A slim needle-shaped multiwire microelectrode for intracerebral recording

Journal of Neuroscience Methods, 40 (1991) 203-209 © 1991 Elsevier Science Publishers B.V. All rights reserved 0165-0270/91/$03.50

203

NSM01311

A slim needle-shaped multiwire microelectrode for intracerebral recording T. J e l l e m a a n d J . A . W . M . W e i j n e n

Department of Social Sciences, PhysiologicalPsycholology Section, Tilburg UniL,ersity,5000 LE Tilburg (The Netherlands) (Received 24 January 1991) (Revised version received 24 August 1991) (Accepted 2 September 1991)

Key words:

C u r r e n t source density analysis; Electrophysiological recording; F i e l d potentials; M u l t i c h a n n e l microelectrode; Multiwire electrode; S o m a t o s e n s o r y cortex; ( R a t )

The construction of a needle-shaped multiwire microelectrode is described. It can be made with simple mechanical tools. The presented electrode assembly consists of 12 insulated nichrome wires (core diameter 25 p.m) which are embedded in epoxylite resin. The straight-cut wire tips are aligned lengthwise and have a relative spacing of 150 /xm. Outer dimensions vary from 100 x 180 /xm at the level of the 1st electrode channel, to 100 x 100 /zm at the level of the 12th channel at the tip. The configuration of this electrode was determined by its application: the laminar analysis of evoked potentials in the cortex of the rat. However, the number of channels, the diameter of the (nichrome) wire which determines the surface area of these channels, and the channel spacing can be easily adjusted during construction to meet other experimental requirements, such as the recording of single-unit activity. The electrode which is composed of biocompatible materials is suited for the study of field potentials and multiple-unit activity, in both acute and chronic experiments, and can be used repeatedly. To demonstrate the performance of the electrode assembly, a depth profile of field potentials is presented, accompanied by the corresponding current source density distribution. The potentials were recorded in the somatosensory cortex of the rat following stimulation of the median nerve under ketamine anesthesia.

Introduction E l e c t r o d e arrays are powerful tools in electrophysiological research involving the r e c o r d i n g of s p a t i o t e m p o r a l p a t t e r n s of n e u r o n a l activity. T h e s i m u l t a n e o u s r e c o r d i n g of p o t e n t i a l s at different points in the b r a i n that c a n be o b t a i n e d with such a n assembly has c e r t a i n distinct a d v a n t a g e s over successive recordings with a single electrode that has to be m o v e d to new locations. This latter

Correspondence: T. Jellema, Department of Social Sciences, Physiological PsychologicalPsychologySection, P-602, Tilburg University, P.O. Box 90153, 5000 LE Tilburg, The Netherlands.

m e t h o d is time c o n s u m i n g a n d is susceptible to c h a n g e s in external c o n d i t i o n s a n d in the i n t e r n a l state of the animal, which are particularly difficult to control in c h r o n i c studies. S i m u l t a n e o u s recording of b r a i n p o t e n t i a l s offers - c o m p a r e d to successive m e a s u r e m e n t s - a m a r k e d reduction in variability of data over time a n d a great increase of efficiency in data collection. It is of crucial i m p o r t a n c e w h e n n o n - r e c u r r e n t events are studied. U n d e r c e r t a i n c o n d i t i o n s a 1 - d i m e n s i o n a l curr e n t source density (CSD) analysis can be perf o r m e d u p o n field p o t e n t i a l s which are r e c o r d e d at n e i g h b b r i n g locations to reveal the locally gene r a t e d c u r r e n t sources a n d sinks (Nicholson a n d

