Direct Interfacing of Neurons to Highly Integrated Microsystems

Abstract

The use of large high-density transducer arrays enables fundamentally new neuroscientific insights through enabling high-throughput monitoring of action potentials of larger neuronal networks (> 1000 neurons) over extended time to see effects of disturbances or developmental effects, and through facilitating detailed investigations of neuronal signaling characteristics at subcellular level, for example, the study of axonal signal propagation that has been largely inaccessible to established methods. Applications include research in neural diseases and pharmacology.

Introduction

To understand how functions and characteristics of neuronal networks arise from the concerted interactions of the involved neurons, it is necessary to have methods that allow for interacting with neuronal functional subunits and ensembles - somas, axons, dendrites, single neurons, as well as entire networks - at high spatiotemporal resolution and in real time [1–4]. Electrical recordings are among the oldest techniques [3–7].

Extracellular electrical recordings by means of microtransducer arrays complement well-established patch clamp techniques [6, 7] and optical or optogenetic techniques [8–11]. Commercially available standard MEAs (circuitless and passive) are an established technology for recording from networks of neurons [12–15]. However, due to their limited spatial resolution (pitch > 30 µm) and limited number of electrodes (< 300), such circuitless MEAs typically do neither allow for recording details of targeted individual neurons nor for monitoring large networks.

High-density MEAs (HD-MEAs) offer the potential to simultaneously record from a large number of sites at high spatiotemporal resolution and in different frequency bands (action-potential (AP) and local-field potential (LFP) band) to obtain information on the properties of individual cells or neurons and on the dynamics in neuronal networks. Each electrode can also be used to locally stimulate [4, 16–18]. Simultaneous recordings of multiple neurons are tedious to realize by patch-clamp, and the cells are only viable for a short time [19, 20]. Optical and optogenetic methods provide multi-unit recording and stimulation over extended times, yet still have limited temporal resolution, and it is difficult to detect or to stimulate single action potentials (APs) in individual cells due to stray light and light scattering [3, 9–11].

CMOS High-Density Systems

CMOS (complementary metal oxide semiconductor) technology-based planar, high-density microelectrode arrays (HD-MEAs) feature several thousands of transducers at densities of > 3’000 transducers per mm² [4, 18, 21–26] (Figure 1), which enables recordings from large neuronal networks at single-cell, or even subcellular-component resolution [27, 28]. The integrated systems feature tightly-spaced transducers and the respective addressing and read-out circuitry on a single chip [4, 21–26].

Figure 1.

CMOS realization of an open-gate field-effect transistor array. (a) Schematic cross section of a CMOS chip; the transistor gate is connected via several metal layers and vias to an electrode on top of the chip, which is covered by a dielectric material, such as high-k Ti-Zr oxide. Adapted from [21]. (b) Chip surface of a fabricated CMOS-based HD-MEA based on open-gate FETs. The much larger stimulation site completely surrounds the sensor site in the center [24].

Several different approaches relying on metal electrodes or open-gate field-effect transistors (FETs), as schematically shown in Fig. 1a, have been pursued [21, 22, 24, 26]. In the case of open-gate FET transducers, additional stimulation spots are needed and distributed between the electrodes. A sensor site and the surrounding stimulation site can be seen in Fig. 1b

Besides the comparably large, planar in-vitro systems, which are in the focus of this contribution, there are also CMOS-based needle-type probes for recording and stimulation in vivo that have been pioneered at the University of Michigan [3, 29–33] and adopted by other groups (Figure 2) [34–37]. Such electrode arrays are inserted into the living brain and facilitate advances in the understanding of the nervous system. Merged with onchip circuitry, signal processing, microfluidics, and wireless interfaces, they are forming the basis for a family of neural prostheses for the possible treatment of disorders, such as blindness, deafness, paralysis, severe epilepsy, and Parkinson’s disease [31].

Figure 2.

Very compact integrated neural recording microsystem frontend on an index finger featuring 64 electrodes: 4 electrodes per shank and 4 chips with 4 shanks each. At the right the cable connection can be seen [29, 30].

