Neuron
Previews pathogenesis involving a neurexin transsynaptic system that could include CNTNAP2 (Clarke and Eapen, 2014), this is not necessarily the case (Willsey et al., 2017). Further consideration of the global genomic architecture will be essential for understanding the likely multiple pathways and interactions involved in the TS phenotype. This will include wholegenome sequencing, with integration of results for rare variants affecting genomic structure (e.g., additional CNVs) and DNA sequence in coding and non-coding regions, and common genetic (e.g., polygenic) background. The results here for TS however are in line with the typical findings for neuropsychiatric conditions, including neurodevelopmental disorders such as intellectual disability, ASD, schizophrenia, speech disorders, ADHD, epilepsy, etc. Genetic heterogeneity within and between patients, and clinical heterogeneity with reduced penetrance
of individually important rare variants, are the norm. REFERENCES Bassett, A.S., Scherer, S.W., and Brzustowicz, L.M. (2010). Am. J. Psychiatry 167, 899–914. Clarke, R.A., and Eapen, V. (2014). Front. Hum. Neurosci. 8, 52. Costain, G., Lionel, A.C., Merico, D., Forsythe, P., Russell, K., Lowther, C., Yuen, T., Husted, J., Stavropoulos, D.J., Speevak, M., et al. (2013). Hum. Mol. Genet. 22, 4485–4501. Hu, J., Liao, J., Sathanoori, M., Kochmar, S., Sebastian, J., Yatsenko, S.A., and Surti, U. (2015). J. Neurodev. Disord. 7, 26. Huang, A.Y., Yu, D., Davis, L.K., Sul, J.H., Tsetsos, F., Rasmensky, V., Zelaya, I., Ramos, E.M., Osiecki, L., Chen, J.A., et al. (2017). Neuron 94, this issue, 1101–1111. Lowther, C., Speevak, M., Armour, C.M., Goh, E.S., Graham, G.E., Li, C., Zeesman, S., Nowaczyk, M.J., Schultz, L.A., Morra, A., et al. (2017). Genet. Med. 19, 53–61.
Mercati, O., Huguet, G., Danckaert, A., Andre´-Leroux, G., Maruani, A., Bellinzoni, M., Rolland, T., Gouder, L., Mathieu, A., Buratti, J., et al. (2017). Mol. Psychiatry 22, 625–633. Miller, D.T., Adam, M.P., Aradhya, S., Biesecker, L.G., Brothman, A.R., Carter, N.P., Church, D.M., Crolla, J.A., Eichler, E.E., Epstein, C.J., et al. (2010). Am. J. Hum. Genet. 86, 749–764. Oguro-Ando, A., Zuko, A., Kleijer, K.T.E., and Burbach, J.P.H. (2017). Mol. Cell. Neurosci. 81, 72–83. Robertson, M.M., Eapen, V., Singer, H.S., Martino, D., Scharf, J.M., Paschou, P., Roessner, V., Woods, D.W., Hariz, M., Mathews, C.A., et al. (2017). Nat. Rev. Dis. Primers 3, 16097. Uddin, M., Tammimies, K., Pellecchia, G., Alipanahi, B., Hu, P., Wang, Z., Pinto, D., Lau, L., Nalpathamkalam, T., Marshall, C.R., et al. (2014). Nat. Genet. 46, 742–747. Willsey, A.J., Fernandez, T.V., Yu, D., King, R.A., Dietrich, A., Xing, J., Sanders, S.J., Mandell, J.D., Huang, A.Y., Richer, P., et al. (2017). Neuron 94, 486–499.
