MRCI: a flexible real-time dynamic clamp system for electrophysiology experiments

doi:10.1016/j.jneumeth.2003.08.002

Journal of Neuroscience Methods

Volume 132, Issue 2, 30 January 2004, Pages 109-123

https://doi.org/10.1016/j.jneumeth.2003.08.002 Get rights and content

Abstract

We present a real-time simulation system that enables modeled dynamical systems to interact with physical experimental systems, and is specifically aimed towards execution of the dynamic clamp protocol. Model reference current injection (MRCI) operates under Real-Time Linux (RT-Linux or RTL) and provides a simple equation-oriented language for describing dynamical system models. Features include scripting of commands to implement repeatable protocols, the ability to output pre-computed waveforms through any variable or parameter of the model, the means to conduct time measurements and assess the computational performance of the real-time system, and an installation program that installs the software and accompanying device drivers with minimal input from the user. Tested models operate as fast as 30 kHz, with actual maximum rates dependent on model complexity. We present sample models that exhibit the main features of the modeling language. Experiments demonstrate the abilities of the system by creating a hybrid network of real and simulated neurons, and playing a pre-defined synaptic waveform into a synaptic conductance variable. We conclude by introducing a waveform reconstruction technique that is useful for establishing the presence of significant experimental error in implementations of the dynamic clamp protocol.

Introduction

Voltage-clamp and current-injection techniques are common tools in cellular electrophysiology. Sharp et al. (1993) introduced the dynamic clamp, the technique of real-time simulation of ion-channel conductances and their associated ionic currents, which are injected into an excitable cell. This technique enjoys use in a variety of applications, including endowing individual neurons with artificial conductances (Sharp et al., 1993, Turrigiano et al., 1996, Hutcheon et al., 1996), creating artificial networks of real neurons (Sharp et al., 1996, Elson et al., 1998) or cardiac myocytes (Verheijck et al., 1998), and coupling computational cell models to in vitro cells (Renaud-LeMasson et al., 1993, LeMasson et al., 1995, Wilders et al., 1996, Kumar et al., 1995).

Past implementations of the dynamic clamp have relied on an approach where the simulation software is run on a desktop operating system such as Microsoft Windows. It has been demonstrated that such an approach introduces an excessive amount of cycle-to-cycle jitter, which significantly affects the speed and accuracy of the real-time computation required for the dynamic clamp technique (see Christini et al., 1999, for review). Recently, several laboratories have presented new techniques for implementing this protocol. Pinto et al. (2001) introduced a Windows-based approach that still does not eliminate the jitter problem, but does measure and account for the duration of each computational cycle. Christini et al. (1999) proposed the real-time operating system Real-Time Linux (commonly referred to as RT-Linux or RTL) as a software platform for applying real-time computing to biological experiments. All of the RTL variants have latencies in the microsecond range, where latency is defined as additional delay in the computational cycle, introduced by software and/or data acquisition operations. Both our lab (Butera et al., 2001) as well as others (Dorval et al., 2001), have implemented dynamic clamp techniques based upon RTL. Our system is called MRCI, which stands for Model Reference Current Injection. Performance evaluations of MRCI (Butera et al., 2001) found that our RTL-based dynamic clamp could reliably operate as fast as 50 kHz. However, this rate was achieved with an early version of MRCI that utilized a hand-coded C software implementation of a model. The model performance in recent versions is not as fast because of added features, potential inefficiencies in machine-generated C code produced by the MRCI translator, and a modular design that separates the model from the run-time system, as described in Appendix A.

While both our previous work (Butera et al., 2001) and Dorval et al. (2001) demonstrate that RTL is a viable real-time computing platform, implementing a computational model still required considerable technical skills, such as proficiency in the C programming language and familiarity with Linux kernel programming.

However, the scientist in the lab is most likely to be interested in the mathematical model and its properties, and does not want to worry about numerical algorithms or other details of computer programming. In the case of the dynamic clamp technique, the questions of selecting an integration algorithm and the particular software implementation are technical issues and do not belong to the problem domain. In this paper, we will describe and demonstrate the capabilities of MRCI, our dynamic clamp model specification and simulation environment. MRCI offers improvements in model specification, ease of installation, and built-in features compared to existing published dynamic clamp systems.

