Adam Fitchett, Jason D. Fabbri, Yaoxing Hu, Justin Cange, Karolina Kozeniauskaite, Kenneth Shepard, David S. Holder, and Kirill Aristovich, “Imaging Circuit Activity in the Rat Brain with Fast Neural EIT and Depth Arrays,” 2023 11th International IEEE/EMBS Conference on Neural Engineering (NER), Baltimore, MD, USA, 2023, pp. 1-4.

Few techniques are specialized for neuroscience at the “mesoscopic” level of neural circuits. Fast neural electrical impedance tomography (fnEIT) is a novel imaging technique that offers affordability, portability, and high spatial (∼100 μm) and temporal (~1 ms) resolution. fnEIT with depth arrays offers the opportunity to study the dynamics of circuits in the brains of animal models. However, current depth array geometries are not optimized for this imaging modality. They feature small, closely packed electrodes with high impedance that do not provide sufficient SNR for high resolution EIT image reconstruction. They also have a highly limited range. It is necessary to develop depth arrays suitable for fnEIT and evaluate their performance in a representative setting for circuit neuroscience. In this study, we optimized the geometry of depth arrays for fnEIT, and then investigated the prospects of imaging thalamocortical circuit activity in the rat brain. Optimization was consistent with the hypothesis that small, closely spaced electrodes were not suitable for fnEIT. In vivo experiments with the optimized geometry then showed that fnEIT can image thalamocortical circuit activity at a high enough resolution to see the activity propagating from specific thalamic nuclei to specific regions of the somatosensory cortex. This bodes well for fnEIT’s potential as a technique for circuit neuroscience.

Sukjin S. Jang, Sarah Dubnik, Jason Hon, Björn Hellenkamp, David G. Lynall, Kenneth L. Shepard, Colin Nuckolls and Ruben L. Gonzalez, Jr. Characterizing the Conformational Free-Energy Landscape of RNA Stem-Loops Using Single-Molecule Field-Effect Transistors. December 22, 2022 J. Am. Chem. Soc. 2023, 145, 1, 402–412

We have developed and used single-molecule field-effect transistors (smFETs) to characterize the conformational free-energy landscape of RNA stem-loops. Stem-loops are one of the most common RNA structural motifs and serve as building blocks for the formation of complex RNA structures. Given their prevalence and integral role in RNA folding, the kinetics of stem-loop (un)folding has been extensively characterized using both experimental and computational approaches. Interestingly, these studies have reported vastly disparate timescales of (un)folding, which has been interpreted as evidence that (un)folding of even simple stem-loops occurs on a highly rugged conformational energy landscape. Because smFETs do not rely on fluorophore reporters of conformation or mechanical (un)folding forces, they provide a unique approach that has allowed us to directly monitor tens of thousands of (un)folding events of individual stem-loops at a 200 μs time resolution. Our results show that under our experimental conditions, stem-loops (un)fold over a 1–200 ms timescale during which they transition between ensembles of unfolded and folded conformations, the latter of which is composed of at least two sub-populations. The 1–200 ms timescale of (un)folding we observe here indicates that smFETs report on complete (un)folding trajectories in which unfolded conformations of the RNA spend long periods of time wandering the free-energy landscape before sampling one of several misfolded conformations or the natively folded conformation. Our findings highlight the extremely rugged landscape on which even the simplest RNA structural elements fold and demonstrate that smFETs are a unique and powerful approach for characterizing the conformational free-energy of RNA.

