WO2012139124A1

WO2012139124A1 - Regenerative microchannel electrode array for peripheral nerve interfacing

Info

Publication number: WO2012139124A1
Application number: PCT/US2012/032784
Authority: WO
Inventors: Akhil SRINIVASAN; Liang Guo; Yoonsu Choi; Ravi Bellamkonda
Original assignee: Georgia Tech Research Corporation
Priority date: 2011-04-07
Filing date: 2012-04-09
Publication date: 2012-10-11

Abstract

A regenerative microchannel scaffold capable of housing microelectrodes to form a high-throughput peripheral nerve interface is described. The regenerative microchannel scaffold was fabricated using polydimethylsiloxane (PDMS) and SU-8 as substrate and structural materials, respectively. This microchannel scaffold is capable of being integrated with a PDMS-based thin-film microelectrode array (MEA) to form a peripheral nerve interface. Preliminary in vitro characterization was performed using dorsal root ganglia cultures to evaluate biocompatibility and the ability of the microchannel design to direct and orient neurite outgrowth. Our findings provided evidence that the microchannel design will be effective in directing axon regeneration to the benefit of a neural interfacing device.

Description

REGENERATIVE MICROCHANNEL ELECTRODE ARRAY FOR PERIPHERAL

NERVE INTERFACING

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of U.S. Provisional Application No. 61/472,959 filed

7 April 2011, the entire contents and substance of which are hereby incorporated by reference.

GOVERNMENT LICENSE RIGHTS

This invention was made with Government support under Agreement/Contract Number CBET 0651716 (RVB), awarded by National Science Foundation. The Government has certain rights in the invention.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention is directed towards electrical stimulation or sensing of neural cells. More particularly, the present invention relates to regenerative microchannel scaffolds capable of housing microelectrodes to form high-throughput peripheral nerve interfaces.

2. Description of the Related Art

In the year 2005, 1.6 million persons were living with a limb amputation and it is projected that by 2050 this number will more than double to 3.6 million. These individuals are forced to live with a severe loss of function, pain, and an overall reduction in quality of life.

In order to mitigate these negative outcomes after amputation, there is great interest in the design of prosthetics that cosmetically and functionally mimic the lost appendages. Currently, neural prosthetics provide only a small fraction of the functionality of a natural limb.

The neural interfacing technologies utilized in these prosthetics can be divided into two major areas comprising central neural interfaces and peripheral neural interfaces. There have been a number of approaches within both of these areas, but all have met with significant difficulties and limitations. Central neural interfaces are plagued by the invasiveness of needing to implant electrodes directly in the brain and the fact that this will cause another injury in addition to the amputation of the limb. Furthermore, even after the implantation and recording of neural signals a huge obstacle of decoding these signals remains. Peripheral neural interfaces are significantly less invasive as they are implanted in peripheral nerves and can be implanted at the time of amputation. Another large advantage peripheral neural interfaces have is that the signals obtained from peripheral nerves, closely represents the signal intended for muscle activation. The decoding has been taken care of by the descending pathways of the brain and spinal cord, obviating the need for artificial decoding.

Existing peripheral prosthetics often use electromyo grams from various sets of muscles. However, this method only allows limited prosthetic control and does not allow for proprioception, tactile feedback, and other sensation. Next generation peripheral nerve prosthetics aim to interface with the original limb's remaining nervous system, which remains viable and functional for years after injury. This includes nerve tracts descending from the cortex traveling through the cerebellum and basal ganglia as well as sensory processing centers in the spine. However, even these next generation interfaces are plagued by issues stemming from damage to the nerve due to interfacing, long term stability, and stimulation and recording specificity. Cuff electrodes, like the Flat Interface Nerve Electrode (FINE, Fig. 1) that wrap around the nerve and deform it, cause relatively minimal damage and provide long-term stability at the tissue-electrode interface. However they do so at the expense of stimulation/recording specificity because they are not in direct contact with individual axons. Instead they tend to record and stimulate nerves on a fascicular level due to the distance of the electrodes from axons. While interfacing with fascicles is useful in some applications, it does not provide the specificity needed to control a neural prosthetic with large degrees of freedom. Additionally many electrodes are rendered useless due to poor contact between the nerve and the electrode.