204

Freeman, 1975; Mitzdorf, 1985). Simultaneous recording of field potentials will, compared to successive recording, greatly enhance the accuracy of such a CSD analysis (Rappelsberger et al., 1981). There are basically 2 different kinds of multichannel microelectrodes which are used in studies requiring the simultaneous recording of signals at different locations: multiwire electrodes and printed-circuit electrodes. Multiwire electrodes, also called bundle electrodes, consist of individual wires glued together in such a way that their cut endings form an array of electrode channels along a straight line at predetermined equal distances. There are arrays in which the electrode channels reach out like the teeth of a comb (e.g., Verloop and Holsheimer, 1984), and arrays in which these channels are positioned above each other on a needle-like carrier (e.g., Barna et al., 1981). As a rule bundle electrodes are handmade at the laboratory of the experimenter. They are usually relatively bulky and frequently lack precision in the spacing of the electrode channels. Printed-circuit electrodes overcome these 2 disadvantages, but at the cost of an advanced lithographic technology. This technology will not be available in most neurophysiological laboratories. Their use in chronic studies has been questioned (Eichenbaum and Kuperstein, 1986). Excellent reviews of both types of multichannel microelectrodes have been published (Pickard, 1979; Kriiger, 1983; Eichenbaum and Kuperstein, 1986; Prohaska et al., 1986). In our research program we investigate laminar patterns of event-related cortical activity. To this end we record field potentials and multipleunit activity, in both acute and chronic preparations. To enable us to apply a 1-dimensional CSD analysis, we need measurements of field potentials along an axis perpendicular to the layers of the lissencephalic neocortex of the rat. For this purpose we decided to develop a slim needleshaped multiwire microelectrode, with a channel spacing of 150 ~m. The main practical problems in the manual construction of such an electrode are: (1) controlling the spacing of the channels of the microelectrode and (2) limiting the outer di-

mensions of the electrode assembly to minimize tissue damage while maintaining strength and electrical insulation.

Materials and methods

Most of the manipulations described below are performed under visual guidance with the help of a dissecting microscope. Nichrome wire, insulated with Formvar, is used. From this wire (outer diameter: 37 ixm, core diameter: 25 izm; Clark Electromedical Instruments) t2 pieces are cut of about 10 cm each. A piece of teflon tubing (Habia Benelux) measuring approximately 15 mm (inner diameter: 300 /zm; outer diameter: 760 /zm) is split lengthwise into 2 identical halves. One half is used as a temporary substrate for fixing the electrode wires. The inner dimension of this length of teflon can be reduced to about 2 0 0 / z m by compressing the lateral walls. With a fine metal needle which is attached to a micromanipulator (David Kopf Instruments), 12 tiny holes are pierced in the teflon tubing from the outer side at equidistant points (150 /zm apart). The holes are made in the middle section of the tubing, in a straight line along the long axis. The metal needle should have a tip diameter of 40 /zm over a length of at least 250 /~m, slightly longer than the thickness of the wall of the teflon tubing. (It can be produced from a hypodermic needle by grinding the tip carefully.) Next, the piece of teflon tubing is put on one of its sides on a mounting plate and is fixed in this position with tape. The 12 nichrome wires are put, one by one, from the outer side of the tubing through the punctured holes and bent at a right angle so that they fit in the hollow part of the teflon tubing (Fig. 1A). The individual wires are straightened and subsequently attached on either side to the mounting plate with heat-resistant tape. Each wire is marked with a channel identification number. Now that the wires are brought into a fixed position, they can be glued together. The teflon tubing is used as a casting mould. Epoxylite resin

205 (Clark Electromedical Instruments) is evenly spread at room t e m p e r a t u r e in the cavity of the tubing, just enough to e m b e d the wires within the mould. One should avoid trapping air bubbles in the resin. After curing, for 30 min at 100°C to drive off the solvents and for another 30 min at 180°C to complete the curing process, the teflon tubing is removed (Fig. 1B), leaving the wires e m b e d d e d in the resin. Teflon is temperature resistant up to 250°C and does not stick to the epoxylite resin. The electrode wires that protrude from the cast at the recording end are removed by bending them back and forth until they break at the point where they emerge from the cast. It is possible to improve the shape of the cast by tapering the tip