Sabina Hillebrandt, Hang, Adriaan J. Taal, Henry Overhauser,Kenneth L. Shepard, and Malte C. Gathe High-Density Integration of Ultrabright OLEDs on a Miniaturized Needle-Shaped CMOS Backplane. Advanced Materials, July 20, 2023

Direct deposition of organic light-emitting diodes (OLEDs) on silicon-based complementary metal–oxide–semiconductor (CMOS) chips has enabled self-emissive microdisplays with high resolution and fill-factor. Emerging applications of OLEDs in augmented and virtual reality (AR/VR) displays and in biomedical applications, e.g., as brain implants for cell-specific light delivery in optogenetics, require light intensities orders of magnitude above those found in traditional displays. Further requirements often include a microscopic device footprint, a specific shape and ultrastable passivation, e.g., to ensure biocompatibility and minimal invasiveness of OLED-based implants. In this work, up to 1024 ultrabright, microscopic OLEDs are deposited directly on needle-shaped CMOS chips. Transmission electron microscopy and energy-dispersive X-ray spectroscopy are performed on the foundry-provided aluminum contact pads of the CMOS chips to guide a systematic optimization of the contacts. Plasma treatment and implementation of silver interlayers lead to ohmic contact conditions and thus facilitate direct vacuum deposition of orange- and blue-emitting OLED stacks leading to micrometer-sized pixels on the chips. The electronics in each needle allow each pixel to switch individually. The OLED pixels generate a mean optical power density of 0.25 mW mm−2, corresponding to >40 000 cd m−2, well above the requirement for daylight AR applications and optogenetic single-unit activation in the brain.

Nanyu Zeng, Taesung Jung, Mohit Sharma, Guy Eichler, Jason Fabbri, R. James Cotton, Eleonora Spinazzi, Brett Youngerman, Luca Carloni and Kenneth L. Shepard A Wireless, Mechanically Flexible, 25μμm-Thick, 65,536-Channel Subdural Surface Recording and Stimulating Microelectrode Array with Integrated Antennas. 2023 IEEE Symposium on VLSI Technology and Circuits (VLSI Technology and Circuits).

This paper presents a fully wireless microelectrode array (MEA) system-on-chip (SoC) with 65,536 electrodes for non-penetrative cortical recording and stimulation, featuring a total sensing area of 6.8mm×7.4mm with a 26.5μm×29μm electrode pitch. Sensing, data telemetry, and powering are monolithically integrated on a single chip, which is made mechanically flexible to conform to the surface of the brain by substrate removal to a total thickness of 25μm allowing it to be contained entirely in the subdural space under the skull.

Adriaan J. Taal, Ilke Uguz, , Sabina Hillebrandt, Chang-Ki Moon, Victoria Andino-Pavlovsky, Jaebin Choi, Changmin Keum, Karl Deisseroth, Malte C. Gather& Kenneth L. Shepard Optogenetic stimulation probes with single-neuron resolution based on organic LEDs monolithically integrated on CMOS. Nature Electronics, August 12, 2023.

The use of optogenetic stimulation to evoke neuronal activity in targeted neural populations—enabled by opsins with fast kinetics, high sensitivity and cell-type and subcellular specificity—is a powerful tool in neuroscience. However, to interface with the opsins, deep-brain light delivery systems are required that match the scale of the spatial and temporal control offered by the molecular actuators. Here we show that organic light-emitting diodes can be combined with complementary metal–oxide–semiconductor technology to create bright, actively multiplexed emissive elements. We create implantable shanks in which 1,024 individually addressable organic light-emitting diode pixels with a 24.5 µm pitch are integrated with active complementary metal–oxide–semiconductor drive and control circuitry. This integration is enabled by controlled electrode conditioning, monolithic deposition of the organic light-emitting diodes and optimized thin-film encapsulation. The resulting probes can be used to access brain regions as deep as 5 mm and selectively activate individual neurons with millisecond-level precision in mice.

Elena Poggio, Francesca Vallese, Andreas J. W. Hartel, Travis J. Morgenstern, Scott A. Kanner, Oliver Rauh, Flavia Giamogante, Lucia Barazzuol, Kenneth L. Shepard, Henry M. Colecraft, Oliver Biggs Clarke, Marisa Brini and Tito Calì Perturbation of the host cell Ca2+ homeostasis and ER- mitochondria contact sites by the SARS-CoV-2 structural proteins E and M. Cell Death and Disease, April 29, 2023.

