Real-time Control of Electrical Stimulation: A Novel Tool for Compound Screening and Electrophysiological Analysis

R. Hollis Whittington, Michael Q. Chen, L. Giovangrandi and Gregory T.A. Kovacs

Electrical Engineering Department, Integrated Circuits Lab, Stanford University

Objectives:

Whole cell assays for compound detection and drug screening have become an increasingly attractive approach that achieves a functional response without the traditional high costs of animal studies. Cardio-active pharmaceutical screening is particu­larly well-suited to cell-based electrical assays, as myocytes have the ability to generate their own electrical signature – the cardiac action potential. However, several subtleties compli­cate the design of systems relying on spontaneous action potentials, including variability of contraction properties. Electrical stimulation affords the investigator a means by which to elicit propagated action potentials that might otherwise be absent. Furthermore, stimulation itself confers certain advantages which are absent in spon­taneous cultures, including the ability to control the beat rate of the culture precisely (and with it rate-dependent biologic processes), the ability to assess excitability via stimulation threshold, and the ability to assess the refrac­tory period. Traditional techniques using manual or computer-controlled open-loop stimulators can be slow, accuracy is limited, transient effects are missed, and pacing itself may generate unobserved physiological responses.

 

This project encompasses an electrical stimulation system that utilizes a closed-loop algorithm to measure and track stimulation threshold in real time, extending the capabilities currently offered by stimulation. Recording of electric field potentials evoked by stimulation through electrodes of the same microelectrode array is made possible by stimulus artifact extraction algorithms that allow detection of APs which could otherwise be obscured. A novel closed-loop algorithm that is tolerant of the high quantization and low data rates inherent in this cell-based system was designed to enable feedback using the efficacy of stimulation, although feedback of any measurable physiological parameter is possible. The hybrid hardware/software stimulation system, along with temperature control circuitry and a custom fluidic perfusion system, comprise a desk­top hardware suite that is easily transportable and flexible, enabling convenient cell-level analysis in other laboratories and with other electrical cell types. The system is characterized using various physiologic stimuli, including temperature variation, ion channel block, electrolyte dis­turbance, and beta-adrenergic receptor stimulation and blockade.

 

Technical Approach:

• HL-1 cardiac myocytes are cultured on planar microelectrode arrays (MEAs) with silicon nitride passivation and platinum electrodes, configured in arrays for both recording and stimulation.
• Electrical pulses of a defined current amplitude, duration, and rate are passed through the stimulation electrodes, and if pulses are supra-threshold, they elicit an action potential that propagates from the site of stimulation outwards, across the recording array, where they are sensed.
• Evoked action potentials, but also artifacts from the stimulation pulse, are amplified, filtered, and digitized.
• Artifacts are removed from the evoked action potential data, and the efficacy of stimulation is determined as “capture fraction”. Capture fraction is the fraction of stimulation pulses, in a given time window, that evoke propagated action potentials.
• Based on the measured capture fraction and the target capture fraction, the stimulus parameters are adjusted automatically by a closed-loop algorithm in order to achieve the target capture fraction with the smallest possible increment in the controlled parameter (e.g., pulse amplitude), achieving high precision in most cases.

Figure 1: Flow diagram showing the major components of the closed-loop control system. Extracted parameter is capture fraction (CF), and the user specifies a target capture fraction (TCF) that the controller acts to achieve via adjustment of pulse parameters like amplitude and duration. Note the cardiomyocytes are represented as a component of the non-stationary plant.

 

Figure 2: Stimulation MEA Chip, composed of a glass MEA wire bonded to a commercial PCB carrier. Stimulation electrodes are connected via custom connectors at each end of the chip.	Figure 3: Micrograph of the stimulation electrode array with recording electrode array at the center of the chip. Large reference electrodes surround the stimulation electrode array on each end of the dice.

 

Results:

• Closed-loop stimulation system tracks the pulse amplitude, duration, or rate necessary to achieve a given target capture fraction from cultures of HL-1 atrial cardiomyocytes.
• Closed-circuit perfusion system was developed to allow seamless addition of drugs during continuous flow of culture medium with minimal artifacts due to change in temperature. This included design of a custom PDMS chamber for interfacing the MEA chip to the fluidic circuit.
• An artifact mitigation strategy was developed for real-time extraction of stimulation artifacts, enabling on-line determination of capture fraction.
• A novel closed-loop algorithm was developed that permitted precision control of pulse parameters to achieve a target capture fraction in the highly quantized environment of cardiomyocyte stimulation
• Responses to ion channel blockers (nifedipine and quinidine), beta-adrenergic agonists and antagonists (isoproterenol and propranolol), extracellular potassium concentration, and rapid temperature steps have been characterized.
• Further research and applications are currently being pursued.

 

Real-time response of stimulation threshold to application of 10 uM and 20 uM concentrations of quinidine, a sodium-channel blocker, at about 6 minutes. Responses are fairly repeatable and each dose results in a 2 – 2.5 uA increase in stimulation threshold. Responses of four separate cell cultures are aligned by subtracting the stimulation threshold immediately prior to drug addition.	Real-time response of stimulation threshold to the addition (10 minutes) and washout (30 minutes) of 3 mM KCl solution. The time courses of these four separate cultures are remarkably consistent, as is the net effect of KCl on the excitability of the culture. Responses are aligned by subtracting baseline threshold values.

Publications:

R. Hollis Whittington, Michael Q. Chen, Laurent Giovangrandi, Gregory T. A. Kovacs (2006). “Temporal Resolution of Stimulation Threshold: A Tool for Electrophysiologic Analysis” IEEE Engineering in Medicine and Biology Conference 2006, New York: in press.

R. Hollis Whittington, L. Giovangrandi and G. T. A. Kovacs (2005). “A Closed-Loop Electrical Stimulation System for Cardiac Cell Cultures.” IEEE Trans Biomed Eng 52(7): 1261-1270.

R. H. Whittington, K. H. Gilchrist, L. Giovangrandi and G. T. A. Kovacs (2003). “A Multi-Parameter, Feedback-Based Electrical Stimulation System for Cardiomyocyte Cultures.” Transducers '03: 983-986.

 

Funding Sources:

 

DARPA , NASA National Center for Space Biological Technologies

 

 

[ Home| Projects | People| Publications | Links| Pictures ]