Beginning with Version 1.0.0, MRCI includes a translator for a general and flexible language which can be used for describing dynamical systems with ordinary differential equations; its data acquisition driver supports two analog input and two analog output channels, using the E-series PCI and ISA boards manufactured by National Instruments.

The interface to the MRCI computational engine has many facilities to aid the experimenter in understanding the dynamics of the simulated model. Some of these are the ability to perform data logging of any variable in the system; the scripting of command sequences for precise implementation of complex protocols; the ability to load waveform tables and “play” the waveform into any parameter or state variable; the real-time modification of any variable or parameter by the user, and on-the-fly switching from one model to another without restarting the system.

A technical description of the software architecture of MRCI is provided in Appendix A; more information is available in the user manual on the MRCI project’s web site, located at http://www.neuro.gatech.edu/mrci. The software that comprises the MRCI system is distributed under the terms of the General Public License of the Free Software Foundation.

Section snippets

Methods

The computational system consists of a Dell Optiplex (Dell Computer, Round Rock, TX) 500 MHz Pentium III. The PC contains a PCI-6052E multi-function data acquisition (DAQ) board (National Instruments, Austin, TX) with a BNC-2090 coaxial cable interface. This board has 16-bit analog input (dynamic range scalable from ±0.1 to ±10 V) and output (±10 V) channels. For the examples presented in this paper, the computer was interfaced to an Axoclamp 2-B (Axon Instruments, Foster City, CA)

Modeling language

This section briefly describes the MRCI modeling language; it is somewhat technical, and readers primarily interested in installing and running the system should skip to the next section. A more complete description of the modeling language is available in the user manual. The model equations to be evaluated and/or integrated comprise a dynamical system. The language used to describe these equations is a form of a structured mathematical description, which can be extended with constructs of the

User interaction

The user starts MRCI by invoking the MRCI start-up script and giving it the name of a model description file, which is compiled and loaded as a MRCI system module. The user interface to the MRCI application is a command-line interpreter interface which resembles the Unix shell: it prompts for commands and executes them until an end-of-line symbol is read (or the user issues an exit command); if called with a script file name, it reads and executes a script from that file.

The script language is

An example session

We will now proceed to describe an example work session with MRCI to supplement the information given in the previous section. We will use the model shown below, and manipulate its quantities while the computational engine is running. This model is identical in functionality to the single conductance model shown earlier, except that table lookups and the multiply–add–update algorithm are used for computational efficiency. For this session, we assume that MRCI is run from the user’s home

Experimental examples

The MRCI User’s Manual has a detailed tutorial, ranging from bench-top testing of an isolated system to actual experiments. The tutorial is based on a Morris–Lecar excitable cell model (Morris and Lecar, 1981), which receives input from up to two external cells. A commented listing of this model is presented in Appendix B. Any deviations from the parameters listed in the file are specified in the figure captions.

For our first example, we connected the MRCI system to motoneuron B5 of the buccal

Performance analysis

There are two major issues that must be considered when analyzing the performance of the MRCI system: sources of error associated with measurement, acquisition, and particular aspects of the electrophysiology experiment (such as DC offset and bridge balance), and the computational performance of the MRCI software.

Summary and future work

We have presented a powerful and easy to use system for implementing interactive models of real-time dynamical systems. The installation script and model specification language help overcome two of the major obstacles that dynamic clamp users have found when trying to use Linux-based dynamic clamp systems (Dorval et al., 2001, Butera et al., 2001): ease of installation and ease of model specification. Future work includes, in order of priority:

1.
Porting MRCI to the RTAI real-time Linux extension (

Acknowledgements

The authors wish to thank Simeon Sessley for patience with testing new versions of the MRCI software. The MRCI modeling language is derived from the modeling language used by the Gnans program, which was written by Bengt Mårtensson at the Institute for Dynamical Systems. Some source code is shared between the translators used by MRCI and Gnans. This project is supported by grants from the James S. McDonnell Foundation and National Science Foundation (DBI-998774) to RJB and a Georgia Tech

References (22)

