Student Seminars and Defenses

Silei Zhu, MS - PhD Candidate, Neuroscience Graduate Program

In the early visual system, functional significance of the feedforward projection, from the retina to the dorsal lateral geniculate nucleus (LGN), and from LGN to the primary visual cortex (V1), is relatively well studied. In contrast, corticogeniculate (CG) feedback to LGN, which mainly originates from V1, is less well understood. Interestingly, CG feedback constitutes more than 30% of total synaptic input onto LGN neurons, outnumbering retinal feedforward synapses (5%-10%). The overall goal of my thesis is to study the behavioral effects of selective manipulation of CG feedback and to develop relevant methods. We first examined the effects of selective optogenetic suppression of CG neurons in anesthetized ferrets. Optogenetic suppression of CG feedback decreased activity among LGN neurons in the absence of visual stimulation but did not affect the visual responses of LGN neurons, suggesting that feedforward visual stimulus drive overrode weak corticogeniculate influence. Optogenetic effects on LGN and V1 neuronal responses depended on the frequency of LED illumination, with higher frequency illumination inducing slow oscillations in V1, dis-inhibiting V1 neurons locally, and producing more suppression among LGN neurons. We then trained ferrets to perform freely moving visual discrimination tasks and tracked their head and eye movements. Heading was predictive of choices as well as biases and decision strategies. While saccades also predicted choices, they were less predictive than heading and occurred after head turning. These findings characterize ferrets' head and eye movements during freely moving visual discrimination tasks and show that when unrestrained, ferrets orient first with their heads and then with their eye movements. Furthermore, these methods also provide a unique paradigm to probe the continuous process of visual decision-making in a more naturalistic manner. We also developed methods of long-term viral expression of genes of interest in ferrets. We injected AAV2 locally in V1 and confirmed viral expression in V1 neurons 14 months post-injection. However, AAV2-retro did not retrogradely label CG neurons after injection into ferret LGN. Instead, FuG-E lentivirus retrogradely infected CG neurons after injection into LGN and we confirmed viral gene expression 9 months post injection. Interestingly, FuG-E lentivirus barely labeled any neurons after local injection in V1. To determine the visual field affected by viral infection, we first mapped the retinotopy of ferret V1 on a ferret MRI brain atlas, and then registered virus expression patterns to this atlas to obtain the elevation and azimuth of each virus-infected neuron. We expressed inhibitory chemogenetic channels in V1 in one ferret but did not detect any behavioral effects of chemogenetic inactivation. To test the behavioral impact of optogenetic activation of CG neurons, we injected FuG-E lentivirus expressing ChR2 into LGN to label CG neurons in V1 retrogradely. After stimulation of CG neurons using wireless optogenetics during freely moving visual discrimination tasks, we detected a significant decrease of task performance in one ferret specifically in the visual field contralateral to the LGN injected with virus. However, at 10 months after the detection of behavioral changes, we did not detect viral expression in post-mortem histology. No significant behavioral effect was found in another animal 4-10 months post-injection, perhaps due to mismatch between the position of the visual stimulus and the retinotopic location of virus-infected neurons. Through these experiments, we learned the importance of completing behavioral tests within 1 year post viral injection to avoid expression decay, and the importance of developing methods to confirm success of viral expression and estimate affected receptive fields in vivo, e.g. using retinal imaging.

 Nov 21, 2024 @ 9:00 a.m.
 Medical Center | K307 (3-6408)

Matthew Adusei, MS - PhD Candidate, Neuroscience Graduate Program

Visual signals follow a feedforward progression: the dorsal lateral geniculate nucleus (LGN) receives and relays signals from the retina to primary visual cortex (V1). V1, in turn, provides reciprocal feedback to the LGN. Because the vast majority of LGN inputs target V1 (and secondary visual cortex or V2), their complementary corticogeniculate neurons are assumed to be similarly restricted to V1 and V2. However, there are direct inputs mainly from koniocellular LGN neurons to mid-level, extrastriate visual areas. Whether there are corticogeniculate neurons in mid-level extrastriate cortex that project to the LGN remains unknown. Additionally, since corticogeniculate neurons in V1 modulate the timing and precision of LGN responses, it would be important to investigate whether extrastriate corticogeniculate neurons serve analogous or different functions. In this thesis work, I investigate these unknown questions using virus-mediated gene delivery to 1) retrogradely trace corticogeniculate circuits and 2) optogenetically activate corticogeniculate feedback in primates and ferrets, carnivores with visual pathways similar to those in primates.

First, we identified and characterized the morphology of multiple distinct corticogeniculate neurons in mid-level extrastriate visual cortical areas of ferrets: posteromedial lateral suprasylvian (PMLS), posterolateral lateral suprasylvian (PLLS), and area 21a, and macaque monkeys: middle temporal (MT), medial superior temporal (MST), and area V4. Importantly, all three areas in both species were dominated by corticogeniculate neurons with spiny stellate morphology, suggesting possible preferential targeting of W/koniocellular LGN layers. We also observed corticogeniculate neurons in other extrastriate visual cortical areas, although we did not systematically characterize them.

Toward the second aim, we found that activating corticogeniculate feedback from PMLS in ferrets shifted the preferences of LGN neurons to low spatial and high temporal frequency stimuli, which aligns with the preference of PMLS neurons for fast-moving stimuli.

