- Email: [email protected]
Long-term gliosis around chronically implanted platinum electrodes in the Rhesus macaque motor cortex
Neuroscience Letters 406 (2006) 81–86
Long-term gliosis around chronically implanted platinum electrodes in the Rhesus macaque motor cortex Ronald W. Griffith ∗ , Donald R. Humphrey Emory University School of Medicine, Department of Physiology, Whitehead Biomedical Research Center, Suite 605S, 615 Michael St., Atlanta, GA 30322, United States Received 17 March 2006; received in revised form 30 June 2006; accepted 7 July 2006
Abstract Chronically implanted microelectrodes have been an important tool used by neuroscientists for many years and are critical for the development of neural prostheses designed to restore function after traumatic central nervous system (CNS) injury. It is well established that a variety of mammals, including non-human primates (NHP), tolerate noble metal electrodes in the cortex for extended periods of time, but little is known about the long-term effects of electrode implantation at the cellular level. While data from rodents have clearly shown gliosis around such implants, there have been difficulties in demonstrating these reactions in NHP. Glial reactions are to be expected in NHP, since any trauma to the mammalian CNS is believed to result in the formation of a glial scar consisting of reactive astrocytes and microglia around the injury site. Because a glial scar can potentially affect the quality of recordings or stimulations from implanted electrodes, it is important to determine the extent of gliosis around implants in NHP. We studied the response of cortical glial cells to chronic electrode implantation in the motor cortices of Rhesus macaques (Macaca mulatta) after 3 months and 3 years duration. Antibodies specific for astrocytes and microglia were used to detect the presence of glial reactions around electrode implant sites. Reactive glia were found within the cortical neuropil surrounding the chronically implanted noble metal electrodes. Reactive gliosis persisted over the time periods studied and demonstrates the importance of developing strategies to minimize this event, even around noble metal implants. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Microelectrode; Glial scar; Astrocyte; Microglia; Primate
The use of brain prosthetic devises (BPD) to circumvent motor deficits related to CNS disorders such as Alzheimer’s [2] and Parkinson’s diseases [12,35], as well as, ALS and paralysis [3,4,18,24] have currently been reported. Hearing deficits resulting from auditory nerve neuromas have also been improved with BPD [21,25]. In order for BPD to be significantly useful to patients, the brain/device interface must remain stable in the CNS indefinitely without causing cellular changes that adversely affect the devices function or result in tissue destruction. To this end, investigators have labored to produce smaller implants and have begun examining the biocompatibility of materials used in their construction [15,16,17]. One of the most common materials used in chronically implanted electrodes is fine wire. Investigators have used chronically implanted electrodes in species
∗
Corresponding author. Tel.: +1 404 727 0346; fax: +1 404 727 7242. E-mail address: [email protected] (R.W. Griffith).
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.07.018
ranging from mice [20] to non-human primates [14,22] to stimulate and record from various CNS regions. While the electrodes appear to be well tolerated by research animals and humans [28], little is known about the cellular responses associated with their long-term presence in the brain. Typically, any perturbation of the adult mammalian CNS results in localized activation of glial cells and the formation of a glial scar. In the mammalian brain, microglia and astrocytes respond to CNS injury relatively quickly and immunohistochemical markers specific for these cells are upregulated during the first few days [10,11] and may remain elevated for months or subside over time depending on the type and severity of the perturbation and the developmental stage ([5,8,30], for review see [6]). Microglial hyperplasia is common while the response of astrocytes in the adult CNS is typically hypertrophic [9]. The consequences of the inflammatory response to electrode viability are not known. However, astrocytic elements ensheath and grow on electrodes [26,32–34] and electrode impedance and resistance can change with time
82
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
after implantation. Early attempts at assessing the effects of electrode implantation and stimulation of the brain were typically based on basic histology employing the use of light or electron microscopy [1,36,28,24,27]. However, more recent studies have taken advantage of glial specific markers to identify reactive cell types associated with implanted electrodes [29]. Expression of extracellular matrix molecules around chronically implanted BPD also increases [33] and may affect the approximation of axons and dendrites with electrodes. In order to better understand the glial response to long-term Pt electrode implantation (Fig. 1), we examined the brains of two Rhesus macaques with chronically implanted electrodes after 3, and 36 months using histological and immunohistochemical techniques. All experiments were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals and the Emory University Institutional Animal Care and Use Committee. Two rhesus macaques were euthanized and transcardially perfused with formalin at the end of a study examining changes in cortical muscle maps. The heads were immersed in formalin and stored at 4 ◦ C for approximately 2 years. Bone was carefully removed from around the electrode implant site and the implanted cortical hemisphere. The brain was bisected sagitally and the hemisphere removed to fresh 4% paraformaldahyde in 0.1 M phosphate buffer. To facilitate clearing the tissue of formalin, fresh paraformaldahyde solution was added daily for 1 week and the electrode implant carefully extracted from the cortex. The electrode implant site was excised and transferred to fresh 0.1 M phosphate buffer pH 7.4 (PB). Several changes of PB over a 2-day period cleared the tissue of fixative. Fifty-micrometer free-floating sections were cut while immersed in 0.01 M phosphate buffered saline pH 7.4 (PBS) at 4 ◦ C using a VibratomeTM . The sections were placed in Falcon 12-well tissue culture plates (two sections per well), washed repeatedly in PBS and stored at 4 ◦ C until the immunohistochemistry was carried out. Sections to be evaluated with cresyl violet stain were dried onto subbed slides. One electrode track was examined from the 3-month time point and three electrode tracks were examined at the 36-month time point. A monoclonal antibody (Sigma clone GA-5) against glial fibrillary acidic protein (GFAP) was used to reveal astrocytes and a monoclonal antibody (LN-3 ICN) used to label microglia. The GFAP primary antibody was prepared in a diluent of PBS containing 0.25% Triton X-100 and 1% normal horse serum (Vector) at a dilution of 1:500. The monoclonal antibody LN-3 was used at the dilution supplied by the manufacturer without Triton X-100. Sections were incubated with the primary antibodies overnight at 4 ◦ C on a rotator then washed in PBS three times for 15 min each. Biotinylated horse anti-mouse secondary antibody was prepared at a dilution of 1:500 in PBS with 1% NHS but without detergent. Sections were incubated in secondary antibody for 1 h at room temperature, washed in PBS three times for 15 min each then incubated with avidin-HRP complex (Vectastain Elite Standard – Vector Laboratories) for 45 min at room temperature. Staining was visualized using a chromogenic substrate of 0.05% 3 3 -diaminobenzidine (Sigma) with 0.005% hydrogen peroxide in 0.05 M Tris buffer at pH 7.0.