Coronavirus disease (COVID-19) is a contagious respiratory disease caused by the SARS-CoV-2 virus. The clinical phenotypes are variable, ranging from spontaneous recovery to serious illness and death. On March 2020, a global COVID-19 pandemic was declared by the World Health Organization (WHO). As of February 2023, almost 670 million cases and 6,8 million deaths have been confirmed worldwide. Coronaviruses, including SARS-CoV-2, contain a single-stranded RNA genome enclosed in a viral capsid consisting of four structural proteins: the nucleocapsid (N) protein, in the ribonucleoprotein core, the spike (S) protein, the envelope (E) protein, and the membrane (M) protein, embedded in the surface envelope. In particular, the E protein is a poorly characterized viroporin with high identity amongst all the β-coronaviruses (SARS-CoV-2, SARS-CoV, MERS-CoV, HCoV-OC43) and a low mutation rate. Here, we focused our attention on the study of SARS-CoV-2 E and M proteins, and we found a general perturbation of the host cell calcium (Ca2+) homeostasis and a selective rearrangement of the interorganelle contact sites. In vitro and in vivo biochemical analyses revealed that the binding of specific nanobodies to soluble regions of SARS-CoV-2 E protein reversed the observed phenotypes, suggesting that the E protein might be an important therapeutic candidate not only for vaccine development, but also for the clinical management of COVID designing drug regimens that, so far, are very limited.

Thierry Tambe, Jeff Zhang, Coleman Hooper, Tianyu Jia, Paul N. Whatmough, Joseph Zuckerman, Maico Cassel Dos Santos, Erik Jens Loscalzo, Davide Giri, Kenneth Shepard, Luca Carloni, Alexander Rush, David Brooks1, Gu-Yeon Wei1 22.9 A 12nm 18.1TFLOPs/W Sparse Transformer Processor with Entropy-Based Early Exit, Mixed-Precision Predication and Fine-Grained Power Management. March 23, 2023, IEEE International Solid- State Circuits Conference (ISSCC).

Large language models have substantially advanced nuance and context understanding in natural language processing (NLP), further fueling the growth of intelligent conversational interfaces and virtual assistants. However, their hefty computational and memory demands make them potentially expensive to deploy on cloudless edge platforms with strict latency and energy requirements. For example, an inference pass using the state-of-the-art BERT-base model must serially traverse through 12 computationally intensive transformer layers, each layer containing 12 parallel attention heads whose outputs concatenate to drive a large feed-forward network [1]. To reduce computation latency, several algorithmic optimizations have been proposed, e.g., a recent algorithm dynamically matches linguistic complexity with model sizes via entropy-based early exit [2]. Deploying such transformer models on edge platforms requires careful co-design and optimizations from algorithms to circuits, where energy consumption is a key design consideration.

Jake Rabinowitz, Andreas J. W. Hartel, Hannah Dayton, Jason D. Fabbri, Jeanyoung Jo, Lars E. P. Dietrich, and Kenneth L. Shepard Charge Mapping of Pseudomonas aeruginosa Using a Hopping Mode Scanning Ion Conductance Microscopy Technique. Anal. Chem., March 15, 2023.

Scanning ion conductance microscopy (SICM) is a topographic imaging technique capable of probing biological samples in electrolyte conditions. SICM enhancements have enabled surface charge detection based on voltage-dependent signals. Here, we show how the hopping mode SICM method (HP-SICM) can be used for rapid and minimally invasive surface charge mapping. We validate our method usingPseudomonas aeruginosaPA14 (PA) cells and observe a surface charge density of σPA = −2.0 ± 0.45 mC/m2 that is homogeneous within the ∼80 nm lateral scan resolution. This biological surface charge is detected from at least 1.7 μm above the membrane (395× the Debye length), and the long-range charge detection is attributed to electroosmotic amplification. We show that imaging with a nanobubble-plugged probe reduces perturbation of the underlying sample. We extend the technique to PA biofilms and observe a charge density exceeding −20 mC/m2. We use a solid-state calibration to quantify surface charge density and show that HP-SICM cannot be quantitatively described by a steady-state finite element model. This work contributes to the body of scanning probe methods that can uniquely contribute to microbiology and cellular biology.