In contrast, the Utah Slanted Electrode Array (USEA, Fig. 2a) with needle like electrodes penetrates into the nerve and resides within individual fascicles, axon bundles. This design provides direct contact between electrodes and axons and leads to enhanced specificity over a large number of channels. However the insertion causes irreparable physical damage to the nerve as shown in Fig. 2b. This image shows a representative histological section of a nerve at the insertion site of one of the electrodes from the USEA. Additionally, the USEA electrodes suffer from chronic inflammation and the eventual formation of scar tissue. Due to these drawbacks, the overall efficacy has been limited to 80% of the electrodes for approximately five months. The Longitudinal Intrafascicular Electrode (LIFE, Fig. 3) electrode platform, which is a wire like electrode that is inserted longitudinally into the nerve, has been shown to strike a balance between specificity and reliability. However, while showing some reliability, the insertion still causes damage to the remaining viable portion of the nerve. Perhaps most importantly, by design the LIFE is limited to a small number of channels when compared to the FINE or USEA and simply does not scale to the level necessary for functional and viable interfaces for amputees. As shown by these hallmark peripheral nerve interfaces, there seems to be a large tradeoff between the ability to selectively interface with individual axons versus the amount of disruption to the nerve and long-term stability as a result of trying to get in close proximity to axons.

In order to combat the negative outcomes of trying to get in close proximity, a regenerative approach is being explored where cut nerves are encouraged to grow into specific electrode geometries rather than being forced into specific electrode geometries or penetrated by electrodes. While this is advantageous when compared to the significant limitations of non- regenerative neural prosthetics, these approaches have met again with limited success.

The Sieve Electrode (Fig. 4), the hallmark regenerative interface, is effectively a thin perforated disk that is attached to the nerve perpendicularly so the nerve is forced to regenerate through the device. Some of the holes in the disk have ring electrodes that contact the axons regenerating through the holes. Theoretically, each hole could have a ring electrode allowing this device to interface with an extraordinarily large number of axons. However, this design has only a few viable electrodes because the measurable extracellular potential of an axon's action potential (AP) is small, and there is a spatial dependence of the electrode to the Nodes of Ranvier where the extracellular potential of the AP is largest.

The measurable extracellular potential of action potentials is small due to the low impedance extracellular volume surrounding axons and nerves and the resulting high ionic diffusion/dispersion. It is important to note that not only are Sieve electrodes forced to endure this low signal, but all peripheral nerve interfaces to date are forced to endure it. This severe limitation of all peripheral nerve interfaces is something this proposal directly addresses. Additionally, all peripheral nerve interfaces that contact the nerve in a perpendicular manner (Sieve electrodes, Utah electrode arrays, etc.) are spatially dependant on the nodes of Ranvier, breaks in axon insulation where the extracellular potential is largest, which occur at least every 2mm. This is yet another severe limitation of state-of-the-art peripheral nerve interfacing technologies and again is something this proposal directly addresses.

Another major area of limitations for all current interfacing technologies stems from the mixed nature of most peripheral nerves and the inherent difficulty in separating sensory and motor signals. No technologies to date have successfully developed an interfacing platform capable of spatially separating and electrically isolating sensory and motor axons for stimulation and recording purposes. Current interfaces are forced to physically treat all axons the same. When the recording and stimulation of axons is attempted it is unknown where the axons spatially are and generally whether they are sensory or motor. In an ideal world however, interfaces would be able to specifically and selectively record from motor axons and stimulate sensory axons without affecting the other modality. The ability to achieve this type of separation would enable the 'mapping' of a regenerated peripheral nerve in terms of sensory vs. motor axonal makeup and allow the device to interact with these different subsets of axons in different ways.