Live or Die? Depends on Who You Are Takaaki Kuwajima1 and Carol Mason1,2,3,* 1Department
of Pathology and Cell Biology of Neuroscience 3Department of Ophthalmology College of Physicians and Surgeons, Columbia University, New York, NY 10032, USA *Correspondence:
[email protected] http://dx.doi.org/10.1016/j.neuron.2017.06.016 2Department
In this issue of Neuron, Welsbie et al. (2017) and Norsworthy et al. (2017) implicate the transcription factor Sox11 as a key player after optic nerve injury—in DLK signaling of RGC cell death, and in RGC regeneration and survival but only in certain RGCs. During development, neurons are born, their cell fate specified, and they differentiate, extending axons to target regions to form connections on target cells. After differentiation, through naturally occurring cell death, the number of neurons in a given population is adjusted, thought to reduce competition for synaptic space during the establishment of neural circuits. Following axon injury to the adult mammalian central nervous system (CNS), axons retract, unwanted cell death ensues, and without efforts for repair,
circuits are ultimately disconnected. In the last few years, a number of studies have pinpointed both intrinsic and extrinsic molecules and their signaling pathways that can implement both axon regeneration and neuronal survival in injured neurons (Crair and Mason, 2016; He and Jin, 2016). However, the molecular mechanisms that mediate survival versus death of specific types of neurons are still unclear. Further, if injured adult neurons do survive, one strategy for repair is to manipulate molecules present in the
adult but suppressed, such as mTOR, inhibited by PTEN, that then stimulates axon extension. An alternative strategy is to reprogram or reactivate factors that function in developmental growth programs. In this issue of Neuron, two papers report on each of these strategies, one elucidating pathways that lead to retinal ganglion cell (RGC) death after injury and the other demonstrating developmental programs that can be harnessed for axon regeneration but only for certain types of RGCs.
Neuron 94, June 21, 2017 ª 2017 Elsevier Inc. 1043
Neuron
Previews Welsbie et al. (2017) reveal a novel RGC signaling cascade via the dual leucine zipper kinase (DLK) that leads to cell death after optic nerve injury. Accumulating evidence has shown that DLK has multiple, but opposite, functions, e.g., enhancing axon outgrowth and degeneration, and neuronal migration and apoptosis, during development and post-injury degeneration in the adult nervous system of different species (Tedeschi and Bradke, 2013). Numerous studies have implicated DLK and its downstream molecules in RGC death and axon degeneration (He and Jin, 2016). Their previous functional genomic screen identified DLK as a mediator of RGC death after optic nerve injury in mouse (Welsbie et al., 2013). Moreover, in DLK-deficient retina, both proapoptotic and regeneration-associated genes emerged, but the precise signaling pathways were not identified (Watkins et al., 2013). In this issue of Neuron, Welsbie et al. (2017) set out to find out why pharmacological inhibition of DLK kinase is better at rescuing RGCs from death after injury than genetic deletion of Dlk. They hypothesized that additional kinases, together with DLK, promote RGC death, whereas simultaneous inhibition of such multiple kinases should promote RGC survival. A functional small interfering RNA (siRNA) screen for kinase targets identified the kinase Lzk (Map3k13) that works in concert with DLK; genetic inhibition of both Dlk and Lzk significantly increased RGC survival compared with single deletion of either Dlk or Lzk. Further, they showed that DLK/LZK-induced RGC death is mediated through activation of the kinases MKK4, MKK7, and JNK1–JNK3. Welsbie et al. (2017) then performed a second round of whole-genome siRNA screening to identify novel mediators of RGC death in a culture setting and found in addition to Dlk and MKK4 and MKK7, the transcription factors Jun, Atf2, Mef2a, and the Bcl-2 family member Puma as RGC death regulators. Knockdown of Atf2, Mef2a, or Puma improved RGC survival in vitro. They also validated the effect of one of these mediators, MEF2A, on RGC death in vivo; depletion of Mef2a increased RGC survival without any functional redundancy from other MEF2 members, e.g., Mef2c and Mef2d.
