 |
|||||
|
|||||
|
|||||
|
|||||
|
|||||
Electronic Letters to:
- DevelopmentPlasticityRepair:
Florence K. Roche, Bonnie M. Marsick, and Paul C. Letourneau- Protein Synthesis in Distal Axons Is Not Required for Growth Cone Responses to Guidance Cues
J. Neurosci. 2009; 29: 638-652[Abstract] [Full text] [PDF]
eLetters: Submit a response to this article
Electronic letters published:
Turning data not supporting the claim
- James Q Zheng (18 February 2009)
- Paul C Letourneau (5 March 2009)
- Samie Jaffrey (10 March 2009)
- Paul C Letourneau (31 March 2009)
Turning data not supporting the claim 18 February 2009 
James Q Zheng,
Professor
Atlanta, GASend letter to journal:
Re: Turning data not supporting the claim
james.zheng{at}emory.edu James Q Zheng
In this paper, Florence et al. claimed that protein synthesis in distal axons is not required for growth cone responses to guidance cues. This conclusion would be significant if supported by convincing data, since it contradicts previous findings from several labs on the involvement of protein synthesis and degradation in growth cone guidance. While I can not comment on the collapsing results, I did take a close examination on the turning data. As depicted in the annotated figure below, I used the turning data presented in original Figure 6 (the only one dealing with growth cone turning responses) for analysis. As one can see, the turning data might have been mis-analyzed since there is no turning at all for the growth cones indicated. This is demonstrated by placing two parallel lines that are drawn according to the trajectory of the growth cone and its adjacent 20 um neurite at the beginning of the assay. The growth cone named #3 actually turned away initially but return to a position that gives rise to no turning. The conclusion that none of the growth cones shown in the figure 6 exhibited any turning response is also confirmed by overlaying the 2 min and 30 min images.
Assuming that the figure 6 represents the best images of the turning data, together with the rather small turning angles (NGF: 12 ± 2.5° without inhibitors, 5.4 ± 2.1° with inhibitors; BSA control: –1.7 ± 2.2°), I do not believe that the authors’ claim (at least for the turning responses) is supported by their data. On the contrary, the fact that protein synthesis inhibition significantly reduced the turning response to NGF pipette supports an important role of local protein synthesis in turning responses.
Response to turning data not supporting the claim 5 March 2009 
Paul C Letourneau,
Professor
University of Minnesota 55455Send letter to journal:
Re: Response to turning data not supporting the claim
letou001{at}umn.edu Paul C Letourneau
These images presented in Figure 6 do not show the largest turning angles that we observed. Our article does cite previously published methods that we used to measure growth cone turning angles. As we reported, the global presence of cycloheximide, starting 15 min before introducing the NGF pipette, did reduce the turning angle, but a positive response still occurred. However, because newly synthesized proteins can be rapidly transported from the perikaryon to the growth cone of a short axon, the effects of globally added protein synthesis inhibitors on turning cannot argue strongly for a requirement for distal axon protein synthesis in growth cone responses to guidance cues. Thus, we used compartmented dishes where neuronal perikarya and distal axons can be exposed to different media. This approach to more directly ask if distal protein synthesis is required to respond to guidance cues cannot be used when in vitro axonal growth is limited, such as Xenopus neurons.
Collapse data not supporting the claim 10 March 2009 Samie Jaffrey,
Associate Professor, Department of Pharmacology
Weill Medical College, Cornell University, New York, NY 10065Send letter to journal:
Re: Collapse data not supporting the claim
srj2003{at}med.cornell.edu Samie Jaffrey
Numerous laboratories have found that certain guidance cues exert protein synthesis-dependent effects on axons. Roche et al. (2009) seemingly contradict these studies; however, experimental flaws make the relevance of their results questionable. Here, we focus on the collapse assays. The turning assays are discussed elsewhere (Zheng, 2009).
The semaphorin 3A (Sema3A)-induced growth cone collapse assay results in prominent axonal shriveling and toxicity, which is evident within 2 min (see Fig. 8b). The collapse rate and cytotoxicity do not accord with most previous studies. The cytotoxicity renders these experiments uninterpretable.
It is puzzling that Roche et al. examine the requirement for axonal protein synthesis using an intense Sema3A stimulation paradigm. It has already been shown that Sema3A-induced growth cone collapse is dependent on protein synthesis only at low Sema3A stimulation levels (Li et al., 2004). The authors may not be aware of this critical information, and concentrations where protein-synthesis dependent collapse is expected were not tested. Indeed, we identified a minimal concentration that elicited collapse over one hour, allowing us to detect a requirement for axonal protein synthesis (Wu et al., 2005). In addition, Sema3A needs to be titered to account for batch-to-batch and culture condition variations, since NGF in the culture media affects the effectiveness of Sema3A (Dontchev and Letourneau, 2002). The physiological relevance of rapid exposure to large amounts of Sema3A is unclear. Furthermore, while hyperstimulation by Sema3A can elicit effects by maximally activating limiting amounts of effector proteins, lower levels of stimulation permit the detection of pathways that are sensitive to limiting amounts of signaling proteins. This scenario is particularly relevant when examining intra-axonal protein translation.
Also problematic are the putative measurements of RhoA in growth cones. The authors measure RhoA in the distal 25 µm of axons after collapse. However, after collapse, this measurement reflects RhoA in the axon, not the growth cone. We used a scenario in which growth cone collapse was delayed to allow us to measure RhoA levels in growth cones prior to collapse (Wu et al., 2005). Along with the absence of normalization and staining specificity controls (e.g. RhoA knockdown), this flaw invalidates their claim to be measuring RhoA in growth cones.
