Investigation of the neural substrates of motor and nociceptive behavior

A comprehensive understanding of the brain’s organization and function remains one of the biggest challenges we face today.
Despite an immense volume of research and tireless efforts to grasp the complexity of the brain, we are bound by the dichotomy between basic understanding of physiological neural function and making efforts to cure or treat conditions that hold a heavy toll on society like Alzheimer’s, Parkinson’s disease or amyotrophic lateral sclerosis (ALS) to name a few, which often hinders the convergence of different lines of research into a concerted effort. This dichotomy determines the course of two major research approaches that can be best exemplified by basic versus clinical research. The former being “knowledge-oriented” that investigates the basic component, mechanism and role of each specific unit that leads to a pathological condition, whereas the latter being “solution-oriented” centers on models that replicate as accurately as possible features of diseases in an effort to directly find more effective treatments or cures. These two pillars of research cannot be completely separated since without basic understanding of brain function it would be impossible to determine if and when pathological scenarios are arising, much like how without light it would be impossible to define what darkness is.
Taking the physiology-to-pathology or vice versa approach has both benefits and drawbacks, this choice is often determined by external factors like availability of appropriate disease models, technical limitations, severity and urgency of the disease studied and ultimately even personal preference.
A case in point are the current therapies used for Parkinson’s disease. The most effective treatments consist of dopamine replacement therapy in the form of L-dopa (Hornykiewicz, 2010) and deep brain stimulation (DBS) (Benabid et al., 2009). In the first case the dopamine precursor is provided orally to patients and helps compensate for the decreased dopamine levels of the disease (Birkmayer and Hornykiewicz, 1998). In the second case usually reserved for the most advanced and impairing forms of the disease when pharmacological treatments are no more efficient, electrodes are surgically implanted in the brain and a sustained electrical stimulation is hypothesized to subdue aberrant neuronal activity hence suppressing the symptoms. Both of these therapies although effective are far from optimal, and come with undesired side effects. L-dopa therapy is known to induce other forms of dyskinesia over time (Ahlskog and Muenter, 2001; Fahn et al., 2004), whereas DBS can induce cognitive impairments (Rezai et al., 2008).This clearly elucidates how there is ample room for improvement of these therapies and the fact that they have remained relatively unchanged for decades is an indication of stagnant technology.
The most obvious benefit of taking the pathology-first route is a potential much shorter time between scientific effort and therapy, on the down side, this type of study tends to be of high-risk/high-reward and can often be built upon limited or incomplete information about the system affected.
Basic research efforts have provided ample evidence of functionally defined pathways in the Basal Ganglia that are differentially affected in dopamine depleted models ( Gittis et al., 2011; Gittis and Kreitzer, 2012). Information processing in the striatum is also differentially affected based on the inputs (Parker et al., 2016). This knowledge is not reflected on current therapies, likely because the type of technologies used in basic research are not viable or adapted for clinical use yet.
Conversely one of the benefits of following the physiology-first approach is that it will ultimately provide an exhaustive blueprint of the intricate function of the brain area or network under examine, which will then give room to much more refined and specific therapies that would minimize if not eliminate side effects, and would be potentially much more effective. This process is however slow, laborious and is just as much limited by technical factors. One added benefit of compiling the so-called blueprint of a specific brain area is that this information would be suitable to apply in any and all diseases affecting it and perhaps even extrapolated onto other brain regions.
After centuries of research on the brain there is a virtually unanimous consensus on the fact that much like an electronic board, the brain is composed of billions of principal cells or neurons, each with the capacity to form thousands of connection nodes or synapses with other neurons, thus creating a remarkably intricate network that somehow encodes and governs every single function that the organism can execute, from sensory perception, to emotional states or physical actions.
One aspect of the challenge to understand the brain, is being able to dissect the sub-units of the overall network and unravel how they are relevant for all the different functions. This task is only possible if we somehow manage to target neuronal populations or pathways that hold a common characteristic be that the coding properties, anatomical features, inputs and outputs, type of neurotransmitter or neuromodulator, even receptor expression. The most encompassing perspective in this case is that of the genetic profile of different cells (Tasic et al., 2016). By having access to the genes expressed by neurons and using that as the targeting or flagging method, it’s possible to take the first step in the classification and profiling of defined neuronal circuits. Although genetic profiling and labeling of specific proteins or receptors has been around for decades, the biggest limitation until recently, was being able to manipulate the activity of these defined sub-classes of neurons in order to probe their function, and this is where recent technological advances in the field of neuroscience have allowed us to dramatically increase the pace in the endeavor to compile a library of neuronal function with a much higher degree of specificity, down to the gene level.