It would be beneficial to provide a regenerative neural interface that can successfully establish high-channel, bi-directional communication between, for example, an amputated nerve and a prosthetic limb. Such an interface would allow for intimate contact between electrodes and small groups of axons while enhancing regeneration. Thus, it is to such a system, device and method that the present invention is primarily directed.

BRIEF SUMMARY OF THE INVENTION

The present invention comprises a peripheral nerve interface through which high- channel, bi-directional communication between an amputated nerve and a prosthetic limb can be established. The interface comprises a regenerative microchannel-scaffold electrode array and an integrated multiplexer/amplifier (a MicroChannel Regenerative Interface).

Encouraging axons to grow through microchannels significantly improves the ability to record from axons by constraining the low impedance extracellular fluid and amplifying the extracellular potential of an action potential. Additionally, the present invention allows for intimate contact between small groups of axons and electrodes while enhancing recording selectivity. The present invention enables recording from regenerated axons using the electrodes housed within the microchannel scaffold. This establishes a neural prosthetic interface that can record volitional motor commands to control a robotic limb with large degrees of freedom.

Thus, the present invention preferably comprises a regenerative neural interface to establish high-channel, bi-directional communication between protoplasmic protrusions and, for example, a prosthetic limb. In an exemplary embodiment, the interface enables intimate contact between electrodes and small groups of axons while enhancing regeneration. The interface can comprise a thin-film PDMS/SU-8 based regenerative microchannel scaffold with microelectrodes incorporated in the microchannels. Confining regenerated axons in microchannels increases the extracellular potential resulting from an AP, thus, enhancing signal strength and recording capabilities.

Preferably, the present invention is a microchannel -based device that successfully records spontaneous single unit action potentials from regenerated axons.

In an exemplary embodiment, the present invention's use of polydimethylsiloxane allows for a highly flexible and stretchable scaffold, while SU-8 polymer allows for mechanical integrity of the microchannels.

In an exemplary embodiment, the microchannels are closed microchannels with incorporated electrodes.

In an exemplary embodiment, the present invention further includes methods for fabricating closed microchannels.

In an exemplary embodiment, the present invention further includes for patterning electrodes on PDMS using photolithography and lift-off processes.

In an exemplary embodiment, the present invention further incorporates electronics that allows for multiplexing and amplification of signals from electrodes.

In another exemplary embodiment, the present invention further incorporates an electrical shielding wire/cage-like structure surrounding the implant. This cage serves a purpose similar to that of a Faraday cage, and greatly reduces noise and disruptive electromyogram signals from local muscles. The present invention, according to a preferred embodiment is a polydimethylsiloxane (PDMS)/SU-8 based regenerative microchannel scaffold that provides the housing basis for integrated microelectrodes for reliable, high-throughput peripheral nerve interfacing. The microchannel-scaffold electrode array has been functionally evaluated through the stimulation of and single unit recordings from regenerated axons in a rat sciatic nerve model. By forcing regenerating axons to grow through microchannels with integrated microelectrodes, the intimate and isolated contact will facilitate a more selective recording and stimulation.

To fabricate an exemplary microchannel-scaffold, a PDMS base layer is first formed, in which a microelectrode array would be fabricated if electrodes were to be incorporated into the scaffold. SU-8 microchannel walls are then patterned on the PDMS surface. Finally, a PDMS cover layer is bonded onto top of the SU-8 walls to form closed microchannels.

Microchannel wall width and height are approximately 20μιη and ΙΟΟμιη, respectively; microchannel widths range from approximately 50 to 150μιη; and channel lengths range from approximately 1 to 5mm. In a final configuration, the scaffold was rolled on itself forming the final implant with a total radius of approximately 1.5mm.

