The Timing of Synapse Formation During Development Drives Social Behavior in Rats

Post by Shireen Parimoo

What's the science?

In the first week after birth, hippocampal neurons first exhibit GABA-mediated (inhibitory neurotransmitter) signaling followed by glutamatergic (excitatory neurotransmitter) signaling. Cortical neuronal populations exhibit glutamatergic activity in the first postnatal week, followed later by GABAergic activity. These developmental sequences of neurotransmitter activity correspond to the maturation of neurons in the brain. Network-level (groups of neurons)patterns of activity are important for complex behaviors like social interaction. However, it is not known whether a particular developmental sequence initiates a shift toward network-level activity that enables the development of more complex behaviors. This week in Nature Communications, Naskar and colleagues used electrophysiology and chemogenetics to investigate the relationship between the development of functional synapses and complex social behavior in newborn rats.

How did they do it?

The authors first recorded synaptic activity from pyramidal cells (neurons) in layers II/III and V of somatosensory cortex slices from rat pups between 2 and 15 days old (i.e. ‘postnatal days’ 2 and 15). They used whole-cell patch clamp electrophysiology to record postsynaptic currents to determine the relative timing of glutamate and GABA synaptogenesis (i.e. the creation of synapses). They also examined whether synaptogenesis was related to structural changes in these somatosensory cortex layers by staining the slices and counting the number of dendritic spines while mice were between 7 and 15 days old. They assessed the development of 1) complex sensory and motor behavior in pups by observing huddling behavior, and 2) more basic sensory and motor behavior by measuring performance on the modified SHIRPA test.development. Behavioral results from the SHIRPA test can be clustered in revealed three main categories of behaviors: (i) primitive behaviors like reflexes, and (ii) motor and (iii) sensory behaviors. Huddling behavior (more complex) included the time that pups spent together, the number of different clusters they formed, and the number of times they switched from one cluster to another.

To investigate the relationship between synaptogenesis and huddling behavior, the authors transfected somatosensory layer II/III precursor cells with inhibitory Designer Receptors Exclusively Activated by a Designer Drug (iDREADDs; synthetic receptors) or a control green fluorescent protein (GFP). The authors then recorded huddling behavior in pups before and after CNO application; clozapine N-oxide (CNO) is a synthetic drug that inhibits activity in cells expressing iDREADDs but has no effect on GFP-expressing cells. Finally, they investigated whether serotonergic transmission affects the relationship between cortical activity and huddling behavior. This was done by measuring serotonin transporter (SERT) levels in cortical slices of projections to the somatosensory cortex and assessing huddling behavior after increasing serotonergic activity with citalopram (inhibits serotonin reuptake) in the pups.

What did they find?

In layer II/III neurons of the somatosensory cortex, glutamatergic currents were observed at postnatal day 5, before GABAergic currents at postnatal day 7, and the frequency of both currents increased between postnatal days 8 and 9. This increase in the frequency of currents was associated with an increase in the number of dendritic spines on somatosensory neurons. By contrast, glutamatergic and GABAergic synapse formation occurred at the same time in layer V somatosensory neurons, whereas GABAergic currents preceded glutamatergic currents in hippocampal neurons. This indicates that developmental sequences are not the same across different cortical layers within a brain region or across different brain regions entirely.

The motor and sensory behaviors assessed in the SHIRPA were observed before the development of functional cortical synapses, suggesting that these basic behaviors are not dependent on the somatosensory cortex. On the other hand, the emergence of the more complex huddling behavior was aligned with the timeline of cortical synaptogenesis. Specifically, pups exhibited greater huddling behavior between postnatal days 8 and 9, consistent with the increase in synaptic current frequency in the somatosensory cortex. In fact, inhibiting somatosensory neurons by CNO in iDREADD-expressing pups reduced huddling behavior compared to control pups. This relationship between somatosensory cortical activity and huddling behavior was modulated by serotonergic transmission. Serotonin neurons project to layers I-IV of the somatosensory cortex, and there was a larger concentration of SERT on axons projecting to the somatosensory cortex on postnatal day 9 compared to postnatal day 7. Moreover, enhancing serotonergic activity with citalopram had no effect on basic behaviors as measured by the SHIRPA, but led to earlier emergence of huddling behavior compared to control pups.

What's the impact?

This study demonstrates that functional synapses follow a non-uniform pattern of development both within and across different regions of the brain. It further shows that early social behaviors occur concurrently with the development of glutamatergic and GABAergic synapses in the rodent somatosensory cortex, and that this is modulated by serotonergic transmission. Overall, this research provides deeper insight into the relationship between the precise timing of cellular processes during development and the emergence of complex, social behaviors.

Naskar et al. The development of synaptic transmission is time-locked to early social behaviors in rats. Nature Communications (2019). Access the original scientific publication here.