Together, our results suggest that: (1) extrastriate corticogeniculate feedback from PMLS may enhance LGN responses to fast-moving (high temporal frequency) stimuli, (2) evolutionary preservation of corticogeniculate neurons throughout visual cortex supports their critical role in visual function, (3) extrastriate geniculo-cortico-geniculate loops, that bypass V1, could provide a substrate for residual vision following V1 damage, (4) other sensory systems may contain corticothalamic neurons beyond primary and secondary sensory cortex that also target first order thalamus, (5) broader characterizations of these circuits could provide additional clues about the overall functional roles of corticothalamic feedback in sensory perception, and (6) the presence of corticogeniculate neurons across visual cortex necessitates a reevaluation of the LGN as a hub for visual information rather than a simple relay.

Zoom Passcode:  659426

 Nov 22, 2024 @ 9:00 a.m.
 Medical Center | K207 (2-6408)

Emma Bryson, Yunshan Cai, & Alex Solorzano - PhD Candidates

Titles: TBD

Faculty Evaluators: Chris Holt & Harris Gelbard

Student Moderator: Nicole Popp

 Nov 25, 2024 @ 4:00 p.m.
 Medical Center | K-207 (2-6408)

Skylar DeWitt, Wen Li, & Gueladouan Jean Setenet - PhD Candidates

Titles: TBD

Faculty Evaluators: Liz Romanski & Jennetta Hammond

Student Moderator: Staci Rocco

 Dec 02, 2024 @ 4:00 p.m.
 Medical Center | K-207 (2-6408)

Tracey Preko, Pavel Rjabtsenkov, & Adam Roszczyk - PhD Candidates

Titles: TBD

Faculty Evaluators: Juliette McGreger & Chris Proschel

Student Moderator: Stacey Pedraza

 Dec 09, 2024 @ 4:00 p.m.
 Medical Center | K-207 (2-6408)

Daulton Myers - PhD Candidate, Neuroscience Graduate Program

The anterior cingulate cortex (ACC) is a heterogenous structure that is strongly connected with the amygdala and contains subdivisions critical for unique limbic and cognitive functions. The subgenual ACC (sgACC, Brodmann area 25/14c), implicated in major depression in humans, is a key node of the salience network and is important for arousal state modulation and valuation of sensory information. The perigenual ACC (pgACC, Brodmann area 32/24b), which is positioned dorsal to the sgACC, is important for a host of cognitive functions including decision-making and conflict monitoring. Despite known functional differences in sgACC and pgACC, the main cortical and thalamic drivers of the ACC subregions are not fully understood in higher species. Our preliminary data in macaque suggests each region is unique. sgACC is weighted towards prefrontal cortical (PFC) and thalamic afferents carrying information about motivational states and the value of sensory cues, including midline thalamic nuclei and Brodmann area 13. In contrast, pgACC receives unique inputs from mediodorsal (MD) thalamus and Brodmann area 9/46 that carry information important for spatial and temporal localization of salient stimuli. In Aim 1, I will use paired retrograde tracer injections and compare ensembles of prefrontal cortical and thalamic afferents to the macaque sgACC and pgACC within the same animals. I hypothesize that unique combinations of inputs drive sgACC and pgACC, with sgACC being weighted towards key areas of the arousal network and pgACC being weighted towards cortical and thalamic areas important for goal-directed behavior and decision-making.

The amygdala is critical for ACC function, but its specific inputs to sgACC and pgACC are not clear. The basal nucleus of the amygdala, a 'cortical-like' nucleus critical for detection of salient cues such as facial expression and facial identity, has strong inputs to the ACC. The basal nucleus of the amygdala is enlarged in primates compared to rodents and is subdivided into a dorsal magnocellular division (Bmc), a ventral parvicellular division (Bpc), and an intermediate subdivision (Bi). These cellular divisions are based on size and density of glutamatergic pyramidal neurons. Since glutamatergic projection neurons are specialized at the molecular level, it is possible that glutamatergic neurons in the basal nucleus exhibit distinct transcriptional profiles that encode their projection targets. In Aim 2A, I will use long-read single-nucleus RNA sequencing to characterize the transcriptional profiles of glutamatergic neurons in the macaque basal nucleus of the amygdala. I hypothesize that a gradient of excitatory neuron subtype-specific gene expression will be revealed, with unique glutamatergic neuron types present in the Bmc, Bi, and Bpc. In Aim 2B, an atlas of differentially express genes in the basal nucleus subdivisions will be validated with spatial transcriptomics (RNAScope).

While the sgACC and pgACC act as discrete functional and connectional hubs, preliminary data show that they receive a common input from the Bi, which may function to coordinate responses to salient social stimuli. I hypothesize that excitatory Bi neurons projecting to sgACC and pgACC will exhibit distinct transcriptional profiles compared to neurons that do not target the ACC. In Aim 3, information from transcriptomic studies (Aim 2) will be used to determine the molecular features of Bi-ACC projection neurons. Using cases from Aim 1, retrogradely labeled neurons in the basal nucleus of the amygdala projecting to sgACC and pgACC will be double labeled for fluorescent in situ hybridization (RNAScope). These results will then be integrated with the results of Aim 1 for a comprehensive analysis of sgACC and pgACC connectivity. Overall, this will provide insight on how the unique functions of sgACC and pgACC are established and coordinated to guide decision-making in the presence of salient social stimuli, and will be informative for understanding ACC dysregulation in psychiatric disorders.

 Dec 17, 2024 @ 9:00 a.m.
 Medical Center | K307 (3-6408)