Reactive oxygen species (ROS) have been extensively studied as damaging agents associated with ageing and neurodegenerative conditions. Their role in the nervous system under non-pathological conditions has remained poorly understood. Working with the Drosophila larval locomotor network, we show that in neurons ROS act as obligate signals required for neuronal activity-dependent structural plasticity, of both pre- and postsynaptic terminals. ROS signaling is also necessary for maintaining evoked synaptic transmission at the neuromuscular junction, and for activity-regulated homeostatic adjustment of motor network output, as measured by larval crawling behavior. We identified the highly conserved Parkinson’s disease-linked protein DJ-1β as a redox sensor in neurons where it regulates structural plasticity, in part via modulation of the PTEN-PI3Kinase pathway. This study provides a new conceptual framework of neuronal ROS as second messengers required for neuronal plasticity and for network tuning, whose dysregulation in the ageing brain and under neurodegenerative conditions may contribute to synaptic dysfunction.

Neurons contain polarised microtubule arrays essential for neuronal function. How microtubule nucleation and polarity are regulated within neurons remains unclear. We show that γ-tubulin localises asymmetrically to the somatic Golgi within Drosophila neurons. Microtubules originate from the Golgi with an initial growth preference towards the axon. Their growing plus ends also turn towards and into the axon, adding to the plus-end-out microtubule pool. Any plus ends that reach a dendrite, however, do not readily enter, maintaining minus-end-out polarity. Both turning towards the axon and exclusion from dendrites depend on Kinesin-2, a plus-end-associated motor that guides growing plus ends along adjacent microtubules. We propose that Kinesin-2 engages with a polarised microtubule network within the soma to guide growing microtubules towards the axon; while at dendrite entry sites engagement with microtubules of opposite polarity generates a backward stalling force that prevents entry into dendrites and thus maintains minus-end-out polarity within proximal dendrites.

Animal behavior is principally expressed through neural control of muscles. Therefore understanding how the brain controls behavior requires mapping neuronal circuits all the way to motor neurons. We have previously established technology to collect large-volume electron microscopy data sets of neural tissue and fully reconstruct the morphology of the neurons and their chemical synaptic connections throughout the volume. Using these tools we generated a dense wiring diagram, or connectome, for a large portion of theDrosophilacentral brain. However, in most animals, including the fly, the majority of motor neurons are located outside the brain in a neural center closer to the body, i.e. the mammalian spinal cord or insect ventral nerve cord (VNC). In this paper, we extend our effort to map full neural circuits for behavior by generating a connectome of the VNC of a male fly.

Microtubules form polarised arrays throughout axons and dendrites necessary for neuronal growth and maintenance. γ-tubulin-ring-complexes (γ-TuRCs) nucleate microtubules at microtubule organising centres (MTOCs), but how microtubules are generated and organised within neurons remains unclear. We show that γ-TuRCs are predominantly recruited to the somatic Golgi, rather than to dendritic Golgi outposts, within Drosophila neurons. Microtubules nucleated from the somatic Golgi grow preferentially towards and into the axon, while growing microtubules that approach dendrites are excluded. Both directed growth and dendritic exclusion depend upon Kinesin-2, which associates with plus ends and directs their growth towards the plus ends of adjacent microtubules. We propose that plus-end-associated Kinesin-2 guides growing microtubules nucleated from the somatic Golgi towards the axon and prevents plus end entry into dendrites when engaging with oppositely polarised microtubules.In summary, microtubules are nucleated from the somatic Golgi within neurons and their guidance is necessary to maintain neuronal microtubule polarity. envelope, and different cells use different MTOCs to help generate and organise their specific microtubule arrays 18 . γ-TuRC recruitment occurs via γ-TuRC "tethering proteins" that simultaneously bind to the MTOC and γ-TuRC, possibly also helping to activate the γ-TuRC. An example is Drosophila Centrosomin (Cnn) and its mammalian homologue CDK5RAP2, which recruit γ-TuRCs to centrosomes during mitosis [23][24][25][26] and whose binding can activate γ-TuRCs 27-29 . γ-TuRCs can be recruited to different MTOCs by different tethering proteins, and it was also recently shown that different isoforms of Cnn can mediate recruitment to different MTOCs 29 . Thus, there is potentially a wide-range of γ-TuRC recruitment mechanisms available and understanding which mechanisms are used in different cells, including neurons, remains poorly understood.Although it is known that γ-tubulin is important 10-14 , it remains unclear how microtubule nucleation is regulated within neurons. During early development of mammalian neurons, the centrosome within the soma nucleates microtubules 30 that are severed and transported into neurites via motor-based microtubule sliding 31 . Microtubules sliding is also important for axon outgrowth in Drosophila cultured neurons 32,33 , and for establishing microtubule polarity [34][35][36][37][38] . Centrosomes are inactivated, however, at later developmental stages 30 and are dispensable for neuronal development in both mammalian and fly neurons 30,39 . No other active MTOCs within the neuronal soma have been described. Nevertheless, microtubules continue to grow within the soma 11,39 , and in mammalian neurons this depends in part on HAUS 11 , which is also important for microtubule growth within axons and dendrites 11,40 . Some MTOCs have been identified within dendrites in different neurons. For example, the basal body, or its surrounding region, within C. elegans ciliated neurons a...

