Technical Comments

Response to Comment on “Nervy Links Protein Kinase A to Plexin-Mediated Semaphorin Repulsion”

Science  22 Jul 2005:
Vol. 309, Issue 5734, pp. 558c
DOI: 10.1126/science.1109259

The comment by Ice et al. (1) questions conclusions we drew concerning the function of the Drosophila protein Nervy (2). They state that the polyclonal antibody anti-ETO Ab-1, directed against the mammalian ortholog of Nervy ETO/MTG8, does not recognize Drosophila Nervy. We purchased the polyclonal antibody in question in July 2002 from EMD Biosciences because it was generated against a region that is highly conserved between the human Nervy/MTG/ETO protein and Drosophila Nervy (including one portion where 19 out of 21 amino acids are identical) (2, 3). Our extensive analyses (2, 3) revealed that the ETO Ab-1 antibody specifically recognized a single prominent band of the size expected for Drosophila Nervy on blots of wild-type Drosophila embryo lysates. Furthermore, this band was absent in lysates derived from nervy mutant embryos [fig. S2D in (2); also see (3)]. More recently, we noted lot-to-lot variation in this reagent and reported this observation to EMD Biosciences in June 2003. We also reported the specific batch of the ETO antibody that we used in our original study (2, 3). In their comment, Ice et al. (1)report that they supplied these lots of ETO antibody to EMD Biosciences but rule out any possible lot-to-lot variation in these polyclonal antibodies. However, they did not detect Nervy in Drosophila embryonic lysates [see figure 1A, right panel, in (1)], whereas we detected a single prominent band that was absent in lysates from nervy mutant embryos (2). Therefore, the assertion of Ice et al. that the band we observed was a background band serendipitously reduced in the nervy mutant is not consistent with their observations, because they did not see prominent band using their lots of anti-ETO Ab-1 on wild-type Drosophila embryo lysates. The faint bands Ice et al. observed with their lots of ETO antibody upon long exposure [figure S1 in (1)] are only evident in Cos-7 cell extracts, not in Drosophila lysates [figure 1A in (1), right panel]. Apparent lot-to-lot differences in the anti-ETO used by Ice et al. can be seen by comparing bands in figure S1, A and B, in (1). We also performed multiple, carefully controlled coimmunoprecipitation (co-IP) studies using our purchased lot of anti-ETO. We observed that Nervy associates with an epitope-tagged neuronal cell surface receptor, Plexin A (PlexA), and with protein kinase A (PKA) but not with the neuronal protein enabled; Nervy did not coimmunoprecipitate in the absence of epitope-tagged PlexA [figures 1 and S2 in (2); (3)]. Therefore, it is extremely unlikely that a single serendipitously detected protein resulting from a background association with the anti-ETO would (i) be observed only at a size identical to Nervy, (ii) be able to selectively coimmunoprecipitate PlexA and PKA, (iii) be selectively recognized as a band of the same size as Nervy from a PlexA co-IP experiment, and (iv) be absent from nervy mutant embryos.

Ice et al. (1) also question whether Nervy/MTG proteins are expressed in the cytoplasm and in axons. Although Nervy/MTG proteins are best known for their nuclear localization, other research groups report that mammalian Nervy/MTG proteins are indeed present in the cytosol (410). The nuclear or cytosolic localization appears to be dependent on both cell type and developmental stage (410). Of particular importance to Nervy/MTG protein localization is that MTG proteins are also reported to be present in axons and at synapses (5, 6). We found that the nervy transcript is also present in axons. Antisense probes for the nervy transcript strongly label neurons and their processes, including motor axons, and this labeling is absent when using a nervy sense-control or the antisense nervy probe in nervy mutant embyros [Fig. 1 and figure 1C, C1, and C2 in (2)]. Immunohistochemical staining of wild-type embryos using our purchased anti-ETO (2, 3) also identified central nervous system neurons and their axonal processes [Fig. 2, C and E; figure 1D in (2); (3)]. This axonal staining was absent in nervy mutant embryos [Fig. 2, B, D, and F; see also (2, 3)]. The inability of Ice et al. (1) to observe immunostaining of neuronal processes with their antibody directed against Drosophila Nervy could reflect differences in immunostaining protocols (11) or differences among the epitopes recognized by their anti-Nvy polyclonal antibody and the ETO antibody used in our study (2). It does not, however, bear on the accuracy of our carefully controlled immunostaining results with a different immunological reagent.

Fig. 1.

Nervy (Nvy) antisense probes reveal Nvy transcript in axons of wild-type embryos but not Nvy mutant embryos. Filleted preparations of stage 16 Drosophila embryos (anterior, left). (A) In situ hybridization with antisense Nvy shows Nvy transcript within neurons and their axons (arrows), some of which can be seen here emanating from a neuronal cell body (ventral unpaired median neuron, white arrow). (B) In situ hybridization with antisense Nvy in a Nvy [PDFKG1 (2)] mutant embryo shows no evidence of Nvy in neuronal cell bodies or axons. (C) In situ hybridization with sense Nvy probe shows no specific staining of neurons or axons. Methods as in (2, 3).

Fig. 2.

Immunohistochemistry with ETO/Nvy antisera shows that in wild-type, but not in the Nvy mutant, the Drosophila embryonic nervous system can be stained. Stage 16/17 embryos immunostained with anti-ETO/Nvy Ab-1 (Lot D16428-1, EMD Biosciences) in 2003 and reported in (2, 3) observed at low (A and B), intermediate (C and D), and high (E and F) power. The nervous system (arrowheads) immunostains in wild-type embryos [(A), (C), and (E)] but not in Nvy [PDFKG1 (2)] mutant embryos [(B), (D), and (F)]. Select immunostained axonal processes are indicated in (C) [arrows in the inset, emanating from a neuron (arrowhead)] and in (E) (arrows indicating multiple commissural axons). Axons were not immunostained in the Nvy mutant. Arrows in (F) indicate position of anterior and posterior commissures, visualized by Nomarski optics. Anterior is to the left. A, anterior commissure; P, posterior commissure; L, longitudinal connectives; asterisks, motor nerve roots. Methods as in (2, 3).

Our other observations (2) are also consistent with Nervy functioning in axons through its interactions with PlexA as a PKA (A kinase) anchoring protein (AKAP). Our largescale yeast interaction screen first identified Nervy as a specific interacting partner with the axon guidance receptor PlexA (2). Nervy mutant analyses and extensive genetic studies revealed no gross abnormalities indicative of major changes in gene expression but, instead, showed axon guidance phenotypes similar to mutations in PlexA signaling components (2). Nvy/MTG proteins have a consensus PKA binding motif and are cytoplasmic AKAPs (810). Antibodies directed against Nervy immunoprecipitate PKA, and an antibody directed against the Myc epitope coimmunoprecipitates PKA when Myc-Nvy is expressed in neurons (2). The consensus PKA binding motif in Nervy is critical for its role in axon guidance, and genetic analyses and co-IP experiments between PKA and PlexA also support a role for PKA in PlexA signaling (2). Our molecular, genetic, in situ hybridization, immunolocalization, and immunoprecipitation data, along with published work from others, support that Nervy can be found in the cytoplasm, interacts with the cytoplasmic domain of the plasma membrane receptor PlexA, and directly modulates PlexA signaling by functioning as an AKAP. Our results in no way exclude the possibility that Nervy serves important roles in neuronal nuclei, but they support a critical role for direct PlexA/Nervy interactions in directing axon guidance decisions.

Supporting Online Material

SOM Text


References and Notes

Related Content

Navigate This Article