and cutting off redundant epoxylite resin along the shaft w i t h a scalpel blade (Fig. 1D). The other ends of the nichrome wires are soldered to micro-miniature connectors (Amphenol, 221 series). Soldering is done in 2 stages. First, the nichrome wire ends are tinned with SnAg5 solder, under application of a soldering fluid that is suitable for stainless steel. Subsequently, the wires can be fixed into the gold-plated connectors with ordinary multicore solder. The connectors are pushed into a block made of connector strips (Amphenol). The same block also holds the connectors for reference and grounding purposes, and in addition a piece of metal tubing cut from a hypodermic needle that will hold the electrode assembly. The strip of cured epoxylite 1 O0 I.lm 180 ~.lm

.-"

C

a

1 2 3 4 5 6 7 8 9 ~ 1 2

Fig. 1. A and B: 2 stages in the construction of the 12-channel multiwire electrode. In (A) the first 2 nichrome wires have been put through punctured holes in the teflon mould and bent subsequently at a right angle. (B) shows the removal of the teflon mould, after curing of the epoxy resin which is employed for embedding the electrode wires. Numbers 1-12 indicate individual wires. C: overall view of the completed electrode assembly;(a) and (b) mark extra leads for reference and grounding puposes. D: enlarged view of the encircled part of (C). This part of the electrode assembly contains the exposed tips of the nichrome wires: left, side view; right, frontal view with the 12 electrode channels (black dots) which are spaced by 150.p,m.

206

resin which contains the electrode channels i~ glued to this piece of metal tubing. The whole assembly is strengthened and insulated at the same time, by applying fast-setting 2-component epoxy adhesive (Fig. 1C).

Evaluation of the electrode

Shape and dimensions The maximal dimensions of the needle electrode contacting brain tissue were 100 × 180/zm. These proportions decreased linearly from the level of the first electrode channel to 100 x 100 /zm at the tip position of the 12th channel (Fig. 1D and Fig. 2). The spacing of the electrode channels was verified with a microscope that was equipped with a calibrated micrometer. The accuracy of this method was approximately 2 p.m. Three electrode assemblies were investigated. The mean interchannel distance was 149.5 _+ 14 /zm (mean + SD). Accurate measurement of the channel spacing, and use of these values in the CSD calculation, is essential. Even small variations in the spacing of the electrode channels can yield a considerable effect upon the CSD distribution. Repeated use of the electrode assembly in up to 8 acute experiments, or chronic implantation for up to 2 weeks, did not affect the channel spacing. The physical appearance of the surface of the electrode channels was quite regular, as revealed by light microscopy (Fig. 2). Breaking off the electrode wires by bending them back and forth resulted in a smooth fracture surface, flush with the surface of the electrode assembly. Cured epoxylite resin still possesses a certain flexibility; this property reduces the chance of accidentally damaging the electrode.

Recording characteristics The impedance of the individual channels, measured with an electrode impedance meter (Corti Electronik Labor) at 1 kHz in a 0.9% NaCI solution, ranged from 0.5 to 1.5 MY/. As the output leads of the electrode channels were directly coupled to a field-effect transistor (FET) headstage (Weijnen and Chehade, 1987), differ-

Fig. 2. Photomicrograph of the needle-shaped multwire my croelectrode. The bar indicates 200/~m (see also Fig. I D).

ences in impedance among the channels are relatively unimportant compared to the very high input impedance of these transistors. Cross-talk between channels was measured by applying a rectangular pulse with an amplitude of 0.5 mV and a duration of 10 ms to individual channels, at a point between the electrode and the attached F E T headstage. The electrode as-

207

easily produced. Passing an anodal current of 5 p,A through a nichrome electrode wire for 25 s produces a spherical lesion with a diameter of about 150/zm (Nissl stain). However, this current causes some erosion of the uninsulated end of the wire, due to dissolution of metal ions into the surrounding medium. It appeared that passing a lesioning current through the same electrode channel in 8 successive experiments increased its impedance 2-4 times (measured at 1 kHz in saline). Selecting different electrode channels every time that marking is required is, therefore, to be recommended.

sembl~ itself was lowered in a grounded bath of 0.9% NaC1. Cross-talk could only be detected in a neighboring channel and did not exceed a value greater than 1.5% of the original signal. Within the frequency band of interest (3 Hz-3 kHz), the noise level of the recording system with attached electrode, which was immersed in a grounded 0.9% NaC1 solution, was 5/xV (eft).