19Neurons are inherently plastic, adjusting their structure, connectivity and excitability in 20 response to changes in activity. How neurons sense changes in their activity level and then 21 transduce these to structural changes remains to be fully elucidated. Working with the 22 Drosophila larval locomotor network, we show that neurons use reactive oxygen species 23 (ROS), metabolic byproducts, to monitor their activity. ROS signals are both necessary and 24 sufficient for activity-dependent structural adjustments of both pre-and postsynaptic 25 terminals and for network output, as measured by larval crawling behavior. We find the 26 highly conserved Parkinson's disease-linked protein DJ-1ß acts as a redox sensor in neurons 27 where it regulates pre-and postsynaptic structural plasticity, in part via modulation of the 28 PTEN-PI3Kinase pathway. Neuronal ROS thus play an important physiological role as 29 second messengers required for neuronal and network tuning, whose dysregulation in the 30 ageing brain and under neurodegenerative conditions may contribute to synaptic 31 dysfunction. 32 33 34Plasticity is fundamental to neuronal development and function. Changes in activity can 35 trigger a raft of different responses, including changes to synaptic size, strength and 36 connectivity, as well as to neuronal excitability (Davis and Müller, 2015; Keck et al., 2017; 37 Wefelmeyer et al., 2015). 'Hebbian' plasticity mechanisms, such as long-term potentiation or 38 depression, lead to altered transmission strength. These are balanced by homeostatic 39 mechanisms, which are compensatory in nature and work toward maintaining neuronal or 40 network activity within a set range. For example, blockade of excitatory inputs in neurons 41 can induce a range of changes that are compensatory, including enlargement of both pre-42 and post-synaptic specializations, increasing synapse, axonal bouton and dendritic spine 43 number as well as reducing dendritic spine elimination (Burrone et al., 2002; Kirov and 44 Harris, 1999; Murthy et al., 2001; Zuo et al., 2005). How neurons monitor their activity 45 levels is not known. Models of action potential firing rate homeostasis suggest that neurons 46 measure the frequency of action potential firing and use homeostatic plasticity mechanisms 47 to maintain this within their set point range (Davis, 2006; Turrigiano, 2012). In support of 48 this view, recent studies in the mammalian visual system demonstrated that following 49 dramatic activity disturbances, such as monocular occlusion, different types of neurons do 50 indeed return to their original cell type-specific firing rate (Hengen et al., 2013; 2016). 51An important second messenger in the regulation of activity-dependent plasticity is 52 calcium, whose intracellular concentration correlates with neuronal activity patterns. 53Calcium regulates cellular and synaptic changes through a raft of signaling pathways, 54 cytoskeletal and transcriptional effector proteins, including calcium/calmodulin-dependent 55 protein...

No abstract

No abstract

No abstract