T.A Kinard et al.
Modulation of the bursting properties of single mouse pancreatic β-cells by artifical conductances
Biophys. J.
(1999)
G LeMasson et al.
From conductances to neural network properties: analysis of simple circuits using the hybrid network method
Prog. Biophys. Molec. Biol.
(1995)
C Morris et al.
Voltage oscillations in the barnacle giant muscle fiber
Biophys. J.
(1981)
R Pinto et al.
Extended dynamic clamp: controlling up to four neurons using a single desktop computer and interface
J. Neurosci. Methods
(2001)
H.P Robinson et al.
Injection of digitally synthesized synaptic conductance transients to measure the integrative properties of neurons
J. Neurosci. Methods
(1993)
R Wilders et al.
Action potential conduction between a ventricular cell model and an isolated ventricular cell
Biophys. J.
(1996)
R.J Butera et al.
A methodology for achieving high-speed rates for artificial conductance injection in electrically excitable biological cells
IEEE Trans. Biomed. Eng.
(2001)
D.J Christini et al.
Practical real-time computing system for biomedical experiment interface
Ann. Biomed. Eng.
(1999)
A Dorval et al.
Real-time Linux dynamic clamp: a fast and flexible way to construct virtual ion channels in living cells
Ann. Biomed. Eng.
(2001)
R.C Elson et al.
Synchronous behavior of two coupled biological neurons
Phys. Rev. Lett.
(1998)

V Gauck et al.

The control of rate and timing of spikes in the deep cerebellar nuclei by inhibition

J. Neurosci.

(2000)

Cited by (38)