In in vitro studies, dorsal root ganglia (DRG's) were explanted from PI rat pups and cultured on open versions of the PDMS-SU-8 scaffolds (lacking the top layer). This characterization using DRG cultures to evaluate biocompatibility, and the ability of the present microchannel design to direct and orient neurite outgrowth, proved successful.

The present invention can comprise a fabricated PDMS-based scaffold with SU-8 microchannel walls, wherein the bottom PDMS layer is approximately 50μιη thick, the SU-8 microchannel walls are approximately 20μιη wide and ΙΟΟμιη thick, and the microchannel width is approximately 50μιη.

DRG's were cultured for one week on open scaffolds. The sample was stained for axons and Schwann cells, and robust growth and proliferation of axons and Schwann cells extending through the microchannels was observed. The axons and Schwann cells alike were aligned and oriented within the microchannels as opposed to radially spreading out where there were no microchannels. The present PDMS -based microchannel scaffold guides DRG neurite outgrowth along the microchannels. This microchannel scaffold design provides an efficient way to guide regenerating axons, and can be used as a new platform to incorporate electronics for recording and stimulation specifically from small groups of axons. By integrating a microelectrode in each microchannel to form a high-throughput electrode array, such a microchannel-scaffold electrode array significantly enhances the efficacy and reliability of peripheral nerve interfacing.

In in vivo studies, the present PDMS/SU-8 scaffolds of varying microchannel dimensions were implanted in a rat sciatic nerve amputee animal model to validate and assess levels of regeneration axons and supporting cell type regeneration through the devices. This characterization to validate regeneration proved successful.

The present invention can further comprise a fabricated PDMS-base scaffold with SU-8 microchannel walls. In these scaffolds, the bottom PDMS layer is approximately 50μιη thick, the SU-8 microchannel walls are approximately 20μιη wide and ΙΟΟμιη thick, the top PDMS layer is approximately ΙΟμιη thick, and the microchannel widths are approximately 50, 100, or 150μιη. These scaffolds were implanted for eight weeks, at which point they were explanted. The tissue inside the microchannels of the explanted scaffolds was analyzed for cell nuclei, axons, and Schwann cells. Robust growth and integration of tissue as viewed by staining for the three protoplasmic components of interest was confirmed. The scaffolds were deemed as reliable and robust vehicles to guide, house, and support regenerating axons in a chronic setting to the benefit of a neural interfaces with electrodes integrated into the microchannels.

In additional in vivo studies, the present PDMS/SU-8 scaffold of approximately 150μιη microchannel width can further comprise microwires integrated into the microchannels serving as microelectrodes. This neural interfacing device was implanted in a rat sciatic nerve animal model with the purpose of recording signals from axons that regenerated through the microchannels containing the microwires. The use of this device to record signals from regenerated axons proved to be successful.

The present invention can further comprise a fabricated PDMS-based scaffold with SU-8 microchannel walls, wherein the microchannels have integrated microwires of approximately 60μιη in diameter. These microwires can be insulated except at the very tip of the wire that resides in approximately the middle of the length of the microchannel. These microwires can extend from the device implanted in the sciatic nerve to lie just under the skin of the rat for the duration of the implantation. At the end of 12 weeks the skin is opened, the wires exposed and connected to recording equipment for the recording of sensory induced and spontaneous single unit action potentials from regenerated axons. It is believed that this embodiment of the present invention is the first microchannel-based device that has successfully recorded spontaneous single unit action potentials from regenerated axons.

The present microchannel neural interface has proven capabilities to record spontaneous and induced action potentials from regenerated axons in vivo. The integrated microelectrodes provide an easy and direct connection to on-board electronics that provide the highly advantageous utilities of amplification and multiplexing.

The present scaffolding material can be porous for increased nutrient exchange into the microchannels of the device.

The present invention can further comprise microfluidics for nutrient and growth factor delivery into the microchannels of the device.

The present invention can further comprise acellular nerve grafts into the microchannels to increase and stabilize axon regeneration into the microchannels of the device.

