Return of GEVI (Genetically encoded voltage indicator): GEVI strikes back against GCaMP.
Optogenetics combined with high-resolution microscopes, which provides increase spatial resolution render neurobiologists to understand those neural circuits involved in specific functions. In other words, it let us to ask some fundamental questions on how information is processed, consolidated and retrieved by different regions of the brain. Till date, most of the optogenetics experiments to perturb circuit function (switching on and off the activity of the neurons) heavily rely on using a genetically encoded fluorescent calcium indicator (GCaMP) as a proxy for measuring the neural activity. The final objective or a dream of a neurobiologist will be translating those activities into experience dependent behavioural responses. This objective has been partially achieved by optogenetics experiments using Gcamps nonetheless they draw a few limitations. 1. They possess slow kinetics, which precludes both the temporal analysis of spike timing and the resolution of spikes in fast spiking trains. Moreover calcium imaging neither measure precisely the sub-membrane threshold potentials nor the dendritic voltage dynamics, the latter play a pivotal role in shaping the integrative properties of how a neuron can bandwidth different inputs impinging on them. Learning how neurons integrate information from functionally different pathways or how individual representations were co-related temporally from functionally different pathways tunnelled on individual neurons are a dream to understand for many system neurobiologists. In other words, visualizing how different inputs (multiple presynaptic activity) are weighed by a neuron and how this integration culminates in an output (a single action potential or bursts) is one of the pressing questions in the field. Mark Schnitzer and colleagues had managed to develop a genetically encoded voltage indicator (GEVI), which allows high-speed recording of neural activity in wake mice and flies (1, the technical details of the methodology can be learned from the publication, here I will only record the potential use of GEVI in answering some of the fundamental questions of how neurons integrate information). Before enumerating the advantages of using GEVI over Gcamp, I will brief an introduction of the computational capabilities of dendrites, which receives the information, process it and send the integrated information to axon, which send those processed information to other receiver (another dendrites). Dendrites can perform a gamut of operations, which involves sub-linear summation of inputs by filtering of synaptic inputs in passive mode whilst generating supralinear spikes during active mode of information processing, there by acting as basic logical operation units (AND, OR, AND-NOT) to calculate complex operation on the incoming signals. This process is crucial to understand not only the computational abilities of dendrites but in general how brain works.
Dendrite structure, dendritic field, dendritic computation and its signal integrating properties.
The passive and active conductance properties of the dendrites dictate the timing and the rate of the output of the neuron as an all or none action potential (2, 3). It has been suggested that dendrites, with their distinct branching pattern in different types of neurons, are involved in specific tasks. Dendrite branching pattern and the dendritic field are involved in the normal functioning of many physiological processes. For instance, bi-tufted neurons modulate the temporal difference between sound inputs, (4). Amacrine cells and tangential cells confer directional selectivity by computing the direction of motion (5). Mitral cells with their functionally distinct apical and lateral dendrite use this particular organization for the specific processing of odor information and for odor discrimination (6). The area covered by the dendritic branching pattern governs the extent of the inputs the neuron can receive and compute, while the complexity of the branching of the dendrite determines its specialized task as mentioned in the examples above. The geometry of the dendritic branches, the number of synaptic sites, the distribution of different types of spines on the dendritic branches, the distribution of different voltage gated channels, and the history of the previous synaptic activity are all factors that determine how the dendrites integrate the incoming information, that the cell receives from multiple inputs spanning across the dendritic branches. Moreover, the computation of the proximal dendrites differs from that of the distal branches and the summation of their response also varies (7).
For the generation of an action potential (AP), the neurons require multiple summations of synaptic potentials because single input is not large enough to overcome the threshold for generating an AP. The farther the synapse is from the soma, the less likely it is that the EPSPs (Excitatory post synaptic potential) will reach to the axon’s initial segment, making it unlikely that an action potential will be generated. Though some neurons can adapt at scaling the input from different synapses, the dendrites acts as a leaky cable as predicted by cable theory (Rall W, 1959 and 1962). The distal dendrites far from the soma tend to be thinner and the voltage attenuation within them becomes asymmetrical. The distance dependent synaptic integration is attenuated both because of the passive properties of the dendrite and because the major players which determine the efficacy of generating action potentials are membrane resistance and dendritic branching pattern.
The “leaky cable’ properties of the dendrite are subdued by expressing different dendritic ion channels, which are voltage – gated and distributed uniformly or non – uniformly across the whole dendritic tree. Synaptic plasticity and co-incident detecting capability of neurons are mediated by the activation of regenerative currents primarily of Na+ and of voltage – gated Ca2+ channels in the dendritic regions leading to the generation of dendritic spikes. Dendritic spikes can detect co-incident synaptic inputs across the dendritic tree. (2,9). Retrogradely the spikes can propagate to the soma to generate burst of action potentials (10). Spatial and temporal activation of the distal dendrites with the back propagating action potential can generate bursts of somatic action potential and prior history of synaptic activity will augment this process. This form of dendritic excitability scales co –incident activation in different dendritic branches rendering the dendrites an alternative source of memory storage. The back propagating action potential summates with the local dendritic spikes thereby relieving the voltage sensors of NMDA channels in the spines, and initiating their production of LTP. The spines are strengthened in this way.