Marking the electrode position in cerebral tissue Electrolytic lesions which can be used to mark the location of selected electrode channels are

F P t

o.smv

~

C S D

~. sink

I

II

&

III IV

2

3

4-

I I

I I

6

V ..

7

__At~l

!

I i

--

I I

8 I

I

I I

I I

l I

I I I

I I

I I

I 10 I

I 15 I

9

b

-~

10

Vl

!

I I

I

I

I I I I I I

I t I I I I

I I

I I

I I I I I 10 I

15 I

i

11

-

p~.-i

n

12

-.

-5

I

0 I

5 I

msec

20 I

25 I

30 I

-5 I

0 I

5 I

20 I

25 I

30 I

msec

Fig. 3. Left-hand side: averaged (n = 100) depth profile of field potentials, evoked by electrical stimulation of the median nerve (single pulses, 4.5 V, 0.3 ms, delivered at t = 0, see arrow) and recorded with the multiwire microelectrode placed into the somatosensory cortex of the rat. T h e positions of the electrode channels (1-12), as determined by histological means, are indicated relative to the layers of the cortex (I-VI). Right-hand side: 1-dimensional CSD analysis of the field potentials shown in the profile on the left.

208 After each experiment the electrode was cleaned with trypsin to remove proteins and was subsequently stored in a dry environment.

spatial derivatives, with respect to the z-axis (perpendicular to the cortical layers), as follows:

Example of the recording performance of the microelectrode

d(2-3) d¢1-2) ]m = --O'z l(d(l_2 ) q.. d(2_3) )

The performance of the multichannel electrode is demonstrated in Fig. 3. At the left-hand side, an averaged (n = 100) depth profile of somatosensory evoked potentials is shown. These field potentials were evoked by electrical stimulation (single monophasic pulses, 4.5 V, 0.3 ms) of the median nerve, just proximal to the carpal ligament, under ketamine anesthesia (initial dose 100 mg/kg, with supplements of 35 mg/kg when indicated). The 12-channel microelectrode had been positioned in that region of the somatosensory cortex of the rat where epicortical potentials reached a maximal amplitude upon median nerve stimulation. The angle of penetration was such that the electrode track was perpendicular to the layers of the cortex. A FET headstage linked the animal to the amplifiers. Bandwidth of the recorded signal ranged from 3.2 Hz to 3.2 kHz ( - 3 dB points). Analog-to-digital conversion of the field potentials was performed at a frequency of 14 kHz/channel. We see a positive wave complex in superficial cortical layer II (peak latency: 9.5 ms) which extends into layer III. A negative wave complex is present in all cortical layers, with a maximal amplitude in layer IV (peak latency: 10.5 ms). The amplitude of this wave diminishes slowly in layers V and VI, and it reaches its peak in layers III and II with gradually increasing latency. The resulting superficial negative wave, peaking at 15 ms in layer II, coincides with a positive wave in layers V and VI. This overall pattern of changing polarity, in time and space, which emerges from the depth profile, appears to be fairly characteristic for the early processing of specific afferent signals in the somatosensory cortex of the rat (see D i e t al., 1990). At the right-hand side of Fig. 3 a 1-dimensional CSD analysis of the field potentials is presented. The estimates of the net densities of current sources and sinks were derived from the field potentials by calculating the second-order