1044 Neuron 94, June 21, 2017
Since LZK has little effect on its own but potentiates DLK in mediating RGC death after injury, Welsbie et al. (2017) performed a third set of whole-genome siRNA screening to explore other possible elements downstream of DLK/LZK in mediating RGC death in retinal cell cultures when Lzk was knocked down. In addition to the identified regulators of RGC death, such as Dlk, Atf2, Jun, Puma, Mkk4 and Mkk7, and Jnk1–Jnk3, the transcription factor Sox11 was pinpointed as a RGC death factor through this screen, quite surprising given Sox11’s recently reported roles in RGC survival and differentiation during development (Chang et al., 2017; Kuwajima et al., 2017). Moreover, while deletion of Sox11 only slightly increased RGC survival in vivo after optic nerve crush, simultaneous deletion of Atf2, Mef2a, Jun, and Sox11 led to RGC survival to the same extent as deletion of both Dlk and Lzk. In sum, Welsbie et al. (2017) performed three different sets of wholegenome siRNA screens and identified several novel death signaling cascades in RGCs injured after optic nerve crush through ATF2, MEF2A, JUN, and SOX11 upon DLK/LZK activation (Figures 1A and 1B). The data from Welsbie et al. (2017) provide a clear path to therapeutic approaches for preventing RGC death after injury and in neurodegenerative diseases such as glaucoma. The work of Welsbie et al. (2017) raises the question: why does Sox11 knockdown only slightly improve RGC survival compared to deletion of DLK and LZK? One explanation is that SOX11 and other kinase targets, such as the transcription factors ATF2, MEF2A, and JUN, may have redundant roles in RGC death, and/or any one of these factors might kill distinct subpopulations of RGCs. Since simultaneous knockdown of these downstream targets rescues almost all RGCs from death, this explanation appears likely. In support of this idea, in this same issue of Neuron, Norsworthy et al. (2017) show that Sox11 does kill one particular RGC subtype after injury, while other RGCs are maintained. In the retina, there are nearly 30 RGC subtypes, and each of these subtypes express different genes, have different dendritic morphology, and project to different visual and non-image processing targets. Previous work from this lab has demonstrated that RGC sub-
types differ in their ability to survive after optic nerve injury. a-RGCs are resistant to cell death after optic nerve injury, and deletion of PTEN, which leads to activation of mTOR, or overexpression of osteopontin and IGF1 promotes axon regeneration only in a-RGCs (Duan et al., 2015). What molecular mechanisms might promote axon regeneration of other RGC subtypes? Norsworthy et al. (2017) took the strategy to restore defined transcription factors that regulate development of specific neuron subtypes to interrogate whether any of the transcription factors would reprogram RGCs to an immature state to acquire ‘‘developmental’’ properties for regeneration. Norsworthy et al. (2017) show that Sox11 plays such a role as a life-or-death decision maker for RGC axon regeneration in specific RGC subtypes. The group first tested seven transcription factors, all of which are involved in RGC specification and differentiation during development and were expected to reprogram adult RGCs to regrow. Norsworthy et al. (2017) overexpressed each transcription factor by viral delivery into the eye, several days before inflicting optic nerve crush. Among these transcription factors, including Sox2, overexpression of Sox11 most significantly promoted axon regeneration after optic nerve injury. Consistent with the data of Welsbie et al. (2017) in this issue of Neuron, Sox11 overexpression slightly decreased pan-RGC survival after injury. Most surprisingly, however, Sox11 overexpression completely killed a-RGCs but stimulated axon regeneration in non-a-RGCs (Figure 1C). To identify functional and regulatory targets of Sox11 in axon regeneration, Norsworthy et al. (2017) performed gene expression profiling in injured RGCs following overexpression of Sox11. The data revealed that genes associated with ceramide biosynthetic and metabolic processes were upregulated, implicating a ceramide-induced cell death pathway (Buccoliero and Futerman, 2003) in Sox11-induced a-RGC death. On the other hand, genes facilitating axonogenesis, cell-cell adhesion, and other axonal differentiative programs, such as Dcx, were upregulated. These data suggest that Sox11 overexpression may revert adult injured RGCs to ‘‘developing’’ RGCs while simultaneously killing specific
Neuron
Previews A
B
C
Figure 1. Regulation of RGC survival or death after optic injury (A) Among the cells in the retina, only RGC axons project to the brain. RPE, retinal pigment epithelium; P, photoreceptor; H, horizontal cell; B, bipolar cell; A, €ller glia. Retinal ganglion cells, (a), a-RGC (orange); (?), unidentified RGC subtypes (fuchsia, dark purple). amacrine cell; M, Mu (B) After injury, knockdown (KD) of Dlk and Lzk (Dlk + Lzk) or their downstream targets, Sox11, Atf2, Mef2a, and Jun (Sox11 + Atf2 + Mef2a + Jun), rescues RGCs from death. However, which RGC subtypes are maintained is unclear. (C) Knockout (KO) of PTEN promotes axon regeneration exclusively in a-RGCs (a), which are resistant to cell death even after axotomy (Duan et al., 2015). Overexpression (OE) of Sox11 kills a-RGCs but promotes axon regeneration in other RGC subtypes, whose identity is unknown. The combination of PTEN knockout and Sox11 overexpression enables axons to extend farther in these unidentified RGC subtypes compared with either PTEN knockout or Sox11 overexpression alone.