One of the main, if not the most influential finding in the area of systems neuroscience is the so-called optogenetic technology. From the very first years of the 21st century, a few selected laboratories were exploring the possibility to use light to control the activity of neurons (Zemelman et al., 2002; 2003). These seminal initiatives reached a pinnacle when in 2005 the first version of the nowadays most used light sensitive cation channel (Channelrhodopsin) that depolarizes neurons, was introduced. This technology has now evolved and in combination with CRE-loxP systems developed in the 90’s (Tsien et al., 1996; Tsien, 2016), it allows to specifically manipulate the activity of genetically defined neuronal populations driving or suppressing their function with pinpoint accuracy. In essence, a plasmid containing the gene that expresses the light sensitive-opsin is flanked by loxP sites and carried by adeno-associated viruses injected in defined regions of the brain of transgenic mice expressing the -CRE recombinase under a specific promoter. Thanks to the invaluable potential of this strategy the production of CRE-expressing transgenic mice has also tremendously developed in the last decade.
The core strength of this type of strategy lies in the ability to provide pathway and genetic specificity effectively isolating one specific tract of the circuit and probing its function allowing us to place this augmented information in the context of pioneering studies that in the 50s and 60s took the first step towards attributing defined behavioral functions to specific brain regions or nuclei through pharmacological lesions or electrical stimulation (Hoebel and Teitelbaum, 1962).
The main body of research described in this thesis takes place in this context of expansion and refinement of previous knowledge exploiting recently developed tools, something often referred to, as the age of optogenetics.
I will now provide a brief overview of the main findings of this thesis that will be introduced and covered in greater detail in the respective chapters.
The Basal Ganglia and motor behavior.
As previously alluded, Parkinson’s disease is one of the leading neurodegenerative conditions and affects motor behavior (Tanner and Aston, 2000). More precisely it leads to slowness of movement, muscle rigidity, uncontrollable tremors among other symptoms also including cognitive decline (Lang and Lozano, 1998). This disease is characterized by the peculiarity that symptoms usually manifest only after 70-80% of dopaminergic neurons in the Substantia Nigra have degenerated (Hornykiewicz, 1975). In essence, once the symptoms arise, it is too late.
This represents a good example where understanding the physiological organization and function of the brain areas involved is crucial to for example attempt a prediction of the cause of the progressive and seemingly irreversible degeneration of these neurons.
This first study describes how implementing a circuit based approach, I have identified two genetically targeted sub-populations of inhibitory neurons within the output nucleus of the Basal Ganglia (the Substantia Nigra pars Reticulata), that present different output target innervation patterns and underlie distinct aspects of motor coordination, furthermore synergistic activity of these two sub-circuits very efficiently shapes movement.
The periaqueductal role in ascending nociceptive signaling
One of the most basic yet indispensable functions of the brain is to identify potential threatening or dangerous situations, subjects or stimuli, integrate all the information available, and evoke an appropriate response in order to insure survival.
Something as simple as a noxious stimulus triggers a cascade of signals that ripple all over the brain, starting at the level of the spinal cord, inducing a reflexive reaction, and reaching far more complex effects such as an emotional reaction to that stimulus. With the novel tools and techniques available it is now possible to trace and decompose such a stimulus and get a picture of how this signal propagates and how different components of this signal are encoded by distinct neuronal populations or pathways.
From a classical point of view, a noxious stimulus arises from surface receptors, decussates at the level of the spinal cord and is then relayed to the brain via several so called ascending pathways that mainly culminate in thalamic and somatosensory areas of the brain. One of these pathways contacts the periaqueductal gray matter (PAG), a brain nucleus known for its involvement in analgesic responses to pain.
This second study has identified two pathways arising from the same genetically defined PAG neuronal population, that relay the neuronal activity evoked by a noxious stimulus to two different postsynaptic areas, likely encoding different components of a single stimulus.
Taken together these results provide evidence of how in very different contexts we are now able to assign specific characteristics of a single brain function to defined neuronal classes and networks opening the door to an unprecedented level of specificity.