Efficient hardware implementation of the subthalamic nucleus–external globus pallidus oscillation system and its dynamics investigation
2017, Neural Networks
Citation Excerpt :
The proposed architecture allows for the applications in the closed-loop electrophysiological experiment to explore brain dynamics at the cellular and microcircuit levels. Electrophysiological experiment is one of the broadest and most successful applications which use intracellular experimental techniques including voltage clamp and dynamic clamp (Raikov, Preyer, & Butera, 2004; Williams & Mitchell, 2008). As demonstrated, the proposed HOSTN model can regenerate the biological dynamics accurately under a stimulation, which makes them especially attractive to closed-loop dynamic-clamping experiment.
Modeling and implementation of the nonlinear neural system with physiologically plausible dynamic behaviors are considerably meaningful in the field of computational neuroscience. This study introduces a novel hardware platform to investigate the dynamical behaviors within the nonlinear subthalamic nucleus–external globus pallidus system. In order to reduce the implementation complexities, a hardware-oriented conductance-based subthalamic nucleus (STN) model is presented, which can reproduce accurately the dynamical characteristics of biological conductance-based STN cells. The accuracy of the presented design is ensured by the investigation of the dynamical properties including bifurcation analysis and phase portraits. Hardware implementation on a field-programmable gate array (FPGA) demonstrates that the proposed digital system can mimic the relevant biological characteristics with higher performance, which means the resource cost is cut down and the computational efficiency is improved by introducing the multiplier-less techniques including novel “shift MUL” approach and piecewise linear approximation. The central pattern generator (CPG) coupled by the presented system is also investigated, which can be applied as an embedded intelligent system in the field of neuro-robotic engineering.
Cell-specific Dynamic Clamp analysis of the role of funny I<inf>f</inf> current in cardiac pacemaking
2016, Progress in Biophysics and Molecular Biology
Citation Excerpt :
RT-Linux, compared to Windows, has the advantage of running two processes, a real-time process (top priority, guaranteed timing) and the standard Linux operating system (low priority, interruptible by the real-time process), as set out in detail by Christini et al. (1999). The ‘Real-Time eXperiment Interface’ (RTXI, Lin et al., 2010; Ortega et al., 2014) was developed starting from RTLDC (Real-time Linux dynamic controller, Dorval et al., 2001) and MRCI (Model reference current injection, Raikov et al., 2004). RTXI is open-source and free; it uses a modular architecture, which allows the construction of complex experimental protocols using a combination of built-in, community-shared, and self-made modules.
We used the Dynamic Clamp technique for i) comparative validation of conflicting computational models of the hyperpolarization-activated funny current, I_f, and ii) quantification of the role of I_f in mediating autonomic modulation of heart rate. Experimental protocols based on the injection of a real-time recalculated synthetic I_f current in sinoatrial rabbit cells were developed.
Preliminary results of experiments mimicking the autonomic modulation of I_f demonstrated the need for a customization procedure to compensate for cellular heterogeneity. For this reason, we used a cell-specific approach, scaling the maximal conductance of the injected current based on the cell's spontaneous firing rate. The pacemaking rate, which was significantly reduced after application of Ivabradine, was restored by the injection of synthetic current based on the Severi-DiFrancesco formulation, while the injection of synthetic current based on the Maltsev-Lakatta formulation did not produce any significant variation. A positive virtual shift of the I_f activation curve, mimicking the Isoprenaline effects, led to a significant increase in pacemaking rate (+17.3 ± 6.7%, p < 0.01), although of lower magnitude than that induced by real Isoprenaline (+45.0 ± 26.1%). Similarly, a negative virtual shift of the activation curve significantly lowered the pacemaking rate (−11.8 ± 1.9%, p < 0.001), as did the application of real Acetylcholine (−20.5 ± 5.1%).
The Dynamic Clamp approach, applied to the I_f study in cardiomyocytes for the first time and rate-adapted to manage intercellular variability, indicated that: i) the quantitative description of the I_f current in the Severi-DiFrancesco model accurately reproduces the effects of the real current on rabbit sinoatrial cell pacemaking rate and ii) a significant portion (50–60%) of the physiological autonomic rate modulation is due to the shift of the I_f activation curve.
Real-time kinetic modeling of voltage-gated ion channels using dynamic clamp
2008, Biophysical Journal
Citation Excerpt :
Equally important is the advance we made in terms of software implementation. To date, most efforts in writing dynamic-clamp software have focused on real-time operating systems (28,29,48) or on real-time hardware (49), with few—and relatively slow—Windows implementations (50). The reason for adopting a real-time operating system, generally at the price of sacrificing a convenient graphical user interface, is the need for consistent real-time performance.
We propose what to our knowledge is a new technique for modeling the kinetics of voltage-gated ion channels in a functional context, in neurons or other excitable cells. The principle is to pharmacologically block the studied channel type, and to functionally replace it with dynamic clamp, on the basis of a computational model. Then, the parameters of the model are modified in real time (manually or automatically), with the objective of matching the dynamical behavior of the cell (e.g., action potential shape and spiking frequency), but also the transient and steady-state properties of the model (e.g., those derived from voltage-clamp recordings). Through this approach, one may find a model and parameter values that explain both the observed cellular dynamics and the biophysical properties of the channel. We extensively tested the method, focusing on Na_v models. Complex Markov models (10–12 states or more) could be accurately integrated in real time at >50 kHz using the transition probability matrix, but not the explicit Euler method. The practicality of the technique was tested with experiments in raphe pacemaker neurons. Through automated real-time fitting, a Hodgkin-Huxley model could be found that reproduced well the action potential shape and the spiking frequency. Adding a virtual axonal compartment with a high density of Na_v channels further improved the action potential shape. The computational procedure was implemented in the free QuB software, running under Microsoft Windows and featuring a friendly graphical user interface.
