Response to: Kim et al., “Axon Regeneration in Young Adult Mice Lacking Nogo-A/B.” Neuron 38, 187–199

Abstract

Department of Neurosciences, University of California at San Diego, La Jolla, CA 92093, USA*Correspondence: osteward@uci.eduDOI 10.1016/j.neuron.2007.04.004IntroductionNogohasbeenproposedtobeamajormyelin-derived inhibitor of axon re-generation in the mammalian centralnervous system (CNS). Three studiestested this hypothesis by assessingregeneration in various Nogo-deficientmice following spinal cord injury (Kimet al., 2003; Simonen et al., 2003;Zhengetal.,2003),yetdifferentregen-erative responses were reported. Instudies of a Nogo gene trap mutant,Kim et al. (2003) reported dramaticallyenhanced sprouting above the lesion,reflected by the presence of CSTaxons in ectopic locations (the lateralfuniculus ipsilateral to the injection) inall (12/12) Nogo knockout mice. Evi-dence of striking long-distance regen-eration was the presence of ectopicBDA-labeled axons more than 5 mmcaudal to the injury in 7/11 Nogoknockout mice that received spinalcord injuries at 7.5–9 weeks of age.Three mice that underwent surgery at11–14 weeks of age did not exhibit ev-idence of long-distance regeneration.The labeled axons in caudal segmentsseemed to be especially compellingevidence of robust but disorderly re-generationoftheCSTbecauselabeledaxons were present bilaterally in thelateral funiculus (an ectopic location).In contrast, Simonen et al. (2003)reported increased sprouting nearthe lesion site in a Nogo-A-targetedknockout(inwhichNogo-Bexpressionwas upregulated) and evidence foren-hanced regeneration below the lesionin a subset (4/16) of mutants, althoughthere was no statistically significantdifferencebetweenknockoutandcon-trols. Our group found no statisticallysignificant difference between Nogo-A,B(orNogo-A,B,C)mutantsandwild-type controls in the extent of CSTsprouting above the lesion or regener-ation past the lesion (Zheng et al.,2003). One enigmatic finding was thatectopic axons similar to those re-ported by Kim et al. were observed inone control mouse (see Figure S2 inthe Supplemental Data of Zhenget al., 2003).Possible reasons for the discrepantresults have been discussed (Woolf,2003), but there has been no satisfac-tory explanation of the apparently ro-bust regeneration in one study andmodest or no enhanced regenerationin the others. Accordingly, we set outto perform a full replication of the ex-periments in the knockout line usedin our study (Zheng et al., 2003) andthe gene trap line used by Kim et al.(2003), which were reported to exhibitstriking regeneration. In the course ofour studies, we discovered that axonsin the white matter stain artifactuallywhenBDAleaksintothecerebrospinalfluid (CSF) in the cerebral ventricle,producing a pattern of ectopic labeledaxons that is remarkably similar towhat was interpreted as massivesprouting/regeneration by Kim et al.Inthisreport,wedocumenttheartifac-tual labeling, show that it occurs asa result of BDA entering the CSF, anddemonstrate that the artifact may beavoided by delaying the tract tracinginjections for 4 weeks after spinalcord lesions.ResultsIn all three studies of Nogo knockoutmice, regeneration was assessed fol-lowing transection of the dorsal halfof the spinal cord (a dorsal hemi-section) atT8, which in mice interruptsdescending CST axons while sparingwhite matter and gray matter in ventralportions of the spinal cord. Possibleregeneration of cortico-spinal tract(CST) axons was then assessed byinjecting biotinylated dextran amine(BDA) into the sensorimotor cortex.Importantly, all three studies used asimilar, short-term protocol: the corti-cal BDA injection was made in thesame operative procedure as the spi-nal cord lesion, and regeneration wasassessed 2–3 weeks later.In our replication experiment, weusedthesameprotocoltoassesspos-sibleCSTregenerationinhomozygousmutant and control (wild-type and het-erozygotes). BDA was injected usingthe coordinates described in Zhenget al. (2003) (Kim et al. [2003] did notreport their coordinates). One smallprocedural difference that turned outto be important was that, in our previ-ous study (Zheng et al., 2003), BDAwas injected at a depth of 0.5 mm inthe cortex, whereas in the repeatexperiment we lowered the syringebelow 0.5 mm and then withdrew tothe final location. This was done toachieve greater accuracy becausethe cortex is frequently displacedslightlyasthemicrosyringe islowered.As discussed below, this had theunin-tended consequence that the syringesometimes penetrated the ventricle.In intact mice (that is, without spinalcord injuries), injections of BDA intothe sensorimotor cortex producerobust labeling of dorsal main CST inthe ventral part of the dorsal columnNeuron 54, April 19, 2007 a2007 Elsevier Inc. 191