RGCs through the regulation of its targets. Interestingly, the developmental genes upregulated in the RGCs that survive do not include later synaptogenic genes, in agreement with findings that synaptogenic programs are at odds with axon regeneration (Tedeschi et al., 2016). Finally, in keeping with their earlier findings, Norsworthy et al. (2017) investigated possible interactions between Sox11 and PTEN by comparing Sox11 overexpression and PTEN deletion in axon regeneration and subtype-specific RGC survival. Whereas Sox11 overexpression or PTEN deletion alone led to only slight axon extension, the combination of Sox11 overexpression and PTEN deletion enabled some axons to reach the optic tract. Furthermore, although PTEN deletion normally maintains a-RGCs, Sox11 overexpression combined with PTEN deletion killed a-RGCs but maintained non-a-RGCs, similar to
Sox11 overexpression alone. These data point to synergistic effects of Sox11 overexpression and PTEN deletion in axon regeneration in non-a-RGCs and antagonistic effects on a-RGC survival. Taken together, the complementary findings of Welsbie et al. (2017) and Norsworthy et al. (2017) provide new insight into the molecular mechanisms regulating RGC, especially subtype-specific RGC survival and death after optic nerve injury. Moreover, one of DLK/LZK downstream targets, Sox11, plays a role in subtypespecific RGC survival and axon regeneration following injury. Future work should aim to clarify how Sox11 regulates RGC death in one subtype yet promotes axon regeneration in others. For example, regulatory regions of target genes could be masked or unmasked, or co-regulatory transcription factors could be expressed in one RGC subtype, but not in others. Addressing these questions will provide
better views of the molecular pathways for subtype-specific RGC development, survival versus death, and axon regeneration and, in turn, enable better therapeutic treatments for reconstructing retinal axon wiring after injury. REFERENCES Buccoliero, R., and Futerman, A.H. (2003). Pharmacol. Res. 47, 409–419. Chang, K.C., Hertz, J., Zhang, X., Jin, X.L., Shaw, P., Derosa, B.A., Li, J.Y., Venugopalan, P., Valenzuela, D.A., Patel, R.D., et al. (2017). J. Neurosci. 37, 4967–4981. Crair, M.C., and Mason, C.A. (2016). J. Neurosci. 36, 10707–10722. Duan, X., Qiao, M., Bei, F., Kim, I.J., He, Z., and Sanes, J.R. (2015). Neuron 85, 1244–1256. He, Z., and Jin, Y. (2016). Neuron 90, 437–451. Kuwajima, T., Soares, C.A., Sitko, A.A., Lefebvre, V., and Mason, C. (2017). Neuron 93, 1110– 1125.e5.
Neuron 94, June 21, 2017 1045
Neuron
Previews Norsworthy, M., Bei, F., Kawaguchi, R., Wang, Q., Tran, N., Li, Y., Brommer, B., Zhang, Y., Wang, C., Sanes, J.R., et al. (2017). Neuron 94, this issue, 1112–1120.
Tedeschi, A., and Bradke, F. (2013). EMBO Rep. 14, 605–614.
Tedeschi, A., Dupraz, S., Laskowski, C.J., Xue, J., Ulas, T., Beyer, M., Schultze, J.L., and Bradke, F. (2016). Neuron 92, 419–434. Watkins, T.A., Wang, B., Huntwork-Rodriguez, S., Yang, J., Jiang, Z., Eastham-Anderson, J., Modrusan, Z., Kaminker, J.S., Tessier-Lavigne, M., and Lewcock, J.W. (2013). Proc. Natl. Acad. Sci. USA 110, 4039–4044.
Welsbie, D.S., Yang, Z., Ge, Y., Mitchell, K.L., Zhou, X., Martin, S.E., Berlinicke, C.A., Hackler, L., Jr., Fuller, J., Fu, J., et al. (2013). Proc. Natl. Acad. Sci. USA 110, 4045–4050. Welsbie, D.S., Mitchell, K.L., Ranganathan, V., Sluch, V.M., Yang, Z., Kim, J., Buehler, E., Patel, A., Martin, S.E., Zhang, P.-W., et al. (2017). Neuron 94, this issue, 1142–1154.
