Symposium EP05—Engineered Functional Multicellular Circuits, Devices and Systems

Micro- and nanoscale technologies are emerging as powerful tools for controlling the interaction between cells and their surroundings for biological studies, tissue engineering, and cell-based screening. In addition, hydrogel biomaterials have been increasingly used in various tissue engineering applications since they provide cells with a hydrated 3D microenvironment that mimics the native extracellular matrix. In our lab we have developed various approaches to merge microscale techniques with hydrogel biomaterials for directing stem cell differentiation and generating complex 3D tissues. In this talk, I will outline our work in controlling the cell-microenvironment interactions by using patterned hydrogels to direct the differentiation of stem cells. In addition, I will describe the fabrication and the use of microscale hydrogels for tissue engineering by using a `bottom-up' and a `top-down' approach. Top-down approaches for fabricating complex engineered tissues involve the use of miniaturization techniques to control cell-cell interactions or to recreate biomimetic microvascular networks within mesoscale hydrogels. Our group has also pioneered bottom-up approaches to generate tissues by the assembly of shape-controlled cell-laden microgels (i.e. tissue building blocks), that resemble functional tissue units. In this approach, microgels were fabricated and seeded with different cell types and induced to self-assemble to generate 3D tissue structures with controlled microarchitecture and cell-cell interactions.

Building new capabilities for the control of biological systems using embedded controllers made of biological parts to program system-level behavior holds tremendous promise to address a wide scope of both civilian and military needs. However, to date, the creation of artificial neuronal circuits that aim to perform specific functions has been proceeding at a rather slow pace due to a number of technical challenges. In order to hack the biological intelligence for behavioral and metabolic control, this work discusses the rational paths to move this field forward. Specifically, circuit design, simulation, and fabrication need to complement with new theory development to provide rational designs with optimal performance. We embody this approach through forward-engineering of simple analogs to innate neural circuits. As a first attempt, we choose to hack the Aplysia’s gill-withdrawal reflex circuit, a well-elucidated one composed of easily identifiable large neurons, so that the technical hurdles are minimized and we can focus on revealing the engineering principles and developing the engineering capability. Specifically, we identify and harvest these big neurons and reconstruct the circuit on a microelectrode array (MEA) by precisely placing and wiring the neurons according to their innate anatomical map in a photo-patterned biomaterial scaffold. The MEA is used to program long-term sensitization into this circuit and monitor its internal states. Ultimately, the advances in fundamental understanding and capabilities resulting from this project will create new opportunities for harnessing this promise of engineered biological systems for human use. The success and extension of this platform technology can have broad impacts on the advancement of medical devices, bio-robots, and human augmentation.

The fact that the most significant life-threatening diseases of our times such as Cardiovascular Diseases (CVD) or Cancer remains the number one killer for over a century suggests that, despite the advancements in science and medicine over the years, there is a huge gap in translating these scientific findings to clinical setting. One of the major reasons for this gap is pre-clinical research’s heavy dependence on young animal models despite the fact that aging is the biggest risk factor for these diseases. For example, the average age for first heart attack is 65.3 years for males and 71.8 years for females, and most breast cancers develop in a postmenopausal, aged mammary gland tissue microenvironment at age of 62. Yet, due to the logistical limitations, current pre-clinical research predominately relies on experimental animals with a human-equivalent age of less than 35 years, which does not faithfully replicate the clinically prevailing aged tissue microenvironment. With increasing appreciation of the role of the tissue microenvironment in regulating disease progression and the response to therapeutics, there is an urgent need to develop, optimize and validate a novel 3D culture systems that fully recapitulates the aged tissue microenvironment in order to reproducibly model natural disease progression.

Here we present the first human induced pluripotent stem cell (hiPSC)-derived cardiomyocyte (iCM)-based, in vitro aged myocardial tissue model as an alternative research platform. Within 4 months, iCMs go through accelerated senescence and show cellular characteristics of aging. Furthermore, the model tissues fabricated using these aged iCMs, with stiffness resembling that of aged human heart, show functional and pharmacological deterioration specific to aged myocardium. We characterized hiPSC-derived cardiomyocytes (iCMs) over a prolonged culture period and determined that maturation process reaches a limit around 55 days in culture and is followed by a functional deterioration at late stages of culture (after 100 days). We show, for the first time, an in vitro aged, human-origin cardiac tissue model that is both physiologically and pathologically relevant for studying CVD such as MI. We achieved this by incorporating iCMs at the late stages into matrices with aged human heart stiffness. We show that late iCMs at cellular level and the respective tissue models at tissue level behave similar to aged human hearts, showing age-related responses, including altered gene expression and proliferative properties, impaired functional properties and stress response, decreased sensitivity to therapeutics and lower long-term viability. Our novel tissue model with age-appropriate physiology and pathology presents a promising new platform for investigating CVD or other age-related diseases.

