This interesting manuscript attempts to delineate a relationship between the ubiquitin system and the axonal degeneration of cultured neurons. Strongly presented data suggests that inhibition of proteases results in the marked slowing of this degeneration. This reviewer shows concern for the assertion that the proteasome has been adequately assayed, however. While a rational candidate for being the responsible party to degrade neuronal components(and the answer is probably right, anyway) the methods used to assert this cannot produce that conclusion. At the heart of this is the specificity of the inhibitors used and their relationship to the proteasome. Figure 1 title is boldly stated. Of the inhibitors used, MG132 is the most directed to the proteasome active site, but even it is not highly specific. The support of a calcium chelator and a tripeptide must be discounted almost summarily. Why was the Lactacystin result not shown in Figure 1? Lactacystin is more specific to the proteasome than the other three. A better title is something that asserts protease inhibitors slow...  Read more

This interesting manuscript attempts to delineate a relationship between the ubiquitin system and the axonal degeneration of cultured neurons. Strongly presented data suggests that inhibition of proteases results in the marked slowing of this degeneration. This reviewer shows concern for the assertion that the proteasome has been adequately assayed, however. While a rational candidate for being the responsible party to degrade neuronal components(and the answer is probably right, anyway) the methods used to assert this cannot produce that conclusion. At the heart of this is the specificity of the inhibitors used and their relationship to the proteasome. Figure 1 title is boldly stated. Of the inhibitors used, MG132 is the most directed to the proteasome active site, but even it is not highly specific. The support of a calcium chelator and a tripeptide must be discounted almost summarily. Why was the Lactacystin result not shown in Figure 1? Lactacystin is more specific to the proteasome than the other three. A better title is something that asserts protease inhibitors slow axonal degradation. There are perhaps a hundred targets of these inhibitors, and the proteasome may be the least effected. Attempts to support this prior result with a UBP2 transfection is also suspect. UBP2 overexpression alone cannot reverse all Ubiquitylation in a cell. There are 17 UBPs in yeast, and many more in human, with such irritatingly narrow specificity that mapping out anything useful about their targets has proved problematic. The hydrolytic specificity of UBP2 is almost completely unknown, and the understanding of the specificity of hydrolysis of ubiquitylated proteins (not peptides) is nearly absent in the literature. Why would a single overexpression of a single, large UBP with unknown target have any definable relationship to this process? The only mentioned ubiquitin protease with specificity to the neuron is UCHL-1/PGP9.5, which when knocked out produces the GAD mouse. Perhaps this could have been used, or another neuron specific UBP? Or a Velcade like boronate? Or a longer, more specific peptide inhibitor? The involvement of calcium dependent proteases is very interesting. If the pathway through calcium proteases to the proteasome (and the relevant substrates) were worked out, perhaps dissecting the targeting E3s and the upregulated proteases would be of use. The work could be followed by "proteomic" attempts at characterizing the upregulated elements of the Ub system, to determine which components are actually being used by the cell to tear down the axon. In whole, while the involvement of the proteasome is likely, and the elements of the teardown process have not been illustrated, the effect of this paper on the ubiquitin field is positive, and deserves note as a heads-up start for some biochemical dissection of the pathways involved.

View all comments by Christopher Larsen