A series of images showing different colors in the same image.

Project 1: Hilar Neurons & Dentate Gyrus Function

Hilar mossy cells are important because they regulate the primary neuron of the dentate gyrus, the granule cell. However, mossy cells can excite or inhibit granule cells, making their effects hard to predict. To clarify the role of mossy cells, we use selective chemo- or optogenetics to test the role of mossy cells on granule cells under different conditions and in behavior.

This project will address the role of mossy cells (MCs) in the dentate gyrus (DG), a cell type that has been implicated in normal DG function and in pathological conditions ranging from traumatic brain injury and epilepsy to psychiatric illness. We used a marker of neural activity to form hypotheses about the role of MCs in behaviors related to DG function and now employ optogenetics to selectively activate or silence MCs to establish their role; voltage imaging and patch clamp recordings will be used in a complementary manner. Together, the results will significantly advance our understanding of the normal DG and impairments that develop when MCs are injured or lost in disease.

Mossy cells (MCs) of the dentate gyrus (DG) are glutamatergic neurons that are considered to be important to normal function, and their injury has been suggested to contribute to neurological and psychiatric disorders, as well as deficits after traumatic brain injury. Anatomical and slice electrophysiology studies have described MCs in detail, but there is a gap between these studies and understanding how MCs contribute to DG-dependent behavior in vivo. To address this issue, we began with a simple approach: mice were engaged in behaviors related to DG function, and MCs were examined afterward using the neural activity marker c-fos. We quickly found that simply exploring novel objects led to a large increase in MC c-fos immunoreactivity (ir). Interestingly, most c- fos-ir did not increase in most of the other DG neurons, suggesting preferential activation of MCs by novelty. However, there was one area of the DG where c-fos-ir was consistent: a subset of GCs in dorsal DG. In contrast, the majority of MCs with c-fos-ir were ventral. Because the main projection of ventral MCs is to dorsal GCs, these data suggest that ventral MCs excite dorsal GCs. This circuitry helps explain how normally quiescent GCs become activated in dorsal DG, which is considered essential for cognitive functions of the DG. In Aim 1, we will use optogenetics to test this hypothesis, taking advantage of new mouse lines that have targeted Cre recombinase to MCs. We will also ask if dorsal MCs have effects analogous to ventral MCs, i.e., dorsal MCs contribute to ventral DG functions. In Aim 2, the underlying circuitry will be addressed. We suggest that optogenetic excitation of MCs in a normal adult mouse will recapitulate the results with c-fos: MCs excite proximal GCs weakly but distal GCs in a more robust manner. This idea has been supported by data from slices that were cut at an angle to preserve MC axons, and will be tested further in Aim 2 using voltage imaging and microelectrodes. In Aim 3, we will address the hypothesis that many of the distal GCs that are activated by MCs are immature. That hypothesis supports a previously published study showing that MCs are a primary source of afferent input to young GCs that are born in adulthood. This is potentially important because immature GCs are considered central to DG functions. Therefore, we will address the additional hypothesis that MCs activate adult-born GCs primarily in distal locations, leading to stronger excitation of distal GCs than proximal GCs. By providing afferent input to immature GCs, MCs could play a critical role in behaviors associated with adult DG neurogenesis. Together, these experiments will significantly advance our understanding of DG circuitry and its contribution to behavior. Because MC injury is associated with several disorders, these experiments will also shed light on impairments in DG functions in those pathological conditions.

Project no: R01 MH-109305
Principal Investigator: Helen E. Scharfman, Ph.D.
This support was from the National Institute of Mental Health at the NIH.

Project no: R37 Javitz award
Principal Investigator: Helen E. Scharfman, Ph.D.
This support was from the National Institute for Neurological Disorders and Stroke at the NIH.  The support is entering its 2nd of 7 years.


Principal Investgator

Project 2: Hippocampal Hyperexcitability in Alzheimer's Disease

Most people think that in Alzheimer's disease, the brain is severely degenerated, and the cause is a build-up of toxic proteins called amyloid plaques and neurofibrillary tangles. That may be true at the end stage of the disease, but we think the brain is actually overly active early in life – and people don't know because the activity occurs during sleep. We are currently testing this hypothesis in mice that simulate the disease.
This project addresses a novel hypothesis for the role of hyperexcitability in Alzheimer's disease (AD). Hyperexcitability is a term that refers to the tendency of neurons to fire too much. Although the term is commonly used in discussions of diseases like epilepsy, hyperexcitability also occurs in other diseases, and AD appears to be one example. We previously showed that a type of hyperexcitability found in epilepsy is a potential early biomarker in AD and a contributor to the progressive deterioration in AD. In this project, we test the idea that a subset of cholinergic neurons causes this type of hyperexcitability by synchronizing neurons. This idea is novel because many investigators assume the cholinergic neurons deteriorate in AD. That may be true at the end of life, but early in the disease, we think these cholinergic neurons are overly active and should be quieted to prevent the onset of the disease.