The integration of living cells with 3D printed soft scaffolds can enable the realization of biomachines for soft robotics and hyper-tissue engineering. Here, we will present our group’s recent efforts towards increasing force in miniature biological robots, and efforts towards integrating neural control in these biological machines. We have developed biological locomotive systems that show millinewton forces capable of driving locomotion at speeds above 0.5 mm s−1. The high force was achieved by combining a ring and strip muscle design to result tissues with increased muscle cell alignment, redesign of the scaffolds, and optimized processes for cell differentiation. We also present approaches for creating neural tissue from mouse embryonic stem cell derived motor neurons that could be created within a scaffold and then integrated with these biological robots towards neuronal control. We also study neuronal plasticity in these in-vitro networks by taking advantage of optogenetic mouse embryonic stem cells, in order to start stimulation as soon as the neural progenitor stage of differentiation. We also characterized the effects of stimulation during neural differentiation, by showing morphological changes in network formation and we were able to show that the modulations in network function could be traced back to alterations to gene expression in response to such stimulation. Such complex cellular systems will be a major challenge for the next decade and beyond, requiring knowledge from tissue engineering, synthetic biology, micro-fabrication and nanotechnology, systems biology, and developmental biology. As these “biological machines” increase in capabilities, exhibit emergent behavior, and potentially reveal the ability for self-assembly and self-repair, questions can arise about the ethical implications of this work. These devices could have potential applications in drug delivery, power generation, and other biomimetic systems.

Most biological systems evolve through multiple stages of hierarchy. Individual components at a given hierarchy interact with one another to arrive at the next hierarchy with increased complexity and functionality. For example, organs and their functions emerge through interactions and cross-talk between individual cells and extracellular matrix. The rules of interactions between living biological components at various hierarchies and spatio-temporal scales are only partially understood at best. Here, we first address the question: is there a critical distance between cells necessary for the crosstalk. We address the question by considering two cell types, fibroblasts and muscle cells (C2C12) independently. We mix fibroblasts and collagen-1 and form free discs. We find, consistent with literature, that the discs contract due to cell-ECM-cell interactions. However, compaction initiates only when the cell density exceeds a critical threshold value. We find a similar critical density for compaction in case of C2C12 cells. We differentiate the compacted C2C12-ECM tissue to myotubes when they generate force, both static and dynamic. We then plate a cluster of stem cell derived neurons in the vicinity of the muscle strip. These neurons interact with one another and form a neural network which fires in synchronous bursts. Some of these neurons are motor neurons. They innervate the muscle tissues and stimulate them in a synchronous fashion. The approach of the motor neurons towards muscle is also guided by the interactions between them possibly mediated by soluble factors. Thus, interactions between muscle cells led to myotubes, and interactions between neurons resulted in a neural network, each arriving at a hierarchy and functionality distinct from their individual components. Finally, cross-talk between myotubes and neural network resulted in a neuro-muscular unit with complex functionality involving both sensing and actuation. Such a unit is applied to develop a living swimmer at low Reynolds number.

Observing the activity of the entire nervous system as an animal responds to chemical and physical queues could reveal the fundamental principles of sensory information processing in dynamic networks of spiking neurons. Conveniently, millimeter-sized transparent organisms like C. elegans and Hydra can be manipulated on-chip using microfluidic technologies raising the prospect of high-throughput interrogation of the entire animal brain in well controlled environmental conditions. Here we describe how electrical, optical, and chemical interrogation capabilities can be integrated into scalable microfluidic devices that allow us to investigate neural information processing at the whole-brain level in living organisms. Experiments that exploit the on-chip technologies described can help reveal how adaptive networks of neurons provide robust control of animal behavior.

Since Hodgkin and Huxley‘s seminal work (1952) on the electrical activity of giant squid neurons, studies of electroactive cells and tissue have been carried out using a variety of recording techniques, including glass micropipette patch-clamp electrodes, voltage-sensitive dyes, multielectrode arrays (MEAs), and planar field-effect transistors (FETs). However, multisite, simultaneous, three-dimensional (3D), and long-term electrophysiological investigation of an engineered tissue has not been possible using these technologies. Specifically, (1) voltage/ion sensitive dyes are toxic to the cells and limited in volume (3D) measurements, (2) patch clamp technique is limited by its recording sites, (3) both patch clamp and voltage sensitive dyes are limited to few minutes to hours, thus do not allow long term electrophysiological investigation of the tissue state, (4) planar FETs and MEA use established microfabrication methods to allow multiplexed detection on a scale not possible with micropipette technology, both exhibit relatively large detection areas confined to 2D substrates that render both cellular and subcellular and 3D electrical recording immensely challenging. We report a self-rolling 3D biosensors array for multiplex, simultaneous electrical measurements from excitable tissues in 3D with subcellular spatial resolution and μsec temporal resolution. Our approach directly monitors the electrical activity of microscale tissues (μtissues, ca. 300μm) engineered from induced pluripotent stem cell (iPSC) derived cardiomyocytes.

