Creating Virtual Ionic Conductances in Living Cells with LabVIEW Real-Time and PXI

Author(s):
Paul HM. Kullman - University of Pittsburgh School of Medicine
John P.. Horn - University of Pittsburgh School of Medicine

Industry:
Life Science, University/Education, Research

Products:
PXI/CompactPCI, LabVIEW

The Challenge:
Mimicking the behavior of ion channels with a method that requires deterministic execution of a control loop running faster than 20 kHz.

The Solution:
Running the control loop on an NI LabVIEW Real-Time PXI system with a multifunction I/O data acquisition board connected via Ethernet to a host application running on a Windows PC.

Ion Channel Research with Dynamic Clamping
Ion channels are a diverse group of membrane proteins that regulate numerous important physiological behaviors in all living cells (Hille, 2001). In neurons and muscle cells, these proteins function as essential mediators of electrical communication by enabling cells to generate action potentials and transduce the actions of sensory stimuli and neurotransmitters. Traditional channel manipulation tools include pharmacological and genetic techniques. The dynamic clamp method (Sharp et al., 1993) provides an alternative electrophysiological approach by injecting appropriately sized currents to mimic or mask the influence of channel proteins on cellular behavior, resulting in virtual channels.

Fast Real-Time Computing Requirements
In order for the dynamic clamp method to create virtual channels, it must continually execute a loop, which involves measuring the cellular membrane potential (Vm), calculating the current that should flow through the virtual channels (i), and generating an appropriate voltage output to the current command circuitry of the recording amplifier. The calculations follow Ohm’s law, i = g * (Vm – Erev), where Erev is the reversal potential for current flow through the channel and g is the ionic conductance of the channels being implemented. Although Erev is a constant that can be determined in advance, g is a variable that must be constantly updated, along with Vm. Depending on the conductance involved, g can be calculated in advance and read from a file, or calculated on a continuing basis by solving kinetic equations that describe its voltage-dependence and evolution over time. For the method to work properly, it is essential that the feedback loop operate in real time at a speed faster than the kinetics of conductance changes in the system. In practice, this means that implementing simple synaptic conductances requires loop speeds greater than 5 kHz, while voltage-dependent conductances can require speeds greater than 20 kHz. Obtaining fast deterministic performance with user-friendly operating systems based on Windows and UNIX (including Mac OS and Linux) is not possible because these systems employ an interrupt structure that cannot be readily disabled. The solution is a real-time operating system with dedicated capabilities.

Selecting National Instruments Real-Time Software and Hardware
As part of our research on synaptic integration in autonomic ganglia, we have developed a dynamic clamp system called G-clamp (Kullmann et al., 2004). We chose to build G-clamp on the National Instruments platform because of several key characteristics:

• Fast deterministic real-time performance
• Easy software and hardware integration
• High-level programming environment that does not require an engineering or computer science background
• Capability of developing an easy-to-learn user interface for our applications
• Program code that would be easy to share and upgrade
• Options for future platform upgrades
• Ability to record and process data, in addition to executing the real-time control loop

Overview of G-Clamp Software
In its present form, G-clamp can implement several voltage-dependent conductances, read conductance templates from data files in memory, and store data (Figure 2). The real-time loop runs reliably at speeds as fast as 43 kHz for the simplest tasks and at 20 kHz for more complex problems involving several voltage-dependent conductances (Kullmann et al., 2004). As part of the experimental software, we have also implemented several routines for automating measurements of steady-state I-V relations, synaptic strength and synaptic gain. The program code for G-clamp was developed under NIH grant NS 21065 and is freely available for research purposes.

Graphical LabVIEW Environment Facilitates Dynamic Clamp
Using an embedded PXI controller together with LabVIEW Real-Time, we implemented a dynamic clamp that is easy to set up and upgrade as faster hardware becomes available. The graphical nature of the LabVIEW programming environment facilitates sharing the software with other researchers and adapting existing programs to new experiments.

References
Hille, B. Ion Channels of Excitable Membranes. Sunderland, MA: Sinauer Associates, 2001.
Kullmann, P. H., Wheeler, D. W., Beacom, J. and Horn, J. P.. "Implementation of a Fast 16-Bit Dynamic Clamp Using LabVIEW-RT." Journal of Neurophysiology 91 (2004), 542-554.
Prinz, A. A., Abbott, L. F. and Marder, E. "The Dynamic Clamp Comes of Age." Trends in Neurosciences 27 (2004), 218-224.

For more information, contact:
Dr. John P. Horn
Dr. Paul H.M. Kullmann
Department of Neurobiology and Center for the Neural Basis of Cognition,
University of Pittsburgh School of Medicine, Pittsburgh, PA 15261
Tel: 412-648-9429
E-Mail: jph@pitt.edu; pkullman@pitt.edu

Browse All Case Studies »