An elegant paper from Hausser and Collaborators (11), combining experimental data and computational stimulation shows how the dendritic morphology and geometry plays a pivotal role in governing the forward and backward propagation of APs. The authors have shown how different channels decorating the dendrite membranes determine the efficacy of back propagation of APs to the soma and to the distal part of the dendrites in different types of neurons. The authors speculate that the propagation of dendritic Ca2+ spikes as well as the regenerative Na+ currents evoked by high frequency stimulation are also determined by the number of dendritic branches besides the temporal and spatial summation of inputs clustered or distributed in the dendritic trees. The increase in the complexity should be also compensated by the increase in the number of synaptic inputs so that there is enough depolarization to overcome the active and passive properties of the dendrites to reach the distal part so that neurons can act as a co –incidence detector for strengthening those synapses which are activated.
How GEVI helps in understanding how neurons integrate information?
Combining GRIN lens (developed in Mark Schnitzer’s lab) or using miniature microendoscope with GEVI opens up further the possibility of measuring voltage conductance in dendrites as well as in soma in freely moving live animals.
1.GEVI helps in understanding how local dendritic spikes are generated during learning as well as during memory consolidation in live animals. Dendritic spikes require impinging of multiple synaptic inputs and this has to spatially localized and synchronized temporally. This causes the linear, sub and supra-linear dendritic integration of inputs co-relating with specific aspects of behavior. Along the line, GEVI helps us in measuring how clustered and dispersed input affect or integrate information processing of different dendrites in task related behaviors during learning.
2.NMDA and calcium spikes in dendritic branches are important for synaptic plasticity in vivo during learning tasks. GEVI helps to measure the intrinsic membrane conducting changes in active and non-active dendrites of a neurons and how that integrative ability shapes the generation of all or none output, action potential.
3.One of the long-standing questions that can be answered with the advent of GEVI with microscopy is that the inputs with same information of certain modality converge on the same dendritic branch or do they diverge on different dendritic branches? In both cases how do the information is processed?
4.Dendritic morphology and geometry plays a pivotal role in governing the forward and back propagation of APs. The branching pattern is of highly important especially since the orientation of the dendrites determines how the dendrites become a distributed circuit (12). The spines present on the dendrites enable an increase in connectivity with the passing by axons allowing them to maximize their input. Different neurons exhibit different dendritic branching patterns. For instance cortical pyramidal neuron differs in their branching patterns with respect to hippocampal or purkinje neurons. This partially explains why purkinje neurons with a large dendritic geometry are incapable of propagating the synaptic potentials to the distal regions. GEVI enables to measure the dendritic integration in different types of neurons and how they are functionally compartmentalized in freely moving animals.
5.GEVI helps in measuring on-going background activity (sub-membrane threshold potential) of the neurons that was excited during a learning task. Those neurons can be time tagged permanently using TRAP (targeted recombination in active population) so that one can permanently label the active neurons genetically which are exposed to learning paradigms (13). The measurements will reflect in the EPSP properties of these neurons such as the amplitude of EPSP peak, the time needed to reach the amplitude peak and also the half – width of the somatic EPSP’s in this activated neurons in vivo. In the latter case, the EPSP’s generated as a measure of distance from the soma reflects the half – width of the somatic EPSP’s. Besides it might be of interest in measuring of how neuromodulators affect the active properties of those neurons?
Taking into consideration that dendritic excitability might act as a substrate for metaplasticity and the integrative capabilities of the dendrites, which enables association, co-operation and integration of different inputs, impinging on the dendritic branches (14, 3,15), high speed recording of neural activity using GEVI combining with high resolution two photon microscopy and GRIN lens to image the voltage changes in dendrites in freely moving animals opens up new frontiers in systematically dissecting the complex picture of how different inputs is processed, integrated and produced as an output, the action potential in the brain.
1.Gong et al; High speed recording of neural spikes in awake mice and flies with a fluorescent voltage sensor, Science, 2015.
2.Larkum M et al; Synaptic integration in tuft dendrites of layer 5 pyramidal neurons: a new unifying principle, Science, 2009.
3.Sjöström PJ et al; Dendritic excitability and synaptic plasticity, Physiological Review, 2008.
4.Agmon-Snir H et al; The role of dendrites in auditory coincidence detection. Nature, 1998.
5.Borst A et al; Fly motion vision, Annual review of Neuroscience, 2010.
6.Dhawale AK et al; Non-redundant odor coding by sister mitral cells revealed by light addressable glomeruli in the mouse, Nat Neuroscience, 2010.
7.Polsky A et al, Computational subunits in thin dendrites of pyramidal cells, Nat Neuroscience, 2004.
8.Rall W, Theory of physiological properties of dendrites.” Annals of the New York Academy of Sciences, 1962.
9.Schiller J et al; NMDA spikes in basal dendrites of cortical pyramidal neurons, Nature, 2000.
10.Stuart G et al; Action potential initiation and back propagation in neurons of the mammalian CNS. Trends in Neurosciences, 1997.
11.Vetter P et al; Propagation of action potentials in dendrites depends on dendritic morphology. Journal of Neurophysiology, 2001.
12.Jia H et al; Dendritic organization of sensory input to cortical neurons in vivo, Nature, 2010.
13.Guenthner CJ et al; Permanent genetic access to transiently active neurons using recombination via TRAP: targeted recombination in active population, Neuron, 2013.
14.Hausser M; Storing memories in dendritic channels, Nat Neuroscience, 2007.
15.Poirazi P and Mel BW; Impact of active dendrites and structural plasticity on the memory capacity of neural tissue, Neuron, 2001.