Functional biological devices made from living cells have the advantage of much less foreign body responses than non-biological implants. To engineer such neurobiological circuits, research to date still stays on the level of random neural patterning without knowing the function of each neuron, whereas neurons are precisely wired and functionalized in intact animals. As a result, the missing recognition of both components’ functions and connections becomes a major obstacle for rational design of neurobiological circuits. Based on the finding that neurite growth can be guided by microfibers, we developed a patterning technique enabling aplysia neurons to be precisely wired with each other without cross-communication. We implement a microfiber scaffold using photolithographically fabricated SU-8 microfibers combined with selective surface coating of poly-l-lysine. Our results show that axons of aplysia neurons adhere to the side walls of microfibers and grow along to reach to adjacent target neurons. This simple and effective technique enables rational design and fabrication of functional neurobiological circuits as potential living neural implants.

All tissues are composed of lineage committed cells and an extracellular matrix (ECM) that is secreted and maintained by the surrounding cells. Stem cells are an important cell population for tissue maintenance, regeneration, and repair because they have the ability to replenish their own population, and to differentiate into a number of cell types to restore tissue function. Both intrinsic as well as extrinsic mechanisms have been shown to be involved with the regulation of stem cell selfrenewal and differentiation, and this interplay between intrinsic and extrinsic cues has made it difficult to direct an entire population of stem cells toward a desired lineage. This talk will describe approaches for interfacing synthetic biology and materials science to engineer genetically interactive biomaterials and programmed ECMs to control intrinsic and extrinsic cellular events to control cell behavior. Altogether, these novel approaches can be used to improve cell fate outcomes, in addition to reprogramming terminally differentiated cells to function as novel therapeutic diagnostic and delivery vehicles.

The study of biological function in intact organisms and the development of targeted cellular therapeutics necessitate methods to image and control cellular function in vivo. Technologies such as fluorescent proteins and optogenetics serve this purpose in small specimens or surgically accessed tissues, but are limited by the poor penetration of light. In contrast, most non-invasive techniques such as ultrasound and magnetic resonance imaging – while based on energy forms that penetrate tissue effectively – are not effectively coupled to cellular function. Our work attempts to bridge this gap by engineering biomolecules with the appropriate physical properties to interact with magnetic fields and sound waves. In this talk, I will describe our recent development of biomolecular reporters and actuators for ultrasound. The reporters are based on a unique class of gas-filled protein nanostructures from buoyant photosynthetic microbes. These proteins produce nonlinear scattering of sound waves, enabling their detection with ultrasound. I will describe our recent progress in understanding the biophysical and acoustic properties of these biomolecules, engineering their mechanics and targeting at the genetic level and expressing them heterologously as acoustic reporter genes. Our actuators are based on temperature-dependent transcriptional repressors, which provide switch-like control of bacterial gene expression in response to small changes in temperature. This allows us to use focused ultrasound to remote-control engineered bacteria in vivo. Finally, I will describe how we have used ultrasound in combination with viral vectors and engineered receptors to provide spatially and cell-type specific non-invasive control over neural activity.

The design of nanoscale reservoirs for efficient delivery of drugs and nucleic acids to cells is an enduring challenge in materials for medicine. Drug-loaded liposomes have enabled numerous FDA approved systems for several diseases. However, many of the applications of molecular materials are often discontinued. This is because they fail to deliver cargo to cells that are hard to penetrate such as neurons and embryonic stem cells. One common culprit is not cell entry but rather entrapment of the delivery reservoir inside endosomes. In this work, we use cryo-EM, X-ray scattering, and cell culture studies to demonstrate that a new class of lipid-based particles loaded with siRNA (cuboplexes) have the ability to efficiently escape endosomes. Such particles have complex nanostructures but can be easily synthesized into sub-100 nm sizes with very little polydispersity by means of a microfluidics reactor.

