PT - JOURNAL ARTICLE AU - Makoto Matsuyama AU - Lei Jin AU - Thomas K. Lavin AU - Heather A. Sullivan AU - YuanYuan Hou AU - Nicholas E. Lea AU - Maxwell T. Pruner AU - María Lucía Dam Ferdínez AU - Ian R. Wickersham TI - “Self-inactivating” rabies viruses are susceptible to loss of their intended attenuating modification AID - 10.1101/550640 DP - 2020 Jan 01 TA - bioRxiv PG - 550640 4099 - http://biorxiv.org/content/early/2020/11/11/550640.short 4100 - http://biorxiv.org/content/early/2020/11/11/550640.full AB - An article in Cell reported a new form of modified rabies virus that was apparently capable of labeling neurons “without adverse effects on neuronal physiology and circuit function” but that nevertheless was able to spread between neurons as efficiently as the widely-used first-generation deletion-mutant (ΔG) rabies viral vectors. The new “self-inactivating” rabies (“SiR”) viruses differed from first-generation vectors only by the addition of a destabilization domain to the viral nucleoprotein. We noticed that the transsynaptic tracing results from that article appeared inconsistent with the strategy described in it: specifically, the viruses were able to spread between neurons even in the absence of the exogenous protease that was meant to be required. We hypothesized that the viruses used were actually mutants that had lost the intended addition to the nucleoprotein, making them de facto first-generation viruses. We obtained samples of two SiR viruses from the authors and show here that the great majority of viral particles in both the “SiR-CRE” and “SiR-FLPo” samples were mutants that had lost the intended modification, consistent with our hypothesis. We also found that SiR-CRE killed 70% of infected neurons in vivo within two weeks, consistent with the prediction that mutants without the intended modification would share the toxic phenotype typical of first-generation rabies viral vectors. We hypothesize that the same or similar mutations were present in the viruses used in the original article and that this explains the paradoxical reported findings. While it may be possible to successfully make SiR viral preparations that are not dominated by such mutants, and while it may also be possible that such intact SiR viruses are indeed nontoxic to neurons, we predict that it will not be possible to replicate the transsynaptic tracing results from the original paper unless using mutants similar to the ones that we report here.Competing Interest StatementThe authors have declared no competing interest.