Fig. 1. Electrode design and placement in the motor cortex. (A) Each electrode bundle consists of six Pt wires of varying lengths and a cannon wafer connector. (B) Multiple electrode bundles are implanted along the anterior bank of the central fissure in the gray matter. AS: arcuate sulcus, PS: precentral sulcus, and CS: central sulcus. (C) Electrodes are inserted tangentially and come to rest in the M1 arm area. Thus, the staggered length design of the electrodes allows for electrode tips to be situated at varying depths within lamina 5. The gray matter of lamina 5 extends approximately 4–5 mm deep in this area. Electrode placement was initially verified electrophysiologically then visually when the electrodes were extracted after fixation.
Nissl stained sections (Fig. 2) show an intact neuropil with no evident loss of neurons near the electrode track. The electrode was well integrated without the presence of cavitation. Neutrophilic profiles and evidence of infection were not detected. Neurons in close proximity to the electrode track appeared normal and did not demonstrate chromatolysis. Close examination of smaller nuclei reveals an apparent numerical increase and most likely demonstrates glial proliferation. The enlargement represented in Fig. 2A and B demonstrates the typical histology of the electrode tracks in sections from animals at each of the survival times. Based solely on Nissl staining the electrode bundles appear to be well integrated and well tolerated for up to 3 years. GFAP labeled sections showed heavy staining along electrode tracks at both survival times. The electrode tracks were lined with astrocytic processes that potentially ensheath the electrodes (Fig. 3). The ensheathing elements closely resemble the glial limitans of the brain and may be a logical extension of this structure. In practical terms, production of a glial limitans around implanted electrodes would effectively isolate the structure “outside” the cortex proper and possibly produce a pathway that shunts current from electrodes and away from the intended neuronal targets. Turner et al. [34] also described astrocytic sheath formation around silicon implants in the rat brain with cellular attachment to the probes and a modest increase in GFAP positive astrocytes in tissue adjacent to the probe. In our studies, reactive glia were identified as far as 1 cm away from the electrode track (Figs. 4 and 5). Astrocytes from areas well away (>1 cm)
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
83
Fig. 2. Cresyl violet staining of Nissl substance along electrode track. (A) The neuropil surrounding the electrode track appears relatively normal in cresyl violet stained sections of M1 3 years after electrode implantation. The electrode was well integrated in the cortical tissue. The box indicates the approximate region enlarged in (B), bar in A is 1 mm. (B) While there is evidence of increased density of small nuclei adjacent to the electrode track, closely associated neurons show normal histology, bar in B is 100 m.
from the electrode tracks exhibited normal morphology without hypertrophy. While the astrocytic reaction was readily visible 3 years after implantation, the increased microglial reaction seen after 3 months (Figs. 4A and 6) was not discernable at the 3-year time point. Thus, the microglial reaction may be more transient than the astrocytic reaction. The glial response to a variety of lesions in the rat CNS is very similar. Axotomy of spinal dorsal roots and subsequent sensory axon degeneration results in qualitatively similar glial scarring to that seen after cortical stab lesions. Additionally, motor neuron degeneration in the spinal cord ventral horn leads to similar gliosis [11]. The glial staining patterns seen in the implanted non-human primate motor cortex are strikingly sim-
Fig. 3. Astrocytic processes are closely associated with the electrode tracks. (A) GFAP immunostained section including an electrode track infiltrated with astrocytic processes reminiscent of the glial limitans surrounding the brain, bar in A is 1 mm. (B) Light micrograph of actual electrode tips. Wire diameter at electrode tip is 50 um, bar in B is 300 m. (C) Enlargement of a typical electrode tip. The wire electrode was cut at an angle of approximately 45◦ , bar in C is 25 m.