A New Neuron Population in the Hypothalamus Regulates Satiety and Prevents Obesity

Post by: Amanda McFarlan

What's the science?

The paraventricular nucleus of the hypothalamus contains a population of neurons expressing the SIM1 protein that have been shown to be critical in regulating feeding behaviour and satiety (feeling full). Neurons expressing the melanocortin-4 receptor (MC4R) were the first subset of SIM1-expressing neurons to be identified for their role in mediating satiety. However, previous studies have shown that inhibition of MC4R-expressing neurons causes increased appetite that only accounts for approximately half of that observed with inhibition of all SIM1-expressing neurons. Thus, there may be an unidentified population of SIM1-expressing cells, anatomically distinct from MC4R-expressing neurons, that also play a role in mediating satiety. This week in the Neuron, Li and colleagues investigated the role of prodynorphin (PYDN)-expressing neurons in the paraventricular hypothalamus in regulating satiety and bodyweight.

How did they do it?

The authors performed histological analysis in transgenic mice to investigate whether PDYN-expressing neurons and MC4R-expressing neurons were distinct cell populations within the paraventricular hypothalamus, and to determine whether PDYN-expressing neurons also expressed SIM1. Next, they explored how PDYN-expressing neurons affect satiety compared to MC4R-expressing neurons and SIM1-expressing neurons. To do this, they targeted the expression of an inhibitory DREADD (Designer Receptors Exclusively Activated by Designer Drugs) to PDYN-expressing neurons, MC4R-expressing neurons or SIM1-expressing neurons in the paraventricular hypothalamus and then observed the effect of inhibiting these neuronal populations on food consumption during a time of low caloric intake (the light cycle for mice).  

Next, the authors investigated the impact of long-term inhibition of PDYN-expressing neurons, MC4R-expressing neurons and both neuronal populations together on food consumption and bodyweight. They targeted either a tetanus toxin (to inhibit synaptic release) or a control virus (non-toxic) to each of the neuronal populations and measured changes in food intake and bodyweight over a one-month period. Next, they identified the brain areas that were innervated by PDYN-expressing neurons by injecting a Cre-dependent anterograde tracer into the paraventricular hypothalamus of a PDYN-Cre transgenic mouse. They used whole-cell recordings and Channelrhodopsin-2-assisted circuit mapping (CRACM) to assess glutamatergic transmission between PDYN-expressing neurons and their downstream connections. Finally, they used optogenetics to either increase or suppress the activity of PDYN-expressing neuronal terminals in the parabrachial complex to investigate the role of this circuit (PDYN-expressing neurons >> parabrachial complex) on food intake and satiety.

What did they find?

The authors determined that nearly all PDYN-expressing neurons also expressed SIM1, but not MC4R, suggesting that PDNY-expressing neurons are an anatomically distinct subset of SIM1-expressing cells. Then, they showed that chemogenetic inhibition of PDNY, MC4R and SIM1-expressing neurons caused an increase in food consumption compared to controls, and the effect of inhibiting PDYN or MC4R-expressing neurons was ~half as great as inhibiting SIM1-expressing neurons. Simultaneous chemogenetic inhibition of both PDNY and MC4R-expressing neurons, increased food intake to a level that was comparable to that observed with inhibition of SIM1-expressing neurons. These findings suggest that PDNY and MC4R-expressing neurons are two functionally independent subpopulations of SIM1-expressing neurons, that together, account for the majority of SIM1-mediated satiety. In addition, the authors found that long-term inhibition of PDYN- expressing neurons caused an increased appetite and a progressive increase in body weight compared to controls.

Next, the authors determined that PDYN-expressing neurons innervated and formed strong glutamatergic connections with neurons in a subregion of the parabrachial complex; this subregion was different than the subregion preferred by MC4R neurons. The results suggest that the pathway from PDNY-expressing neurons to parabrachial complex might be implicated in regulating satiety. Finally, they determined that optogenetic activation of PDNY-expressing neuronal terminals in the parabrachial complex decreased food intake, suggesting that the parabrachial complex is involved in regulating food intake. Conversely, they revealed that optogenetic inhibition of these neuronal terminals increased food intake, suggesting that the parabrachial complex is also necessary for satiety.

What's the impact?

This is the first study to identify a novel subpopulation of SIM1-positive neurons in the paraventricular hypothalamus, expressing PDYN, that are anatomically and functionally independent from MC4R-expressing neurons. These PDYN-expressing neurons were shown to play a key role in regulating feeding behaviour and satiety. Notably, PDYN and MC4R-expressing neurons were shown to have an additive effect that accounted for the totality of SIM1-expressing neuron-mediated satiety. This study has provided insight into the circuitry underlying feeding behaviour and satiety and may be critical in understanding how to better treat conditions such as obesity.  