Staring at the Clock Face in Drosophila Ezio Rosato1 and Charalambos P. Kyriacou1,* 1Department Genetics & Genome Biology, University of Leicester, Leicester, LE1 7RH, UK *Correspondence:
[email protected] http://dx.doi.org/10.1016/j.neuron.2017.06.005
Liang et al. (2017) demonstrate how neuropeptides from two groups of clock cells appear to be responsible for the fly’s circadian neurons becoming active at different times of day. By delaying the activity of their clock cell targets, they give rise to morning and evening behavior. We live in a world of relentless daily light and temperature cycles and so most eukaryotes and some bacteria have evolved an endogenous circa-24 hr (circadian) clock that synchronizes the processes of life with its rhythmic environment. In animals, behavior is rhythmic thanks to the action of clock neurons in the brain. In the fruit fly Drosophila melanogaster, 75 clock neurons per hemisphere are distributed into six main clusters, three laterally (s-LNv, l-LNv, and LNd) and three dorsally (DN1, DN2, and DN3) (see Figure 1). The molecular components of the clock cycle in unison in all of them. For instance, under both lightdark cycles (LD) and constant darkness (DD) the clock protein PERIOD (PER) reaches a peak at the end of the night and a trough at the end of the day. However, under LD and for the first few days in DD, the rest/activity cycle of flies is bimodal, with a peak of locomotor activity in the morning (the ‘‘M’’ component) and another in the evening (‘‘E’’). Genetic dissection of these clock neurons revealed that the s-LNv cells are typically responsible for early morning activity (hence ‘‘M’’ cells), and likewise, the LNds generate the evening component, so called ‘‘E’’’ cells. But how can a unimodal and synchronous molecular clock
inform the different cells of their different M or E duties? Liang and co-authors began addressing this issue in a previous publication (Liang et al., 2016). They used light-sheet microscopy (Holekamp et al., 2008) and a genetically encoded Ca2+ sensor GCaMP6s, to measure intracellular levels of Ca2+ in all circadian neuron groups over a 24 hr period. They observed that the different groups show maximal Ca2+ levels, hence neuronal activity, at different times of the day. The s-LNv neurons peak around dawn, the LNd around dusk, the l-LNv around midday, and DN1 and DN3 around midnight. In simple terms, although all circadian clusters agree on what time it is via their canonical molecular clock, the timing of their peak excitabilities are distinct and sequential. While this was an important and highly significant step in the right direction, it still did not illuminate the mechanisms that sequentially time neuronal activation. For the fly, light is the major entraining cue, and the neuropeptide Pigment Dispersing Factor (PDF) produced by the s- and l-LNv cells has long been recognized as a major synchronizer of the molecular clock in circadian neurons. In a new paper in this issue of Neuron, Liang et al. (2017) have now investigated
1046 Neuron 94, June 21, 2017 ª 2017 Elsevier Inc.
the effects of light and PDF on the sequence of excitability within the network. They examined mutants that eliminate the production or the response to PDF under LD or DD. They observed that the LNds respond to the light-dark regime and to PDF by delaying their Ca2+ rhythm. In contrast, only PDF signaling was able to delay the DN3 while l-LNv and DN1 were insensitive to the neuropeptide. The phase of s-LNv was not affected by these manipulations but their Ca2+ ‘‘wave’’ became wider. When PDFR, the PDF receptor, was re-introduced specifically in those cells in an otherwise PDFR mutant background, the normal width of the Ca2+ ‘‘wave’’ was restored. Hence PDF signaling delays activation of the LNds and contributes to curbing the activation of the s-LNvs. Bath applications of synthetic peptide showed that in both cell types, the effects of PDF can be explained by a reduction of Ca2+ levels and that those effects require the presence of PDFR. Ectopic expression of PDFR in the naturally PDFR null l-LNv further suggested a cell-autonomous mechanism. Moving back to the effect of light, Liang et al. (2017) investigated how light pulses delivered at the beginning or the end of the night influenced the Ca2+ waves on