ilar to those seen in the rat brain where gliosis persists for at least 4 weeks after electrode implantation [29], suggesting that the glial response to injury is also similar across species. These data suggest that a long-term astrocytic scar may persist around electrode tracks in the mammalian CNS as long as the electrodes are in place. Since electrode viability can degrade over time it is possible that the glial scar resulting from implantation and/or the presence of electrodes isolates the implants to some degree and may impede or shunt current flow away from the intended target. The pathological and/or physiological roles of reactive glia associated with chronically implanted electrodes are not known. However, the glial scars present in other types of CNS injury are known to inhibit axonal regeneration. Reactive astrocytes secrete proteoglycans that inhibit the growth of neurites in injured spinal cord and brain models [19]. If inhibitory proteoglycans are present in glial scars around electrodes, then the growth of neurites severed upon electrode implantation may be affected. Whether the approximation of regenerating neurites with nearby electrodes is important remains to be determined. Encapsulation of electrodes by reactive glia may also impede current flow and isolate the electrode from the neurons with which they are designed to communicate. Reactive gliosis may also be beneficial to repair of the injured CNS and it seems that a delicate balance exists between the positive and negative consequences of glial scar formation [23,31]. Approaches aimed at improving electrode integration and long-term viability must take into account both neuronal survival and the effects of gliosis. Improved device biocompatibility may be achieved by limiting neuronal injury due to implantation and/or electrical stimulation. Careful studies by Szarowski et al., investigating the effects of electrode size, method of insertion and surface characteristics on reactive gliosis have been carried out in the rodent. Results from these studies indicate that the initial glial response decreases as the cross-sectional area of the electrode decreases. It was noted that the larger electrodes had sharp corners and more surface irregularities than smaller electrodes and may account for greater initial damage upon insertion. Furthermore, it was found that
84
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
Fig. 4. Montage images of electrode tracks. (A) The electrode track 3 months after implantation shows heavier microglial staining adjacent to the dashed line indicating the insertion path. (B) The reactive astrocytic component of the scar at 36 months can be seen along the dashed lines indicating two distinct electrode tracks. The white and black boxes indicate areas enlarged for Fig. 5C and D, respectively. Bar: 1 mm. (C) Black box indicates area magnified in A. (D) Black box indicates area magnified in B. PCS: precentral sulcus. CS: central sulcus, and AS: arcuate sulcus. Bars: 5 mm.
the sustained glial response was independent of electrode size. Based on these findings, it is plausible that rough edges on our electrodes (Fig. 3) may have influenced the degree of gliosis early on and that the glial response seen in our study equates more with the sustained response seen in the Szarowski study [29]. It is also possible that the initial glial response to electrode implantation might be managed pharmacologically. The Nmethyl-d-aspartate receptor antagonist MK-801 has been shown
to provide some protection against neuronal injury caused by electrode implantation and use [1]. Furthermore, MK-801 blocks injury induced glial scar formation in the rat brain and spinal cord [7,13]. These findings, taken together with results from this study, suggest that further experiments aimed at reducing the glial scar around chronically implanted electrodes are warranted and that a comprehensive approach encompassing electrode design, as well as, pharmacological management of gliosis is feasible.
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
85
Fig. 5. Reactive astrocytes are present along the electrode track 3 months and 36 months after implantation. GFAP immunostaining in unaffected areas of the cortex (A) 3 months and (C) 3 years after electrode implantation. Relatively little GFAP staining is evident in the control regions. GFAP immunostaining in cortical tissue adjacent to the electrode tracks (B) 3 months and (D) 3 years after implantation reveals robust staining of hypertrophied reactive astrocytes. Bar in D is 100 m.
Fig. 6. Reactive microglia are present along the electrode track 3 months after electrode implantation. (A) In control regions, LN3 immunostaining reveals ramified microglia with fine processes which are characteristic of resting microglia. (B) However, microglia immediately adjacent to the electrode track exhibit a more ameboid morphology with shorter processes and a more hypertrophied cell bodies characteristic of reactive microglia. Bar in B is 100 m.
86
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
Acknowledgements Supported by NIH Grant No. NS40437 to Donald R. Humphrey. Thanks to Dan Kalman and Alyson Swimm for assistance with the microscopy. References [1] W.F. Agnew, D.B. Mcreery, T.G.H. Yuen, L.A. Bullara, MK-801 protects against neuronal injury induced by electrical stimulation, Neurosci. 52 (1993) 45–53. [2] A.L. Benabid, Deep brain stimulation for Parkinson’s disease, Curr. Opin. Neurobiol. 13 (2003) 696–706. [3] J.P. Donoghue, Connecting cortex to machines: recent advances in brain interfaces, Nat. Neurosci. 5 (Suppl.) (2002) 1085–1088. [4] J.P. Donoghue, A. Nurmikko, G. Friehs, M. Black, Development of neuromotor prostheses for humans, Clin. Neurophysiol. 57 (Suppl.) (2004) 593–606. [5] I. Dusart, S. Marty, M. Peschanski, Glial changes following an excitotoxic lesion in the CNS-II. Astrocytes, Neuroscience 45 (1991) 541–549. [6] J.W. Fawcett, R.A. Asher, The glial scar and central nervous system repair, Brain Res. Bull. 49 (1999) 377–391. [7] C.J. Garrison, P.M. Dougherty, S.M. Carlton, GFAP experession in lumbar spinal cord of na¨ıve and neuropathic rats treated with MK-801, Exp. Neurol. 129 (1994) 237–243. [8] S.A. Gilmore, T.J. Sims, J.E. Leiting, Astrocytic reactions in spinal gray matter following sciatic axotomy, Glia 3 (1990) 342–349. [9] M.B. Graeber, W. Tetzslaff, W.J. Streit, G.W. Kreutzberg, Microglial cells but not astrocytes undergo mitosis following rat facial nerve axotomy, Neurosci. Lett. 85 (1988) 317–321. [10] R. Griffith, J. Soria, J.G. Wood, Regulation of microglial tyrosine phophorylation in response to neuronal injury, Exp. Neurol. 161 (2000) 297–305. [11] R. Griffith, J. Sutin, Reactive astrocyte formation in vivo is regulated by noradrenergic axons, J. Comp. Neurol. 371 (1996) 362–375. [12] C.F. Haberler, P.R. Alesch, P. Mazal, K. Pilz, M. Jellinger, M. Pinter, J.A. Hainfellner, H. Budka, No tissue damage by chronic deep brain stimulation in Parkinson’s disease, Ann. Neurol. 48 (2000) 372–376. [13] D.G. Herrera, A.C. Cuello, MK-801 affects the potassium-induced increase of glial fibrillary acidic protein immunoreactivity in rat brain, Brain Res. (1992) 293–386. [14] D.R. Humphrey, L.R. Hochberg, D.J. Reed, Cortical excitability decreases and MI forelimb muscle maps fractionate with acquisition of more precise motor task, Soc. Neurosci. Abstr. 23 (1997) 1556. [15] L. Kam, W. Shain, J.N. Turner, R. Bizios, Axonal outgrowth of hippocampal neurons on micro-scale networks of polylysine-conjugated laminin, Biomaterials 22 (2001) 1049–1054. [16] L. Kam, W. Shain, J.N. Turner, R. Bizios, Correlation of astroglial cell function on micro-patterned surfaces with specific geometric parameters, Biomaterials 20 (1999) 2343–2350. [17] L. Kam, W. Shain, J.N. Turner, R. Bizios, Selective adhesion of astrocytes to surfaces modified with immobilized peptides, Biomaterials 23 (2002) 511–515.