It has been suggested that neuronal hyperexcitability is an important characteristic of Alzheimer's disease (AD) because it contributes to the impairment in memory and increasing levels of amyloid β (Aβ) that characterize the disease. Using animal models of AD neuropathology, we suggest that the most common form of hyperexcitability is a synchronized spike in hippocampus and cortical neurons that is similar to the spikes between seizures in epilepsy, called interictal spikes (IIS). Our data suggests that IIS occurs very early and is very common, yet seizures are rare. Therefore, we have a potential opportunity to characterize a novel biomarker, IIS, and clarify the relationship between hyperexcitability and AD. Preliminary data have primarily used a mouse model where the precursor to amyloid precursor protein (APP), the precursor to Aβ, is mutated to simulate a Swedish family with familial AD and expressed widely in the brain. By 5 weeks of age, months before Aβ deposition, we have found IIS as the animals are sleeping. With age, the animals develop frequent IIS that occur in other brain states besides sleep, and there are also memory impairments and plaque formation. In other animal models, IIS also occurs, yet seizures are rare. When examining the brains of the young mice, we find that the basal forebrain (BF) cholinergic neurons and dentate gyrus granule cells show signs of elevated activity instead of hypoactivity that characterizes the brain at older ages. We suggest that BF cholinergic neurons stimulate the granule cells, and this leads to synchronized action potentials. In support, the muscarinic cholinergic antagonist atropine reduces the IIS in sleep, as well as in vitro measurements in hippocampal slices that we think reflect the abnormal activity. We now propose experiments to test these hypotheses with multiple methods, including viral-mediated silencing of cholinergic neurons in vivo. In the last part of the proposal, we will examine two strategies that our pilot experiments show can reduce IIS to determine if cognition and neuropathology are ameliorated. One of these has already been tested in a mouse model of Down's syndrome, a condition where AD is prevalent: maternal choline supplementation. The second, a reduction of the neurotrophin receptor p75 (p75NTR), has been tested in one of the mouse models we will use, the Tg2576 mouse, and it is already known that it ameliorates memory impairments in the mice. In summary, this project will address an area of AD research that has been difficult to clarify and controversial: hyperexcitability in AD. We suggest that there are early signs of hyperexcitability, IIS, that present opportunities for better mechanistic understanding and intervention.

Project no: R01 AG-055328
Principal Investigator: Helen E. Scharfman, Ph.D.
This support was from the National Institutes of Aging at the NIH.

A book cover and an image of the same.
A red and green image of the same picture

This is an image of the mouse brain showing a cross section through the hippocampus where area CA2 and area CA3 are located. In green is the pattern of expression of PCP4, which is present in area CA2 as well as other areas. On the right is expression of mCherry to show the area CA2 neurons only.

How we did it: A mouse was used that expresses Cre recombinase in area CA2. Under anesthesia, the mouse was injected in area CA2 with a benign virus that expresses mCherry if neurons express Cre recombinase.  Then the mouse was euthanized and perfused with fixative. The brain was sectioned and immunofluorescence was used to label cells with PCP4 in one section. In the next section imaging was done to show mCherry, which is a red fluorescent protein.

Project 3: The Role of Area CA2 in Epilepsy & Social Comorbidity

Area CA2 is a very small area of the brain, but it seems that it has many effects despite its small size. In the brain of a person with epilepsy, CA2 is even more important because it resists the damage that often occurs in areas nearby CA1 and CA3. Past work suggests that changes in area CA2 occur in epilepsy and may promote seizures. It also may contribute to behavioral changes that are comorbidities because area CA2 normally regulates social behavior. To test these hypotheses, we use a mouse model of epilepsy and selectively inhibit CA2 with chemogenetics to ask if seizures and behavior improve. Ultimately, we anticipate area CA2 could be a therapeutic target in epilepsy.

The CA2 region of the hippocampus, which has been shown to be important for social memory and aggression in mice, has been proposed to be important in temporal lobe epilepsy in humans. This proposal will use a novel mouse line to directly test the hypothesis that CA2 contributes to seizures and abnormal social behaviors associated with epilepsy in two animal models, with the promise of identifying novel therapeutic approaches.

A key challenge in temporal lobe epilepsy (TLE) is to determine the neural mechanisms contributing to seizures because current drugs fail to treat all seizures and usually come with debilitating side effects. Moreover, current drugs do not treat alterations in social behavior often manifested by individuals with epilepsy, including increased social aggression. This project will challenge current dogma as to the pathophysiological bases of how the hippocampus contributes to seizures in TLE, which has focused on three major regions of the hippocampus: dentate gyrus, CA3, and CA1. Instead, we examine area CA2, a relatively small region of the hippocampus that has received little attention but is known to survive relatively intact in TLE patients and rodent models and may serve as a seizure focus or facilitate seizure propagation. Experimental tools developed in the laboratories of the two Principal Investigators now enable the direct investigation of the importance of CA2 in mouse TLE models by employing a mouse line that expresses Cre recombinase relatively selectively in CA2 principal neurons. Cre-dependent viral vectors will be used to express genetically encoded tools in CA2 principal neurons to examine both alterations in CA2 circuitry in TLE and the effects of CA2 acute or chronic silencing on seizures. Thus, we will determine whether CA2 controls the pharmacological induction of acute seizures in the healthy brain and/or chronic seizures in the epileptic brain. We will also determine the importance of CA2 in reported deficits in social cognition and social aggression in mouse models of acquired TLE, as we find that CA2 is required for social recognition memory and is implicated in social aggression. As the social hormone arginine vasopressin promotes social memory and social aggression by enhancing CA2 input and output and has been shown to regulate seizures in animals, we will examine the role of CA2 regulation by this hormone on social and behavioral alterations in epileptic mice. By evaluating the role of CA2 in epilepsy, this project offers the promise of providing both basic mechanistic insight into seizures and social and behavioral comorbidity and may validate novel drug targets highly enriched in CA2 neurons. Difficult to clarify and controversial: hyperexcitability in AD. We suggest that there are early signs of hyperexcitability, IIS, that present opportunities for better mechanistic understanding and intervention.