Unlike traditional computers, neuromorphic systems aim to collocate processing and memory.
Achieving this integration of functions would transform architectures to enable more energy-efficient computers and offer material-level signal processing and memory. To date, neuromorphic hardware (mostly silicon-based) has sought to imitate the variable conductance of synaptic junctions between neurons (i.e., synaptic plasticity) via memory resistance, a property that loosely mimics the reconfigurable and storable states of conductivity in synapses. Specifically, memristors are two-terminal devices that can reconfigure their conductance based on prior electrical activity and are more energy efficient compared to transistor-based synaptic mimics. Even still, state-of-the-art, silicon-based memristors fail to capture the variety of mechanisms and dynamics, including the morphological growth and remodeling, that are exhibited by biological synapse, limiting their integration in functionally dense neural networks. In contrast, soft, biomimetic lipid membranes that mimic the composition, structure, and biophysical properties of synaptic membranes provide a new, bio-inspired approach for mimicking synaptic plasticity.

Here, we demonstrate that the voltage-driven geometric reconfigurability of an insulating lipid membrane can be used in combination with a voltage-independent pore-forming peptide, gramicidin, to create memristive devices with tunable short-term plasticity, which enables signal processing, learning, and memory. Our work builds on a recent study in which we demonstrated that insulative lipid membranes formed at the adjoining interface of lipid-coated aqueous droplets in oil exhibit memory capacitance stemming from voltage-driven modulation in the area and thickness of the hydrophobic core of the membrane. We show how introducing spontaneous, pore-forming, gramicidin peptides into the insulating lipid membrane creates conductive pathways—transforming the inherently memcapacitive device into a memristor that exhibits a pinched hysteretic current-voltage relationship. Experiments are performed to characterize the sensitivities and dynamics of membrane geometry changes and channel forming kinetics in response to voltage. Further, we developed a mathematical model that describes the physical states (membrane area, thickness, and channel kinetics) of the system involved in the memristive response. From both, we learn that the kinetics of channel formation by gramicidin peptides are extremely fast in comparison to the rates of dynamic modulations in bilayer area and thickness—suggesting that hysteresis and memory are dominated by the rates of change in membrane geometry. Our modular peptide doped bio-membrane also exhibits short-term synaptic-like facilitation and depression, emulating mechanisms of short-term plasticity exhibited by ion channels in pre- and post-synaptic neurons. Thus, our findings forecast an alternative paradigm of smart hardware using soft materials with simple composition and fabrication yet capable of providing inherent synaptic properties

It is through modifications of synaptic transmission (i.e., synaptic plasticity) between neurons that the brain is capable of adapting, learning, and modifying its functions in response to stimuli. Synaptic transmission can be either enhanced or depressed by brain activity, and these changes can last from milliseconds to a lifetime. They can occur in the pre-synaptic terminal (e.g., facilitation, PPF and PTP, depression, PPD, and augmentation) or in the post-synaptic terminal (e.g., long-term potentiation, LTP, and depression, LTD). Undoubtedly, each of these mechanisms is essential to the brain and plays major roles such as spatio-temporal filtering, spatial learning, emotional memory, and habituation. Therefore, neuromorphic computing systems that aim to emulate the brain’s parallel architecture and co-location of memory and processing, will require hardware that closely mimics synapses, including possibly multiple mechanisms of plasticity. To date, two-terminal metal-oxide-metal memory resistors (i.e., memristors) have been built, and, while they succeed at emulating mostly long-term memory, they bypass various synaptic functionalities and ignore the actual mechanisms governing synaptic plasticity. Moreover, they remain highly power-hungry, noisy, and rigid—unlike actual synapses. Conversely, neuromorphic hardware needs to be energy efficient, fault tolerant, and preferably soft. To this end, we have previously built a biomolecular memristive device that consisted of an insulating biomembrane doped with voltage-driven, pore-forming peptide, alamethicin. We demonstrated memristance through pinched hysteresis in the current-voltage plane.

Here, we discuss how multiple mechanisms of presynaptic plasticity can be achieved using a different peptide, known as Monazomycin. In response to supra-threshold potentials (>100 mV), monazomycin inserts into the insulating membrane and aggregates to form memristive ionic conductive pathways. When the voltage bias is dropped below the threshold, the peptides exit the membrane—which reclaims its insulating state. Our current-voltage results displayed diode-like pinched-hysteresis, confirming its memristive nature. Compared to alamethicin, monazomycin exhibited a significantly larger hysteresis and much slower formation and decay rates. In response to a stepwise increase in voltage from 10 mV to 150 mV the current exhibited an S-shape response with a time constant of around 20 s. Dropping from the voltage to 10 mV showed that the pores remained open for around 3 minutes before fully exiting the membrane. We demonstrate that in response to a train of pulses the system exhibited facilitation (PPF and PTP), depression, and augmentation. We discuss how the key mechanisms responsible of plasticity in our device depend on channel kinetics, which are similar to those found in biological synapses.