[18] P.R. Kennedy, M.T. Kirby, M.M. Moore, B. King, A. Mallory, Computer control using human intracortical local field potentials, IEEE Trans. Neural Sys. Rehabil. Eng. 12 (2004) 339–344. [19] R.J. McKeon, R.C. Screiber, J.S. Rudge, J. Silver, Reduction of neurite outgrowth in a model of glial scarring following CNS injury is correlated with the expxression of inhibitory molecules on reactive astrocytes, J. Neurosci. 11 (1991) 3398–3411. [20] J.I. Oka, M. Imanishi, A new technique for implanting fine-wire microelectrodes for chronic recording of unit activity from freely-moving mice, Neurosci. Res. 36 (2000) 93–96. [21] S.R. Otto, D.E. Brackman, W.E. Hitselberger, R.V. Shannon, J. Kuchta, Multichannel auditory brainstem implant: update on performance in 61 patients, J. Neurosurg. 96 (2002) 1063–1071. [22] M.C. Park, A. Belhaj-Saif, P.D. Cheney, Properties of primary motor cortex output to forelimb muscles in rhesus macaques, J. Neurophysiol. 92 (2004) 2968–2984. [23] M. Pekny, M. Nilsson, Astrocyte activation and reactive gliosis, Glia 50 (2005) 427–434. [24] R.H. Pudenz, W.F. Agnew, L.A. Bullara, Effects of electrical stimulation of brain, Brain Behav. Evol. 14 (1977) 103–205. [25] J.P. Rauschecker, R.V. Shannon, Sending sound to the brain, Science 295 (2002) 1025–1029. [26] P.M. St. John, L. Kam, S.W. Turner, H.G. Craighead, M. Issacson, J.N. Turner, W. Shain, Preferential glial cell attachment to microcontact printed surfaces, J. Neurosci. Meth. 75 (1997) 171–177. [27] S.S. Stensaas, L.J. Stensaas, Histopathological evaluation of materials implanted in the cerebral cortex, Acta Neuropath. (Berl.) 41 (1978) 145–155. [28] S. Suner, M.R. Fellows, C. Vargas-Irwin, G.K. Nakata, J.P. Donoghue, Reliability of signals from a chronically implanted silicon-based electrode array in non-human primate primary motor cortex, IEEE Trans. Neural Sys. Rehabil. 13 (2005) 524–541. [29] D.H. Szarowski, M.D. Andersen, S. Retterer, A.J. Spence, M. Isaacson, H.G. Craighead, J.N. Turner, W. Shain, Brain responses to micro-machined silicon devices, Brain Res. 983 (2003) 23–35. [30] K.S. Topp, B.T. Faddis, V.K. Vijayan, Trauma-induced proliferation of astrocytes in the brains of young and aged rats, Glia 2 (1989) 201–211. [31] G. Trendelenburg, U. Durnagl, Neuroprotective role of astrocytes in cerebral ischemia: focus on ischemic preconditioning, Glia 51 (2005) 307–320. [32] A.M.P. Turner, N. Dowell, S.W.P. Turner, L. Kam, M. Isaacson, J.N. Turner, H.G. Craighead, W. Shain, Attachment of astroglial cells to microfabricated pillar arrays of different geometries, J. Biomed. Mat. Res. 51 (2000) 430–441. [33] J.L. Turner, B.A. Sipple, H.G. Craighead, M. Isaacson, A.J.H. Spence, W. Shain, Changes in extracellular matrix expression following insertion of model neural prosthetic devises, J. Neurochem. 47 (2000) S82. [34] J.N. Turner, W. Shain, D.H. Szarowski, M. Andersen, S. Martins, M. Isaacson, H. Craighead, Cerebral astrocyte response to micromachined silicon implants, Exp. Neurol. 156 (1999) 33–49. [35] J.L. Vitek, T. Hashimoto, J. Peoples, M.R. Delong, R.A. Bakay, Acute stimulation in the external segment of the globus pallidus improves parkinsonian motor signs, Mov. Disord. 19 (2004) 907–915. [36] T.G.H. Yuen, W.F. Agnew, Histological evaluation of polyesterimideinsulated gold wires in brain, Biomaterials 16 (1995) 951–956.
