MEGF8 is a modifier of BMP signaling in trigeminal sensory neurons

1) This paper provides a mechanistic explanation for the defasciculation defect observed in Megf8 mutants, but does not explain the foreshortening of many nerves, including their focus in this paper: the ophthalmic branch (Figure 4H)? If removing the response to a repellent, wouldn't one predict that the axons might grow longer, as seen in Figure 6?

There are two possibilities for the foreshortening observed in the ophthalmic branch of the trigeminal nerve. One possibility is that the foreshortening is secondary to the defasciculation. In other words, individual ophthalmic axons grow to the same length (or longer) in Megf8 mutants, but because the axons aberrantly extend into surrounding tissues the overall path of the nerve appears shorter. The other possibility is that in addition to its role in mediating BMP4 inhibitory cues, Megf8 also facilitates the growth of ophthalmic axons into the periphery and thus these axons are undergrown in Megf8 mutants. This second possibility is consistent with the phenotype observed in the spinal nerves, which are undergrown in Megf8 null embryos when compared to controls (Figure 4A,D). In addition, the maxillary branch of the trigeminal nerve is slightly undergrown in Megf8 mutants compared to control (Figure 4 legend). Taken together, these results suggest a role for Megf8 in promoting axon growth of TG and DRG sensory neurons. However, our data do not determine whether this occurs through modification of BMP4 signaling.

2) The phenotype observed in Figure 5B is very interesting, suggesting that Megf8 is required to stabilize or maintain Bmp4 expression. Is this phenotype also fully penetrant (add to Table 1?)? Could the authors provide some quantification of Megf8 mutants to more clearly assess the extent to which Bmp4 activity staining is lost in the region surrounding the ophthalmic nerve compared to controls? Importantly, the authors barely discuss this very striking result in the Discussion! Doesn't it suggest that the loss of Megf8 has a non-autonomous negative effect on Bmp4 expression? If so, how mechanistically do they think a membrane bound receptor is suppressing Bmp4 expression?

The disruption of Bmp4-lacZ expression around the TG ophthalmic branch is fully penetrant in Megf8^-/- embryos, and this point is now made in the Table in the revised paper. Interestingly, Bmp4-lacZ expression is normal in all other parts of the embryo. This can be appreciated in Figure 5–figure supplement 1, which shows the Bmp4-lacZ expression pattern throughout Megf8^-/- and Megf8^+/+ embryos and includes images of the whole embryo, DRG, and limb.

We have provided commentary on the changes in the Bmp4-lacZ expression pattern in the Megf8^-/- embryo and the implications of these changes in the revised manuscript. Our findings suggest that the loss of BMP4 expression at the site of defasciculation of the TG ophthalmic branch could result from two possible mechanisms. Either BMP4 expression is dependent on signaling from incoming TG axons (which is disrupted by defasciculation in the Megf8^-/- embryo) or Megf8 expressed in the craniofacial target region regulates BMP4 expression. This is an interesting point, as we know from our ISH results that Megf8 is expressed in TG neurons as well as in the developing craniofacial area. Because neural crest cells contribute to this developing craniofacial area, the Wnt1-Cre; Megf8^f/f embryo unfortunately does not allow us to distinguish between these two possibilities.

3) Is the interaction between Megf8 and BMP4 the likely mechanistic explanation for the other phenotypes? In which case, why is Bmp4 expression not lost in other affected organs (Figure 5)?