The use of CMOS helps to overcome the connectivity problem of how to interface thousands of tightly-spaced transducers through provision of addressing and multiplexing units [4, 18, 26]. The use of CMOS also helps to improve signal-to-noise characteristics, as signal conditioning is done on chip next to where the partially very small signals (< 10 µV) are generated. Using CMOS devices, local field potentials (LFPs), i.e., collective electrical activity of many neurons, have been recorded in brain slices, action potentials have been detected in dissociated neuronal cultures and brain slices, or various types of retinal ganglia have been identified [23, 38–43].

A disadvantage of CMOS chips is that silicon is not transparent to visible light in contrast to standard cell culture substrates used in biology. Additionally, the chip or its components can corrode upon operation and long-term exposure to liquids (salt water). Therefore, a good packaging solution is needed, on the one hand, to protect the chip against metabolic products and chemicals of the cell culture, and, on the other hand, to prevent the cells from being poisoned or disturbed by toxic materials released by the chip, such as the CMOS metals aluminum or copper that dissolve in saline solution.

In this contribution, we will demonstrate how highly integrated CMOS microelectrode array systems featuring several thousands of transducers (> 3’000 transducers per mm²) can be used to record from or stimulate potentially any individual neuron or subcellular compartment on a CMOS chip [4, 21, 26, 28].

As one example, we show in Fig. 3 a HD-MEA system featuring a sensing area of 3.85 × 2.10 mm² hosting 26’400 Pt electrodes of 7 µm diameter at a center-to-center pitch of 17.5 µm [26, 28]. The switch matrix allows for simultaneously routing user-configurable selections of electrodes to 1024 recording channels with 10-bit 20-kS/s A/D conversion and 32 stimulation units at the array periphery. The readout channels provide programmable bandwidth and gain up to 78.3 dB. The integrated noise voltage of the full readout chain is 2.4 μV_rms in the action-potential (AP) signal band (300 Hz–10 kHz). The switch matrix can be reprogrammed within 1.4 ms to adapt to the morphology of the biological sample. Possible configurations include single electrodes, sets of electrodes at points of interest, or multiple contiguous high-density blocks of up to 23×23 electrodes.

Figure 3.

CMOS-based bidirectional HD-MEA system. 26’400 metal electrodes (Pt) at the center of the chip (3.85 x 2.1 mm²) are surrounded by readout, stimulation, and addressing circuitry units (scale bar: 1 mm) [26].

Measurements and Results

With this system it was possible to record subcellular-resolution data in various preparations. Figure 4 shows the electrical signals of 3 neurons, recorded by the HD-MEA at full resolution and superimposed to a fluorescence image of the cell culture (MAP2-staining). The signals of three neurons, as obtained from the respective electrode sites (rounded white rectangles), are displayed in green (top left neuron), red (top right) and blue (bottom right). Figure 5 shows in the left panel a high-resolution zoom-in on a small colony of neurons in a network. From both figures it is evident that many electrodes always record from the same neuron and that every electrode simultaneously records activity of several neurons.

Figure 4.

Electrical signals of 3 neurons, recorded by the HD-MEA at full resolution and superimposed to a fluorescence image of the cell culture (MAP2 staining). Spike-triggered signal averages (50 trials) in green (top left neuron), red (top right) and blue (bottom right) for each electrode site (rounded white rectangles). Only signals exceeding 50 μV are displayed for clarity [28].

Figure 5.

(a): High-resolution zoom-in on a small colony of neurons in a network. Five identified neurons with overlapping electrical footprints were obtained from all 209 electrodes in the area. Black-yellow circles indicate the 3 electrodes providing best separation of the signals of these 5 neurons, the respective spike-triggered average waveforms are given at the right [28]. (b-c): Axonal signals across a branching point recorded by simultaneously using 841 electrodes. (b): “Electrical image” of the axonal signals (rat cortical neurons at DIV 18). (c): Waveforms of the action potentials traveling along the axon as recorded from the electrodes marked in red on the center map. Typical axonal signal characteristics are observed: triphasic, first-positive signal [28].