(b3 -- ~ 2

(~2 -- ~1

In this formula ~ , ~b2 and ~b3 represent field potentials which have been recorded with 3 consecutive electrode channels. I m is the estimated CSD at the location of channel 2, ~z is the tissue conductivity along the z-axis (assumed to be constant), do_2) and d(2_3) , are the actual distances between channels 1 and 2 and 2 and 3, respectively, as measured with a calibrated micrometer (see above). The CSD distribution reveals the basic pattern of extracellular synaptic currents, which to a large extent comprise the generators of the field potentials. It is dominated by a large sink complex in layers III and IV, with a maximal amplitude in layer III 10 ms after stimulation. This sink is flanked by sources in superficial and deeper layers: layers II and V, respectively. A smaller sink complex appears, with similar latency, in the deeper part of layer V and is flanked by sources in the upper part of layer V and in layer VI. Finally, a small sink develops in layer II (peak latency at about 15 ms), accompanied by a source in layer III. A full report on these experiments will be published elsewhere.

Discussion

The described method permits the construction of a multiwire microelectrode of minimal dimensions. Advanced techniques are not required. The electrode can be used repeatedly; although the marking of electrode positions in cerebral tissue by passing an anodal current through individual wires will cause some erosion of the exposed metal surface of the electrode channels, which results in an increase of the impedance. The control over the spacing of the electrode channels during construction is satisfactory and

209

the obtained spacing remains stable with repeated use of the microelectrode in different experiments. The actual distances between consecutive channels can be measured with high accuracy and must be taken into account during CSD calculation. The construction of the electrode assembly can be easily adjusted for application in chronic preparations. Shortening of the piece of metal tubing, which is used for mounting the strip containing the electrode channels, reduces the overall size of the assembly and makes it suitable for implantation. Our evoked potential studies in chronic experiments required a few days of testing; similar results were obtained over this period.

Acknowledgement The authors thank Dr. W.J. Wadman for his comments on the manuscript.

References Barna, J.S., Arezzo, J.C. and Vaughan, Jr., H.G. (1981) A new multielectrode array for the simultaneous recording of field potentials and unit activity, Electroenceph. Clin. Neurophysiol., 52: 494-496.

Di, S., Baumgartner, C. and Barth, D.S. (1990) Laminar analysis of extracellular field potentials in rat vibrissa/ barrel cortex, J. Neurophysiol., 63: 832-840. Eichenbaum, H. and Kuperstein, M. (1986) Extracellular neural recording with multichannel microelectrodes, J. Electrophysiol. Tech., 13: 189-209. Kriiger, J.A. (1983) Simultaneous individual recordings from many cerebral neurons: techniques and results, Rev. Physiol. Biochem. Pharmacol., 98: 177-233. Mitzdorf, U. (1985) Current source-density method and application in cat cerebral cortex: investigation of evoked potentials and EEG phenomena, Physiol. Rev., 65: 37-100. Nicholson, C. and Freeman, J.A. (1975) Theory of current source-density analysis and determination of conductivity tensor for anuran cerebellum, J. Neurophysiol., 38: 356368. Pickard, R.S. (1979) A review of printed circuit microelectrodes and their production, J. Neurosci. Methods, 1: 30t-318. Prohaska, O., Olcaytug, F., Pfundner, P. and Dragaun, H. (1986) Thin-film multiple electrode probes: possibilities and limitations, IEEE Trans. Biomed. Eng., 33: 223-229. Rappelsberger, P., Pockberger, H. and Petsche, H. (1981) Current source density analysis: methods and application to simultaneously recorded field potentials of the rabbit's visual cortex, Pfliigers Arch., 389: 159-170. Verloop, A.J. and Holsheimer, J. (1984) A simple method for the construction of electrode arrays, J. Neurosci. Methods, 11: 173-178. Weijnen, J.A.W.M. and Chehade, N. (1987) Recording of brain potentials with FET-circuits: hazard of inadvertent lesions, Brain Res. Bull., 19: 617-618.