Indeed, given the phenotypic similarities between Megf8 mutant and BMP4 loss of function lines, we hypothesize that the functional interaction between Megf8 and BMP4 is the mechanistic explanation for the other phenotypes observed in Megf8 mutants, including polydactyly, disruption of left-right patterning, and heart defects. As we have shown, BMP4 expression in Megf8 null embryos is only disrupted around the TG ophthalmic branch; it is intact in all other areas of the embryo including the DRG, heart, and limbs. These findings suggest that Megf8 modifies BMP4 signaling by acting on the BMP4 ligand/receptor complex or downstream components of the signaling pathway, not through modifying expression of BMP4 itself. Indeed, this is consistent with our in vitro findings (Figure 6), in which we found that in the presence of exogenous BMP4 ligand, Megf8 null axons respond poorly to BMP4 as an inhibitory cue. If Megf8’s role in BMP4 signaling is primarily to modify BMP4 expression, one would expect that Megf8 null axons would still be inhibited by exogenous BMP4 ligand, unlike our results in Figure 6. In fact, we were surprised to find that BMP4 expression is disrupted around the TG ophthalmic nerve. We suggest that this finding (Figure 5) in conjunction with our in vitro analysis (Figure 6) indicates that Megf8 modifies BMP4 signaling in TG ophthalmic axons in two ways: 1) Megf8 expressed in the axon is required to mediate the BMP4 inhibitory cue and 2) loss of Meg8 in the axon, loss of Megf8 in the surrounding craniofacial tissue, or loss of normal TG axon extension (secondary to loss of Megf8) disrupts BMP4 expression. This is discussed in the revised manuscript.

4) The authors are careful to assay only the response of the ophthalmic branch of the TG nerve to BMP4 stimulation in vitro. Did they try a similar in vitro assay with the maxillary and mandibular branches? It would be interesting to know whether these branches are impervious to BMP4 in vitro as suggested by the results in vivo.

Our initial explant experiments included the entire ganglion and we did see robust inhibition of axon outgrowth by BMP4 (as well as loss of this inhibition in the Megf8^-/- explants). However, we decided to focus on the dorsal half of the TG, which contains the ophthalmic lobe and part of the maxillary lobe, so as to enrich for ophthalmic neurons. Given the large amount of maxillary neurons present (roughly 50% of the explant), our results suggest that BMP4 can inhibit maxillary axon outgrowth, although it is unclear if BMP4 inhibits maxillary neurons to the same extent as ophthalmic neurons. We did not attempt a separate analysis of the maxillary and mandibular branches.

5) Similarly, presumably other classes of neurons are affected, as there are also abnormalities of spinal roots although the spinal root abnormalities were not analyzed in detail.

Megf8 null embryos show undergrowth of DRG spinal nerves as well as defasciculation of the vagus and glossopharyngeal nerves. In addition, Megf8 is broadly expressed throughout the developing neuroepithelium (Figure 2), which suggests that it may play a role in different neuronal populations in the CNS and PNS. However, we have not yet asked whether BMP signaling is disrupted in other classes of neurons. In future experiments it will be important to assess the role of Megf8 in other axon guidance events controlled by BMP signaling, including the ventral axonal trajectory of neurons whose cell bodies reside within the dorsal spinal cord.

6) Although the link to BMP signaling is intriguing, the current evidence is rather indirect and based primarily on their one in vitro experiment. A simple way to provide strong evidence for a direct effect on BMP signaling in vivo would be to assess Phospho-Smad staining levels in mutant versus wild type trigeminal ganglia. The authors should attempt this experiment.

Phospho-SMAD immunohistochemistry on wild type and Megf8 null TG was attempted but, unfortunately, we encountered a widespread pattern of very strong staining in virtually all cell types and tissues in the preparations, and very high background. Thus, we were unable to ascertain whether phospho-SMAD expression is intact or diminished in Megf8 null embryos (which would imply that Megf8 modifies BMP4 signaling via a non-SMAD dependent mechanism), or if we could not observe a difference that was present because of issues of specificity or because this assay would not allow us to detect subtle changes in phospho-SMAD levels. We are therefore not confident that P-SMAD staining provides a specific and reliable quantitative assay for comparing Megf8^-/- TG to wild type littermates and have not pursued this direction.

7) In the Discussion, it is critical that the authors add further comment on how they think Megf8 regulates BMP4 signaling.

We have included discussion of mechanisms by which Megf8 regulates BMP signaling in the revised manuscript.