Long-term gliosis around chronically implanted platinum electrodes in the Rhesus macaque motor cortex Ronald W. Griffith ∗ , Donald R. Humphrey Emory University School of Medicine, Department of Physiology, Whitehead Biomedical Research Center, Suite 605S, 615 Michael St., Atlanta, GA 30322, United States Received 17 March 2006; received in revised form 30 June 2006; accepted 7 July 2006
Abstract Chronically implanted microelectrodes have been an important tool used by neuroscientists for many years and are critical for the development of neural prostheses designed to restore function after traumatic central nervous system (CNS) injury. It is well established that a variety of mammals, including non-human primates (NHP), tolerate noble metal electrodes in the cortex for extended periods of time, but little is known about the long-term effects of electrode implantation at the cellular level. While data from rodents have clearly shown gliosis around such implants, there have been difficulties in demonstrating these reactions in NHP. Glial reactions are to be expected in NHP, since any trauma to the mammalian CNS is believed to result in the formation of a glial scar consisting of reactive astrocytes and microglia around the injury site. Because a glial scar can potentially affect the quality of recordings or stimulations from implanted electrodes, it is important to determine the extent of gliosis around implants in NHP. We studied the response of cortical glial cells to chronic electrode implantation in the motor cortices of Rhesus macaques (Macaca mulatta) after 3 months and 3 years duration. Antibodies specific for astrocytes and microglia were used to detect the presence of glial reactions around electrode implant sites. Reactive glia were found within the cortical neuropil surrounding the chronically implanted noble metal electrodes. Reactive gliosis persisted over the time periods studied and demonstrates the importance of developing strategies to minimize this event, even around noble metal implants. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Microelectrode; Glial scar; Astrocyte; Microglia; Primate
The use of brain prosthetic devises (BPD) to circumvent motor deficits related to CNS disorders such as Alzheimer’s [2] and Parkinson’s diseases [12,35], as well as, ALS and paralysis [3,4,18,24] have currently been reported. Hearing deficits resulting from auditory nerve neuromas have also been improved with BPD [21,25]. In order for BPD to be significantly useful to patients, the brain/device interface must remain stable in the CNS indefinitely without causing cellular changes that adversely affect the devices function or result in tissue destruction. To this end, investigators have labored to produce smaller implants and have begun examining the biocompatibility of materials used in their construction [15,16,17]. One of the most common materials used in chronically implanted electrodes is fine wire. Investigators have used chronically implanted electrodes in species
∗
Corresponding author. Tel.: +1 404 727 0346; fax: +1 404 727 7242. E-mail address: [email protected] (R.W. Griffith).
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.07.018
ranging from mice [20] to non-human primates [14,22] to stimulate and record from various CNS regions. While the electrodes appear to be well tolerated by research animals and humans [28], little is known about the cellular responses associated with their long-term presence in the brain. Typically, any perturbation of the adult mammalian CNS results in localized activation of glial cells and the formation of a glial scar. In the mammalian brain, microglia and astrocytes respond to CNS injury relatively quickly and immunohistochemical markers specific for these cells are upregulated during the first few days [10,11] and may remain elevated for months or subside over time depending on the type and severity of the perturbation and the developmental stage ([5,8,30], for review see [6]). Microglial hyperplasia is common while the response of astrocytes in the adult CNS is typically hypertrophic [9]. The consequences of the inflammatory response to electrode viability are not known. However, astrocytic elements ensheath and grow on electrodes [26,32–34] and electrode impedance and resistance can change with time
82
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
after implantation. Early attempts at assessing the effects of electrode implantation and stimulation of the brain were typically based on basic histology employing the use of light or electron microscopy [1,36,28,24,27]. However, more recent studies have taken advantage of glial specific markers to identify reactive cell types associated with implanted electrodes [29]. Expression of extracellular matrix molecules around chronically implanted BPD also increases [33] and may affect the approximation of axons and dendrites with electrodes. In order to better understand the glial response to long-term Pt electrode implantation (Fig. 1), we examined the brains of two Rhesus macaques with chronically implanted electrodes after 3, and 36 months using histological and immunohistochemical techniques. All experiments were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals and the Emory University Institutional Animal Care and Use Committee. Two rhesus macaques were euthanized and transcardially perfused with formalin at the end of a study examining changes in cortical muscle maps. The heads were immersed in formalin and stored at 4 ◦ C for approximately 2 years. Bone was carefully removed from around the electrode implant site and the implanted cortical hemisphere. The brain was bisected sagitally and the hemisphere removed to fresh 4% paraformaldahyde in 0.1 M phosphate buffer. To facilitate clearing the tissue of formalin, fresh paraformaldahyde solution was added daily for 1 week and the electrode implant carefully extracted from the cortex. The electrode implant site was excised and transferred to fresh 0.1 M phosphate buffer pH 7.4 (PB). Several changes of PB over a 2-day period cleared the tissue of fixative. Fifty-micrometer free-floating sections were cut while immersed in 0.01 M phosphate buffered saline pH 7.4 (PBS) at 4 ◦ C using a VibratomeTM . The sections were placed in Falcon 12-well tissue culture plates (two sections per well), washed repeatedly in PBS and stored at 4 ◦ C until the immunohistochemistry was carried out. Sections to be evaluated with cresyl violet stain were dried onto subbed slides. One electrode track was examined from the 3-month time point and three electrode tracks were examined at the 36-month time point. A monoclonal antibody (Sigma clone GA-5) against glial fibrillary acidic protein (GFAP) was used to reveal astrocytes and a monoclonal antibody (LN-3 ICN) used to label microglia. The GFAP primary antibody was prepared in a diluent of PBS containing 0.25% Triton X-100 and 1% normal horse serum (Vector) at a dilution of 1:500. The monoclonal antibody LN-3 was used at the dilution supplied by the manufacturer without Triton X-100. Sections were incubated with the primary antibodies overnight at 4 ◦ C on a rotator then washed in PBS three times for 15 min each. Biotinylated horse anti-mouse secondary antibody was prepared at a dilution of 1:500 in PBS with 1% NHS but without detergent. Sections were incubated in secondary antibody for 1 h at room temperature, washed in PBS three times for 15 min each then incubated with avidin-HRP complex (Vectastain Elite Standard – Vector Laboratories) for 45 min at room temperature. Staining was visualized using a chromogenic substrate of 0.05% 3 3 -diaminobenzidine (Sigma) with 0.005% hydrogen peroxide in 0.05 M Tris buffer at pH 7.0.