Li et al. The Paraventricular Hypothalamus Regulates Satiety and Prevents Obesity via Two Genetically Distinct Circuits. Neuron (2019).Access the original scientific publication here.

Can We Alter the Progression of Huntington’s Disease?

Post by Anastasia Sares

What's the science?

The Huntingtin (HTT) gene has a number of roles in our brain, including neural development and transport of neuronal cell components, and we still don’t understand everything about it. We do know that the gene has an area where the base pairs “CAG” repeat a number of times. Sometimes, during DNA replication, the “CAG” gets stuck on repeat: if there are over 35 repeats, this leads to Huntington’s disease. The more repeats, the earlier the onset of symptoms, which include chorea (dance-like movements), muscle rigidity, lack of coordination, dementia, and depression. Statistically, 50% of the children of a person with Huntington’s will also have the disease.

In the mere 25 years since the discovery of the gene causing Huntington’s disease in 1993, there are now a myriad of possible approaches to treat this devastating genetic disease. This week in Neuron, Tabrizi and colleagues inventoried different treatment options for Huntington’s disease at the DNA, RNA, and protein level, showing how far in clinical trials each one has progressed, and evaluating their pros and cons.

What do we know?

In our body’s cells, genetic material (DNA) lives in the nucleus. In order to make functional proteins that do work in the rest of the cell, the DNA must first be transcribed into RNA, a messenger that takes the instructions outside of the nucleus, and then translated into proteins. The many repeats of “CAG” base pairs in the mutant HTT gene get translated into a long chain of abnormal material in the resulting protein. Because of HTT’s integral role in cell, these bad proteins have a variety of different effects, not least of which is that they can fragment off and cause neurofibrillary tangles. The tangles may lead to cell death in important brain regions like the striatum, which is responsible for movement selection and initiation.

When it comes to treating the disease, there are many different plans of attack. It might be possible to directly modify the mutant Huntingtin gene itself, chopping it out of the DNA. We could also target RNA, the messenger. Finally, we could intervene at the level of the Huntingtin protein, breaking down the mutants before they have a chance to affect other parts of the cell. However, silencing HTT, especially early in life, can cause a host of problems. A successful therapy must either silence ONLY the mutant HTT, or find a balance between reducing mutant HTT and leaving enough normal HTT for successful neural development. There’s another problem, too. Because of the mutations in the HTT gene, the cell doesn’t always follow the normal rules about where it should start and stop in the process of creating RNA or proteins. This can result in a number of non-standard proteins which are also toxic. An optimal therapy would be able to remove or reduce these non-standard proteins.

What’s new?

To make a treatment acceptable for use in humans, the method must first be demonstrated to be effective in cell cultures, other mammals, and non-human primates. It then proceeds to rigorous multi-phase clinical testing. Recent advances in DNA technologies like the CRISPR/Cas9 system allow for precision manipulation of DNA, and go directly to the source of the problem for a one-time treatment (this means the non-standard proteins will be taken care of as well). However, these technologies are very new and are still in the preclinical stage. Most DNA treatments, including CRISPR/Cas9, and also some RNA treatments, are currently very invasive, requiring insertion of foreign viral proteins directly into the brain. This is irreversible and might provoke inflammation or other immune responses, not to mention the high risks of brain surgery in general.

The most clinically advanced treatments for Huntington’s disease are RNA-targeting methods, especially antisense oligonucleotides (ASOs). Unlike the highly-invasive DNA treatments, ASOs can be administered via lumbar puncture. However, ASOs might have to be administered repeatedly, which isn’t ideal, and they can’t target all of the abnormal proteins generated by the mutant HTT gene. At the protein level, one therapeutic method would be to stimulate the cell’s native machinery to degrade mutant Huntingtin proteins faster (through PROTACS). However, this is also preclinical and needs to be developed further, as we don’t yet know the best way to deliver them to the central nervous system or what side effects they might have. No matter which method is chosen, silencing both normal and mutant HTT seems more promising since it won’t have to be as personalized for patients with different numbers of CAG repeats. However, if we are to decrease HTT on a system-wide level, the timing of the intervention is critical. It would be important to delay start time of the therapy  to avoid the period of neural development but also start treatment early enough for it to be effective. Having better detection methods for Huntington’s disease progression will be crucial to this endeavor.

What’s the bottom line?

The principles behind Huntington gene therapies also extend to many other genetic diseases. The main problem is how to successfully deliver these therapies. The most powerful and specific therapies are also the most invasive and dangerous, and editing DNA comes along with ethical concerns. There is still much work to be done in bringing these therapies to the clinic, and future research will need to focus on providing a safe delivery of therapies while mitigating harmful side effects.

Tabrizi et al. Huntingtin lowering strategies for disease modification in Huntington’s disease. Neuron (2019). Access the original scientific publication here.