“Electromicrobiology” is an emerging field of study that investigates microbial electron exchange with external electrodes and microbial electrochemical functionalities. Future research efforts in this field will continue to revolutionize the fields of bioenergy, bioremediation, biosensing, biocomputing and biosynthesis. Significant advances in electromicrobiology have already occurred, led by impressive discoveries of “Exoelectrogens,” which are microorganisms capable of direct electron transfer to electrodes in the absence of exogenous redox mediators. By incorporating exoelectrogens into a bioelectrochemical cell, the combined biotic-abiotic system offers a solution for environmental sustainability by generating renewable bioelectricity with organic waste while producing value-added chemicals/biofuels; it can also perform many other functions, such as water desalination and toxicity detection. Despite its vast potential, however, the technology’s promise has not yet been translated as application (or market demand) because of a lack of in-depth understanding of the key mechanisms underlying electron harvesting from various microorganisms, and of the fundamental factors responsible for determining efficient electron exchange with external devices. These gaps relegate this promising technology to the status of a laboratory curiosity. Here, critical investigations of microbial electrochemical technology in the broad context of renewable bioenergy science and biosensing applications will be discussed.

With the increased interest in developing cost-effective and readily available energy sources, the research in the microbial fuel cell (MFC) field has gained an increased attention. Here, we report the use of graphene’s high electron mobility coupled with Geobacter sulfurreducens’ biocatalyst nature to increase the current response of a MFC. It is well-accepted that one of the limiting factors of current generation of a Geobacter based MFC is the electron transport process in the anodic chamber. We attempt to address this limitation by utilizing the enhanced electrical conductivity of graphene. To study the interface of graphene- Geobacter, Raman spectroscopy and scanning electron microscopy (SEM) were employed. The Raman characterization studies of the interface revealed an n-doping effect on graphene, with an approximate shift of 5 cm^-1. These results demonstrate that the graphene- Geobacter interface provides an approach to overcoming the challenges of conventional MFCs.

While it is clear that bacteria in the gut can impact brain function, many fundamental questions about the information processing along the brain-gut-axis remain. We developed a unique model for probing the earliest in-vivo brain-body interface, using a Xenopus laevis (frog) developmental model, revealing that the nascent brain, even before being fully formed, contains and sends instructive information for morphogenesis of remote tissues. This brain-body information interchange is mediated by ancient bioelectrical mechanisms, suggesting the possibility of a well-conserved computational code that could be shared by the microbiome. Our latest results demonstrated communication between brain and pathogenic bacteria, showing that the brain is involved in regulation of innate immune response to infection. Transcriptome data indicate pathways and sub-networks involved in brain-dependent and brain-independent regulation of the infection, proposing several novel targets for top-down control of information processing across different scales of biological organization. We are thus undertaking a computational analysis of the information content of the cross-talk between bacteria and neurons in vitro. Using techniques from neuroscience, biophysics, molecular biology, synthetic bioengineering, and information theory, we are developing and fabricating the first integrated electrical-optical brain-bacteria interface (BBI), a multi-site stimulation and recording platform specifically suited to extract information in real-time across highly diverse biological entities. The BBI is a microfluidic, optogenetic, and electrophysiological platform consisting of i) a cultured vertebrate brain or brain organoid, ii) a bacterial biofilm, and iii) a closed-loop software-controlled actuation and sensing system which extracts bioelectrical measurements from the brain and the bacteria under a series of real-time perturbation and uses mutual information analysis to determine how much each side knows about what is happening to the other. Currently, we are validating the health of the two biological components (long-term co-culture of neurons and bacteria), verifying that we can read and write information into the system. We are monitoring information transfer both optically, through fluorescent genetically-encoded ion reporters, and electrically, through the design and constructing of customized micro-electrode-array on microfluidic chambers. Chemical stimuli (drug screening) are provided via microfluidics. Next steps will characterize the communication channel and generate a quantitative model of cross-level communication. Finally, we will manipulate and re-write active communication between components to test and refine specific hypotheses. The extraction of informational content from signaling between prokaryote and metazoan systems will produce exciting new knowledge of both basic and applied (biomedical) impact, bridge a capabilities gap that establishes a new direction for synthetic bioengineering, and serve as an enabling technology and proof-of-principle analysis that many other labs will be able to use to addressing major challenges from different perspectives (from cellular/biological or synthetic constructs to multicellularity and evolution).