Fig. 1. Electrode design and placement in the motor cortex. (A) Each electrode bundle consists of six Pt wires of varying lengths and a cannon wafer connector. (B) Multiple electrode bundles are implanted along the anterior bank of the central fissure in the gray matter. AS: arcuate sulcus, PS: precentral sulcus, and CS: central sulcus. (C) Electrodes are inserted tangentially and come to rest in the M1 arm area. Thus, the staggered length design of the electrodes allows for electrode tips to be situated at varying depths within lamina 5. The gray matter of lamina 5 extends approximately 4–5 mm deep in this area. Electrode placement was initially verified electrophysiologically then visually when the electrodes were extracted after fixation.
Nissl stained sections (Fig. 2) show an intact neuropil with no evident loss of neurons near the electrode track. The electrode was well integrated without the presence of cavitation. Neutrophilic profiles and evidence of infection were not detected. Neurons in close proximity to the electrode track appeared normal and did not demonstrate chromatolysis. Close examination of smaller nuclei reveals an apparent numerical increase and most likely demonstrates glial proliferation. The enlargement represented in Fig. 2A and B demonstrates the typical histology of the electrode tracks in sections from animals at each of the survival times. Based solely on Nissl staining the electrode bundles appear to be well integrated and well tolerated for up to 3 years. GFAP labeled sections showed heavy staining along electrode tracks at both survival times. The electrode tracks were lined with astrocytic processes that potentially ensheath the electrodes (Fig. 3). The ensheathing elements closely resemble the glial limitans of the brain and may be a logical extension of this structure. In practical terms, production of a glial limitans around implanted electrodes would effectively isolate the structure “outside” the cortex proper and possibly produce a pathway that shunts current from electrodes and away from the intended neuronal targets. Turner et al. [34] also described astrocytic sheath formation around silicon implants in the rat brain with cellular attachment to the probes and a modest increase in GFAP positive astrocytes in tissue adjacent to the probe. In our studies, reactive glia were identified as far as 1 cm away from the electrode track (Figs. 4 and 5). Astrocytes from areas well away (>1 cm)
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
83
Fig. 2. Cresyl violet staining of Nissl substance along electrode track. (A) The neuropil surrounding the electrode track appears relatively normal in cresyl violet stained sections of M1 3 years after electrode implantation. The electrode was well integrated in the cortical tissue. The box indicates the approximate region enlarged in (B), bar in A is 1 mm. (B) While there is evidence of increased density of small nuclei adjacent to the electrode track, closely associated neurons show normal histology, bar in B is 100 m.
from the electrode tracks exhibited normal morphology without hypertrophy. While the astrocytic reaction was readily visible 3 years after implantation, the increased microglial reaction seen after 3 months (Figs. 4A and 6) was not discernable at the 3-year time point. Thus, the microglial reaction may be more transient than the astrocytic reaction. The glial response to a variety of lesions in the rat CNS is very similar. Axotomy of spinal dorsal roots and subsequent sensory axon degeneration results in qualitatively similar glial scarring to that seen after cortical stab lesions. Additionally, motor neuron degeneration in the spinal cord ventral horn leads to similar gliosis [11]. The glial staining patterns seen in the implanted non-human primate motor cortex are strikingly sim-
Fig. 3. Astrocytic processes are closely associated with the electrode tracks. (A) GFAP immunostained section including an electrode track infiltrated with astrocytic processes reminiscent of the glial limitans surrounding the brain, bar in A is 1 mm. (B) Light micrograph of actual electrode tips. Wire diameter at electrode tip is 50 um, bar in B is 300 m. (C) Enlargement of a typical electrode tip. The wire electrode was cut at an angle of approximately 45◦ , bar in C is 25 m.