It was also possible to detect small signals of action potentials traveling along thin axons (~100 nm diameter) as shown in Figure 5 center and right panels [27, 28]. Axonal signals across a branching point were recorded by simultaneously using 841 electrodes. The action potential waveforms that have been recorded from the electrodes marked in red in Fig. 5 center panel are shown in the right panel. Typical axonal signal characteristics are observed: tri-phasic, first-positive signals [27, 28]. The grey lines represent individual signal traces, the black line the spike-triggered average signal.

Additionally, CMOS microtransducer arrays and integrated microsystems offer the capability to bidirectionally interact, also in closed loop and real time, with potentially every single neuron in a given neuronal network [17, 27, 44]. We investigated the relationship between stimulation location and reliability of neuronal responses to high-frequency threshold stimulations. To verify that classified signals matched neuronal morphology, we superimposed spike-triggered average footprints and micrographs of lipofected neurons (Fig. 6a) [44]. We produced excitability profiles and defined stimulation thresholds for the respective electrodes.

Figure 6.

Selective stimulation of single neurons [44]. (a) Spontaneous activity footprint (black traces) superimposed over the morphology of the respective neuron and its proximal compartments (colored in gray). Neuronal morphology was inferred from a microscopy image of the lipofected neuron. (b) Stimulation map superimposed over the footprint and neuron morphology (same neuron as in A). Site-specific stimulation thresholds are color-coded, and the electrodes in the center of the spontaneous activity footprint are numbered (1-7). (c) Excitability profiles for stimulation through the 7 center electrodes. Numbers correspond to those in (b) [44].

Neuron-wide stimulation over a range of voltages revealed the spatial distribution of sites with low thresholds for extracellular stimulation. The stimulation site featuring the lowest threshold typically was found within the center region of the neuron’s spike-triggered average footprint (Fig. 6). The stimulation threshold was defined as the minimal voltage that triggered an AP on the readout electrodes in 100% of the trials. Stimulation was applied at 4 Hz for voltages from ±10 to ±1,000 mV, with steps of ±10 mV. The stimulation frequency of 4Hz was selected as it is, on the one hand, not high enough to affect the reliability of neuronal responses, but, on the other hand, sufficiently high to allow for consecutive stimulation of hundreds of electrodes within 2 hours.

We found that stimulation of the most sensitive site provided selective single-cell neuronal activation in 96 % of the cases when voltages at threshold values were applied. “Selective” in this context means that only the targeted neurons were activated. We performed stimulation-triggered recordings to construct stimulation maps (example in Fig. 6b) and excitability profiles (example in Fig. 6c) for 52 different neurons in 27 preparations. For all neurons, 1-3 sites had steep excitability profiles with remarkably low activation thresholds (example in Fig. 6c: electrodes 1, 2, 3), and the most sensitive sites typically resided near the peak of the AP footprints (example in Fig. 6b, electrode 1). In general, activation thresholds of the most sensitive sites had values between ~±30 and ~±90 mV.

Outlook

Since neuronal signal transmission is (electro-)chemical, and since local fluctuations of ion concentrations, as well as anisotropic and inhomogeneous electrical conductivity of tissue and culture medium are critically influencing electric interaction with neurons, the next development step includes a multi-function high-density MEA system [45]. This monolithic system incorporates a large high-density electrode array of 2.43 × 4.48mm² including 59’760 electrodes at 13.5µm pitch (5'487 electrodes/mm²) to record from large preparations/networks at subcellular resolution. Moreover it includes different functional circuitry units: 2048 action-potential readout channels featuring low noise levels of 3.2 µV_RMS (300 Hz to 10 kHz) to retrieve, for example, tiny axonal signals, 32 local-field-potential readout channels, 16 dual-mode voltage/current stimulation units, 32 impedance units, 28 neurotransmitter measurement units, and 32 current recording units. Separate AP and LFP recording units allow for improving specific amplifier characteristics without increasing overall readout unit complexity.

The system architecture also made use of the switch-matrix approach, as the intention was to deliver every system function to any arbitrarily selectable electrode/electrode sets to perform, at the same time, stimulation and measurements of various cell or network parameters in parallel. It was and is not feasible to repeat the various functional circuitry units within each pixel.