the MTG protein family function in the nucleus as transcriptional corepressors (3, 4). Although we cannot rule out a cytoplasmic function for Nervy, we suggest that the axonal migration phenotypes observed in nervy mutant Drosophila embryos (1) may be due to alterations in gene expression rather than a failure to anchor PKA to the plasma membrane.

Supporting Online Material

Materials and Methods


References and Notes

Navigate This Article