Neurocircuitry underlying the antidepressant effect of retrograde facial botulinum toxin in mice

Abstract

Abstract Backgrounds Botulinum toxin type A (BoNT/A) is extensively applied in spasticity and dystonia as it cleaves synaptosome-associated protein 25 (SNAP25) in the presynaptic terminals, thereby inhibiting neurotransmission. An increasing number of randomized clinical trials have suggested that glabellar BoNT/A injection improves depressive symptoms in patients with major depressive disorder (MDD). However, the underlying neuronal circuitry of BoNT/A-regulated depression remains largely uncharacterized. Results Here, we modeled MDD using mice subjected to chronic restraint stress (CRS). By pre-injecting BoNT/A into the unilateral whisker intrinsic musculature (WIM), and performing behavioral testing, we showed that pre-injection of BoNT/A attenuated despair- and anhedonia-like phenotypes in CRS mice. By applying immunostaining of BoNT/A-cleaved SNAP25 (cl.SNAP25197), subcellular spatial localization of SNAP25 with markers of cholinergic neurons (ChAT) and post-synaptic membrane (PSD95), and injection of monosynaptic retrograde tracer CTB-488-mixed BoNT/A to label the primary nucleus of the WIM, we demonstrated that BoNT/A axonal retrograde transported to the soma of whisker-innervating facial motoneurons (wFMNs) and subsequent transcytosis to synaptic terminals of second-order neurons induced central effects. Furthermore, using transsynaptic retrograde and monosynaptic antegrade viral neural circuit tracing with c-Fos brain mapping and co-staining of neural markers, we observed that the CRS-induced expression of c-Fos and CaMKII double-positive neurons in the ventrolateral periaqueductal grey (vlPAG), which sent afferents to wFMNs, was down-regulated 3 weeks after BoNT/A facial pre-administration. Strikingly, the repeated and targeted silencing of the wFMNs-projecting CaMKII-positive neurons in vlPAG with a chemogenetic approach via stereotactic injection of recombinant adeno-associated virus into specific brain regions of CRS mice mimicked the antidepressant-like action of BoNT/A pre-treatment. Conversely, repeated chemogenetic activation of this potential subpopulation counteracted the BoNT/A-improved significant antidepressant behavior. Conclusion We reported for the first time that BoNT/A inhibited the wFMNs-projecting vlPAG excitatory neurons through axonal retrograde transport and cell-to-cell transcytosis from the injected location of the WIM to regulate depressive-like phenotypes of CRS mice. For the limited and the reversibility of side effects, BoNT/A has substantial advantages and potential application in MDD.