Project no: R01 NS-106983-06
Principal Investigators: Helen E. Scharfman, Ph.D., Steven A. Seigelbaum, PhD.
This support is from the National Institutes of Neurological Disorders and Stroke at the NIH. The support is entering its 6th of 10 years.

Project 4 Mossy Cells in Temporal Lobe Epilepsy

Most people think that mossy cells of the dentate gyrus inhibit the primary cell type, granule cells, and this inhibits seizures. However, that idea is not consistent with several studies, such as the evidence mossy cells make numerous excitatory synapses on granule cells. We are testing the hypothesis that mossy cells can excite the granule cells strongly, and do so during the initial insult that causes epilepsy to develop. Therefore, mossy cells are a "bad guy" in epileptogenesis. In the epileptic brain, however, the circuitry changes, and then mossy cells are more powerful in their inhibition. Therefore, they are a "good guy" by inhibiting chronic seizures.

Temporal lobe epilepsy (TLE) is a disorder with recurrent, debilitating seizures as well as comorbidities that greatly decrease quality of life. Many patients respond poorly to medications, making research important to develop new treatments. A focus of research has been a part of the hippocampus called the dentate gyrus (DG) and a glutamatergic cell type in the DG called the mossy cell (MC). MCs have a direct excitatory projection to the main neuronal cell type, granule cells (GCs), so MCs are theoretically in an important position to regulate the role of the DG in TLE. Although MCs can excite GCs, many investigators consider MC-GC excitation to be weak. Instead, MCs are thought to primarily activate DG GABAergic neurons that inhibit GCs. In this proposal, we hypothesize that both the excitatory and inhibitory actions of MCs on GCs have important roles, particularly when an initial insult leads to TLE. Our central hypothesis is that during the initial insult, MC excitation of GCs plays a critical role because it strengthens greatly, leading to strong excitation of GC targets and excitotoxicity. In contrast, in chronic epilepsy, we hypothesize a very different MC role. We suggest that MCs resume their normal role to activate DG GABAergic neurons and inhibit GCs, which reduces chronic seizures. Therefore, during the initial insult, MCs should be inhibited for the best outcomes, and during chronic epilepsy, the MCs should be activated. If supported, this hypothesis would be a paradigm shift by changing the view of MCs in TLE. In addition, the proposed experiments will fill major gaps in knowledge because little is known about MCs during the initial insult, latent period, and chronic epilepsy.

Notably, MCs regulate behavior and cognitive tasks in normal mice. We recently showed MCs regulate tasks related to anxiety and cognition, which are comorbidities in TLE. Therefore, we hypothesize the role of MCs in the behavioral comorbidities in TLE.

Together, these experiments will challenge prevailing views and fill several knowledge gaps about MCs, the DG, and TLE. Furthermore, the experiments will potentially give rise to new approaches for therapeutics.

Project no: R01 NS-126529
Principal Investigator: Helen E. Scharfman, PhD.
This support was from the National Institutes of Health at the NIH.

A spectrogram of the frequency and direction of sound waves.
A yellow and blue background with red squares.

Project 5 Estrogen, Androgen & BDNF in Hippocampus

This project began with the identification by others that estrogen has a response-like element on the gene for brain-derived neurotrophic factor, and administering estrogen to a rat without its ovaries can increase BDNF in the brain. We had previously shown in male rats that BDNF increases glutamatergic transmission in a part of the hippocampus. Therefore we asked if estrogen did that also. We used rats that had their ovaries so we could ask if normal levels of estrogen had this effect. We timed experiments to coincide with stages of the ovarian cycle when estrogen was high or low. We found that estrogen led to an increase in BDNF protein and increased glutamatergic transmission as we predicted (Scharfman et al., 2003). We then showed the effects in rats that had no ovaries but the were treated so that they had a similar level of estrogen (Scharfman et al., 2007). This work led to many other findings about estrogen and BDNF that have translational implications. (See Recent Publications)

Project no.: R01 NS-37562

Principal Investigator: Helen Scharfman, PhD.

Title: BDNF and Hippocampal Hyperexcitability. 

This support has ended.