ilar to those seen in the rat brain where gliosis persists for at least 4 weeks after electrode implantation [29], suggesting that the glial response to injury is also similar across species. These data suggest that a long-term astrocytic scar may persist around electrode tracks in the mammalian CNS as long as the electrodes are in place. Since electrode viability can degrade over time it is possible that the glial scar resulting from implantation and/or the presence of electrodes isolates the implants to some degree and may impede or shunt current flow away from the intended target. The pathological and/or physiological roles of reactive glia associated with chronically implanted electrodes are not known. However, the glial scars present in other types of CNS injury are known to inhibit axonal regeneration. Reactive astrocytes secrete proteoglycans that inhibit the growth of neurites in injured spinal cord and brain models [19]. If inhibitory proteoglycans are present in glial scars around electrodes, then the growth of neurites severed upon electrode implantation may be affected. Whether the approximation of regenerating neurites with nearby electrodes is important remains to be determined. Encapsulation of electrodes by reactive glia may also impede current flow and isolate the electrode from the neurons with which they are designed to communicate. Reactive gliosis may also be beneficial to repair of the injured CNS and it seems that a delicate balance exists between the positive and negative consequences of glial scar formation [23,31]. Approaches aimed at improving electrode integration and long-term viability must take into account both neuronal survival and the effects of gliosis. Improved device biocompatibility may be achieved by limiting neuronal injury due to implantation and/or electrical stimulation. Careful studies by Szarowski et al., investigating the effects of electrode size, method of insertion and surface characteristics on reactive gliosis have been carried out in the rodent. Results from these studies indicate that the initial glial response decreases as the cross-sectional area of the electrode decreases. It was noted that the larger electrodes had sharp corners and more surface irregularities than smaller electrodes and may account for greater initial damage upon insertion. Furthermore, it was found that
84
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
Fig. 4. Montage images of electrode tracks. (A) The electrode track 3 months after implantation shows heavier microglial staining adjacent to the dashed line indicating the insertion path. (B) The reactive astrocytic component of the scar at 36 months can be seen along the dashed lines indicating two distinct electrode tracks. The white and black boxes indicate areas enlarged for Fig. 5C and D, respectively. Bar: 1 mm. (C) Black box indicates area magnified in A. (D) Black box indicates area magnified in B. PCS: precentral sulcus. CS: central sulcus, and AS: arcuate sulcus. Bars: 5 mm.
the sustained glial response was independent of electrode size. Based on these findings, it is plausible that rough edges on our electrodes (Fig. 3) may have influenced the degree of gliosis early on and that the glial response seen in our study equates more with the sustained response seen in the Szarowski study [29]. It is also possible that the initial glial response to electrode implantation might be managed pharmacologically. The Nmethyl-d-aspartate receptor antagonist MK-801 has been shown
to provide some protection against neuronal injury caused by electrode implantation and use [1]. Furthermore, MK-801 blocks injury induced glial scar formation in the rat brain and spinal cord [7,13]. These findings, taken together with results from this study, suggest that further experiments aimed at reducing the glial scar around chronically implanted electrodes are warranted and that a comprehensive approach encompassing electrode design, as well as, pharmacological management of gliosis is feasible.
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
85
Fig. 5. Reactive astrocytes are present along the electrode track 3 months and 36 months after implantation. GFAP immunostaining in unaffected areas of the cortex (A) 3 months and (C) 3 years after electrode implantation. Relatively little GFAP staining is evident in the control regions. GFAP immunostaining in cortical tissue adjacent to the electrode tracks (B) 3 months and (D) 3 years after implantation reveals robust staining of hypertrophied reactive astrocytes. Bar in D is 100 m.
Fig. 6. Reactive microglia are present along the electrode track 3 months after electrode implantation. (A) In control regions, LN3 immunostaining reveals ramified microglia with fine processes which are characteristic of resting microglia. (B) However, microglia immediately adjacent to the electrode track exhibit a more ameboid morphology with shorter processes and a more hypertrophied cell bodies characteristic of reactive microglia. Bar in B is 100 m.
86
R.W. Griffith, D.R. Humphrey / Neuroscience Letters 406 (2006) 81–86
Acknowledgements Supported by NIH Grant No. NS40437 to Donald R. Humphrey. Thanks to Dan Kalman and Alyson Swimm for assistance with the microscopy. References [1] W.F. Agnew, D.B. Mcreery, T.G.H. Yuen, L.A. Bullara, MK-801 protects against neuronal injury induced by electrical stimulation, Neurosci. 52 (1993) 45–53. [2] A.L. Benabid, Deep brain stimulation for Parkinson’s disease, Curr. Opin. Neurobiol. 13 (2003) 696–706. [3] J.P. Donoghue, Connecting cortex to machines: recent advances in brain interfaces, Nat. Neurosci. 5 (Suppl.) (2002) 1085–1088. [4] J.P. Donoghue, A. Nurmikko, G. Friehs, M. Black, Development of neuromotor prostheses for humans, Clin. Neurophysiol. 57 (Suppl.) (2004) 593–606. [5] I. Dusart, S. Marty, M. Peschanski, Glial changes following an excitotoxic lesion in the CNS-II. Astrocytes, Neuroscience 45 (1991) 541–549. [6] J.W. Fawcett, R.A. Asher, The glial scar and central nervous system repair, Brain Res. Bull. 49 (1999) 377–391. [7] C.J. Garrison, P.M. Dougherty, S.M. Carlton, GFAP experession in lumbar spinal cord of na¨ıve and neuropathic rats treated with MK-801, Exp. Neurol. 129 (1994) 237–243. [8] S.A. Gilmore, T.J. Sims, J.E. Leiting, Astrocytic reactions in spinal gray matter following sciatic axotomy, Glia 3 (1990) 342–349. [9] M.B. Graeber, W. Tetzslaff, W.J. Streit, G.W. Kreutzberg, Microglial cells but not astrocytes undergo mitosis following rat facial nerve axotomy, Neurosci. Lett. 85 (1988) 317–321. [10] R. Griffith, J. Soria, J.G. Wood, Regulation of microglial tyrosine phophorylation in response to neuronal injury, Exp. Neurol. 161 (2000) 297–305. [11] R. Griffith, J. Sutin, Reactive astrocyte formation in vivo is regulated by noradrenergic axons, J. Comp. Neurol. 371 (1996) 362–375. [12] C.F. Haberler, P.R. Alesch, P. Mazal, K. Pilz, M. Jellinger, M. Pinter, J.A. Hainfellner, H. Budka, No tissue damage by chronic deep brain stimulation in Parkinson’s disease, Ann. Neurol. 48 (2000) 372–376. [13] D.G. Herrera, A.C. Cuello, MK-801 affects the potassium-induced increase of glial fibrillary acidic protein immunoreactivity in rat brain, Brain Res. (1992) 293–386. [14] D.R. Humphrey, L.R. Hochberg, D.J. Reed, Cortical excitability decreases and MI forelimb muscle maps fractionate with acquisition of more precise motor task, Soc. Neurosci. Abstr. 23 (1997) 1556. [15] L. Kam, W. Shain, J.N. Turner, R. Bizios, Axonal outgrowth of hippocampal neurons on micro-scale networks of polylysine-conjugated laminin, Biomaterials 22 (2001) 1049–1054. [16] L. Kam, W. Shain, J.N. Turner, R. Bizios, Correlation of astroglial cell function on micro-patterned surfaces with specific geometric parameters, Biomaterials 20 (1999) 2343–2350. [17] L. Kam, W. Shain, J.N. Turner, R. Bizios, Selective adhesion of astrocytes to surfaces modified with immobilized peptides, Biomaterials 23 (2002) 511–515.