The discussion in Roche et al. is misleading:
-- Roche et al. state that previous investigators relied on transected axons to monitor the effects of axonal protein synthesis inhibition, potentially introducing artifacts, while they used compartmentalized culturing chambers. However, local protein synthesis dependence was previously demonstrated in uninjured axons using compartmentalized culturing devices (Hengst et al., 2006).
-- Roche et al. suggest that a reason for the difference in results may be due to their use of mouse DRGs while others used Xenopus and rat. However, mouse DRGs were previously shown to exhibit protein-synthesis dependent Sema3A-induced collapse (Li et al., 2004).
-- Roche et al. state that Wu et al. (2005) found negligible levels of RhoA in growth cones prior to Sema3A stimulation. In fact, RhoA levels were approximately half the levels in unstimulated axons compared to stimulated (see Fig. 3f in Wu et al.).
In summary, significant flaws call into question the conclusion that protein synthesis is not required for growth cone responses to guidance cues.
REFERENCES
Dontchev VD, Letourneau PC (2002) Nerve growth factor and semaphorin 3A signaling pathways interact in regulating sensory neuronal growth cone motility. J Neurosci 22:6659-6669.
Hengst U, Cox LJ, Macosko EZ, Jaffrey SR (2006) Functional and selective RNA interference in developing axons and growth cones. J Neurosci 26:5727-5732.
Li C, Sasaki Y, Takei K, Yamamoto H, Shouji M, Sugiyama Y, Kawakami T, Nakamura F, Yagi T, Ohshima T, Goshima Y (2004) Correlation between semaphorin3A-induced facilitation of axonal transport and local activation of a translation initiation factor eukaryotic translation initiation factor 4E. J Neurosci 24:6161-6170.
Wu KY, Hengst U, Cox LJ, Macosko EZ, Jeromin A, Urquhart ER, Jaffrey SR (2005) Local translation of RhoA regulates growth cone collapse. Nature 436:1020-1024.
Zheng JQ (2009) Turning data not supporting the claim. J Neurosci online, see: http://www.jneurosci.org/cgi/eletters/29/23/638.
Re: Collapse data not supporting the claim 31 March 2009 Paul C Letourneau,
Professor
University of Minnesota, Minneapolis MN 55455Send letter to journal:
Re: Re: Collapse data not supporting the claim
letou001{at}umn.edu Paul C Letourneau
Dr. Jaffrey stated that flaws and misleading statements have made our paper (Roche et al, J Neurosci 29:6380) irrelevant. We should respond.
Dr. Jaffrey assumed that our collapse assays are cytotoxic and uninterpretable. This is unlikely, as we used a similar concentration of Sema3A as Dr. Jaffrey (500 ng/ml vs. 450 ng/ml), and images of Sema3A- treated collapsed growth cones in our Figures 4, 5 and 8 are similar to those published by Dr. Jaffrey in his paper (Nature 436:1020) and in countless other papers that depict growth cone collapse. Rapid collapse seen in our Figure 8b reflects the low attachment of growth cones to a laminin-coated coverslip not treated with polylysine.
Dr. Jaffrey cites Li et al (J Neurosci 24:6161) and suggests we test the protein synthesis-dependence of growth cone responses to low Sema3A stimulation. We recently did this using similar Sema3A concentrations as Li et al and found no effects of cycloheximide on the reduced collapse response to lower Sema3A concentrations. Dr. Jaffrey’s point about studying the effects of low levels of stimulation is valid, and we look forward to him measuring axonal RhoA synthesis in response to 10 ng/ml Sema3A (as in Li et al) instead of 450 ng/ml, as in his Nature paper.
We regret not citing Dr. Jaffrey’s elegant paper (Hengst et al, 2006), which used a compartmented dish to nicely show that axonal RhoA is required for a growth cone collapse response to Sema3A. We did immunocytochemistry for RhoA in compartmented dishes in which DRG distal axons were in cycloheximide-medium for 24 hr. These growth cones contained the same levels of RhoA staining intensity as control axons not exposed to cycloheximide. This is consistent with the collapse response of these growth cones to Sema3A.
Dr. Jaffrey does not believe our measurements of RhoA in DRG growth cones. We used the same antibody from Santa Cruz Biotechnology as he used in his Nature paper, and our paper clearly described precautions we took in image acquisition and analysis. As described in our paper, we cultured these neurons on an L1CAM substrate to avoid growth cone and axon retraction, allowing us to accurately measure RhoA levels in the terminal 25 µm axons (which includes growth cones) of control and Sema3A-treated axons.
The images of RhoA staining we presented in Figure 5k-n are actual images from which we made measurements of RhoA staining intensity. Hopefully, this is not the case for the images in Dr. Jaffrey’s Nature paper (Figure 3a-e). The pixel intensity levels of these stained growth cones are highly saturated and unsuitable for quantification.
Unlike Drs. Jaffrey and Zheng, we did not explain differences between our results and theirs by characterizing their papers as flawed, unconvincing or misleading. We consider their results to be valid and suggested the differences are due to metabolic differences arising from the different neurons and culture conditions that are used in different labs. We stated that in vivo studies are needed, and until that happens, we suggest that our cultures of neurons removed at the stage when their axons are actively pathfinding and maintained in vitro for 24 hr or less might reflect in vivo properties at least as well as studies involving neurons cultured for 3 days or more before being used, as in Dr. Jaffrey’s Nature paper, or that use neurons from adult animals, as in Li et al.
Paul Letourneau, Bonnie Marsick and Florence Roche
|
|
|
|
 |
|