NeuReal: An interactive simulation system for implementing artificial dendrites and large hybrid networks
2008, Journal of Neuroscience Methods
The dynamic clamp is a technique which allows the introduction of artificial conductances into living cells. Up to now, this technique has been mainly used to add small numbers of ‘virtual’ ion channels to real cells or to construct small hybrid neuronal circuits. In this paper we describe a prototype computer system, NeuReal, that extends the dynamic clamp technique to include (i) the attachment of artificial dendritic structures consisting of multiple compartments and (ii) the construction of large hybrid networks comprising several hundred biophysically realistic modelled neurons. NeuReal is a fully interactive system that runs on Windows XP, is written in a combination of C++ and assembler, and uses the Microsoft DirectX application programming interface (API) to achieve high-performance graphics. By using the sampling hardware-based representation of membrane potential at all stages of computation and by employing simple look-up tables, NeuReal can simulate over 1000 independent Hodgkin and Huxley type conductances in real-time on a modern personal computer (PC). In addition, whilst not being a hard real-time system, NeuReal still offers reliable performance and tolerable jitter levels up to an update rate of 50 kHz. A key feature of NeuReal is that rather than being a simple dedicated dynamic clamp, it operates as a fast simulation system within which neurons can be specified as either real or simulated. We demonstrate the power of NeuReal with several example experiments and argue that it provides an effective tool for examining various aspects of neuronal function.
A scriptable DSP-based system for dynamic conductance injection
2008, Journal of Neuroscience Methods
A variety of software and hardware systems have been developed to inject controlled electrical conductances into excitable cells, to investigate the physiological mechanisms of action potential generation. These systems face several challenges: the need to model complex conductances, including voltage-gated ion channels, synaptic conductances controlled by electrical models of entire cells or even networks of cells, to do so rapidly and stably, with precisely controlled update intervals of 20 μs or less, and to present an easy and flexible interface to the user, allowing new experiments to be designed and executed easily. In this paper I describe a new software system (SM-2) which is designed to meet these requirements, and which runs on the current generation of digital-signal-processing (DSP) analog input–output I/O boards, hosted in Windows PCs. Its key innovation is its configurability by simple user-written text scripts, or “scriptability”, which gives it a high flexibility of purpose, and allows non-programmers the capacity to rapidly design and use new Hodgkin–Huxley-type active conductances, conductances with arbitrary current–voltage relationships, Markov process conductance mechanisms with user-specified rate matrices, and hybrid networks of virtual cells. At the same time, the hardware platform allows this to be achieved with a fast and accurately timed input–computation–output cycle.
Effects of imperfect dynamic clamp: Computational and experimental results
2008, Journal of Neuroscience Methods
In the dynamic clamp technique, a typically nonlinear feedback system delivers electrical current to an excitable cell that represents the actions of “virtual” ion channels (e.g., channels that are gated by local membrane potential or by electrical activity in neighboring biological or virtual neurons). Since the conception of this technique, there have been a number of different implementations of dynamic clamp systems, each with differing levels of flexibility and performance. Embedded hardware-based systems typically offer feedback that is very fast and precisely timed, but these systems are often expensive and sometimes inflexible. PC-based systems, on the other hand, allow the user to write software that defines an arbitrarily complex feedback system, but real-time performance in PC-based systems can be deteriorated by imperfect real-time performance. Here, we systematically evaluate the performance requirements for artificial dynamic clamp knock-in of transient sodium and delayed rectifier potassium conductances. Specifically, we examine the effects of controller time step duration, differential equation integration method, jitter (variability in time step), and latency (the time lag from reading inputs to updating outputs). Each of these control system flaws is artificially introduced in both simulated and real dynamic clamp experiments. We demonstrate that each of these errors affect dynamic clamp accuracy in a way that depends on the time constants and stiffness of the differential equations being solved. In simulations, time steps above 0.2 ms lead to catastrophic alteration of spike shape, but the frequency–current relationship is much more robust. Latency (the part of the time step that occurs between measuring membrane potential and injecting re-calculated membrane current) is a crucial factor as well. Experimental data are substantially more sensitive to inaccuracies than simulated data.

View all citing articles on Scopus

¹: Ivan Raikov is an undergraduate student in the College of Computing.

²: Amanda Preyer is a graduate student in the School of Electrical and Computer Engineering.

³: Robert Butera is an Assistant Professor in the School of Electrical and Computer Engineering and School of Biomedical Engineering.

View full text

MRCI: a flexible real-time dynamic clamp system for electrophysiology experiments

Abstract

Introduction

Section snippets

Methods

Modeling language

User interaction

An example session

Experimental examples

Performance analysis

Summary and future work

Acknowledgements

Biophys. J.

Prog. Biophys. Molec. Biol.

Biophys. J.

J. Neurosci. Methods

J. Neurosci. Methods

Biophys. J.

A methodology for achieving high-speed rates for artificial conductance injection in electrically excitable biological cells

IEEE Trans. Biomed. Eng.

Practical real-time computing system for biomedical experiment interface

Ann. Biomed. Eng.

Real-time Linux dynamic clamp: a fast and flexible way to construct virtual ion channels in living cells

Ann. Biomed. Eng.

Synchronous behavior of two coupled biological neurons

Phys. Rev. Lett.

The control of rate and timing of spikes in the deep cerebellar nuclei by inhibition

J. Neurosci.