[18] P.R. Kennedy, M.T. Kirby, M.M. Moore, B. King, A. Mallory, Computer control using human intracortical local field potentials, IEEE Trans. Neural Sys. Rehabil. Eng. 12 (2004) 339–344. [19] R.J. McKeon, R.C. Screiber, J.S. Rudge, J. Silver, Reduction of neurite outgrowth in a model of glial scarring following CNS injury is correlated with the expxression of inhibitory molecules on reactive astrocytes, J. Neurosci. 11 (1991) 3398–3411. [20] J.I. Oka, M. Imanishi, A new technique for implanting fine-wire microelectrodes for chronic recording of unit activity from freely-moving mice, Neurosci. Res. 36 (2000) 93–96. [21] S.R. Otto, D.E. Brackman, W.E. Hitselberger, R.V. Shannon, J. Kuchta, Multichannel auditory brainstem implant: update on performance in 61 patients, J. Neurosurg. 96 (2002) 1063–1071. [22] M.C. Park, A. Belhaj-Saif, P.D. Cheney, Properties of primary motor cortex output to forelimb muscles in rhesus macaques, J. Neurophysiol. 92 (2004) 2968–2984. [23] M. Pekny, M. Nilsson, Astrocyte activation and reactive gliosis, Glia 50 (2005) 427–434. [24] R.H. Pudenz, W.F. Agnew, L.A. Bullara, Effects of electrical stimulation of brain, Brain Behav. Evol. 14 (1977) 103–205. [25] J.P. Rauschecker, R.V. Shannon, Sending sound to the brain, Science 295 (2002) 1025–1029. [26] P.M. St. John, L. Kam, S.W. Turner, H.G. Craighead, M. Issacson, J.N. Turner, W. Shain, Preferential glial cell attachment to microcontact printed surfaces, J. Neurosci. Meth. 75 (1997) 171–177. [27] S.S. Stensaas, L.J. Stensaas, Histopathological evaluation of materials implanted in the cerebral cortex, Acta Neuropath. (Berl.) 41 (1978) 145–155. [28] S. Suner, M.R. Fellows, C. Vargas-Irwin, G.K. Nakata, J.P. Donoghue, Reliability of signals from a chronically implanted silicon-based electrode array in non-human primate primary motor cortex, IEEE Trans. Neural Sys. Rehabil. 13 (2005) 524–541. [29] D.H. Szarowski, M.D. Andersen, S. Retterer, A.J. Spence, M. Isaacson, H.G. Craighead, J.N. Turner, W. Shain, Brain responses to micro-machined silicon devices, Brain Res. 983 (2003) 23–35. [30] K.S. Topp, B.T. Faddis, V.K. Vijayan, Trauma-induced proliferation of astrocytes in the brains of young and aged rats, Glia 2 (1989) 201–211. [31] G. Trendelenburg, U. Durnagl, Neuroprotective role of astrocytes in cerebral ischemia: focus on ischemic preconditioning, Glia 51 (2005) 307–320. [32] A.M.P. Turner, N. Dowell, S.W.P. Turner, L. Kam, M. Isaacson, J.N. Turner, H.G. Craighead, W. Shain, Attachment of astroglial cells to microfabricated pillar arrays of different geometries, J. Biomed. Mat. Res. 51 (2000) 430–441. [33] J.L. Turner, B.A. Sipple, H.G. Craighead, M. Isaacson, A.J.H. Spence, W. Shain, Changes in extracellular matrix expression following insertion of model neural prosthetic devises, J. Neurochem. 47 (2000) S82. [34] J.N. Turner, W. Shain, D.H. Szarowski, M. Andersen, S. Martins, M. Isaacson, H. Craighead, Cerebral astrocyte response to micromachined silicon implants, Exp. Neurol. 156 (1999) 33–49. [35] J.L. Vitek, T. Hashimoto, J. Peoples, M.R. Delong, R.A. Bakay, Acute stimulation in the external segment of the globus pallidus improves parkinsonian motor signs, Mov. Disord. 19 (2004) 907–915. [36] T.G.H. Yuen, W.F. Agnew, Histological evaluation of polyesterimideinsulated gold wires in brain, Biomaterials 16 (1995) 951–956.