The present invention can further comprise chemical and/or biological factors to increase and stabilize regeneration. Examples include, among others, laminin and/or nerve growth factor (NGF). These factors additionally could be contained within degradable or non-degradable hydrogels, nanoparticles or other encapsulants.

The present invention can further comprise chemical and/or biological factors to allow the separation of motor and sensory axons. This would allow certain microchannels to house motor axons and other microchannels to house sensory axons that could then be interfaced with in a more specific manner.

The present invention can further comprise cells including stem cells and/or Schwann cells into the microchannels for increasing and stabilizing axon regeneration.

The present invention can further comprise a plurality of electrodes in each channel. The additional electrode(s) could be used as a reference electrode, for unidirectional stimulation, and for stimulation and conduction blocking purposes in conjunction with each other. The present invention can further comprise an electrical shielding cage or wire-like structure around the implant to serve a purpose similar to that of a Faraday cage and reduce noise and/or disruptive electro myogram (EMG) signals from muscles.

The present invention can further comprise wireless components into the On board' electronics, obviating the need for wires leading to a purcutaneous headcap. This would require the addition of a trancutaneous power source or an implantable battery for powering the recording and wireless transmitting unit.

The present invention can be used in neural prosthetic interfacing, allowing amputees to control a prosthetic device by recording from nerves and acquire sensation from the prosthetic device by stimulating nerves. This includes prosthetic limbs and retinal prosthetics among others. The device could be implanted in either peripheral nerves, the spinal cord, and/or the optic nerve.

The present invention can be used in functional electrical stimulation (FES), allowing individuals suffering from various disabilities to regain function. This includes individuals suffering from many types of paralysis stemming from a spinal cord injury and could restore limb movements as well as bowel and bladder control. This can alleviate individuals suffering from the 'foot drop' syndrome

The present invention can be used in conduction blocking, allowing for pain modulation by controlling nerves that are conducting pain signals to the brain or that have aberrant activity resulting in pain signals.

The present invention further can be used as a tool by researchers and scientists to determine function of nerves and axons as they relate to various sensory and/or motor functions. Additionally, the present invention can be used to study how axons, their behavior, and how they interact with other tissues, change over time and after an injury. The present device, in conjunction with tools to assess cortical activity, can also be used to study how information in the periphery as coded by axons is translated to cortical activity.

The present invention can be used in measuring impedance, allowing doctors to monitor nerve regeneration in an individual with nerve damage. This would allow doctors to monitor bone regeneration in an individual with a bone fracture. This would allow doctors to monitor skin regeneration/growth in the subdural layers not visible to the human eye.

The present invention is a vast improvement over conventional device that use polyimide as the substrate material. The general design of the conventional polyimide device includes just the bottom layer. A major difference between such a design and the present invention is the addition of a top PDMS layer. This fully closes the microchannels. A benefit is complete electrical isolation between neighboring microchannels. If this top layer is not present, then neighboring microchannels can share the same low impedance extracellular fluid and matrix surrounding the axons. This leads to increased noise and a decreased ability to differential signals from different axons. The overall result is a decreased ability to specifically and selectively interface with individual and small groups of axons.

Yet another difference between the conventional polyimide device and the present invention is the use of PDMS as the substrate material and SU-8 as the microchannel wall/structural material. PDMS is a highly conformable and flexible material, and much closely matches the mechanical properties of a nerve.

The SU-8 provides the structure needed for device integrity while maintaining the advantages of the PDMS. This choice of materials allows the present invention to better conform to regenerating axons and provide a scaffold for regeneration that closely matches the nerves natural environment. Other materials, such as polyimide, are rigid in comparison and do not provide the need flexibility and conformability to maintain a regenerating nerve over chronic periods of time.

The use of PDMS, a highly hydrophobic material, also limits and/or prevents the accumulation of scar tissue on the implant, greatly benefitting the ability of the device to operate in a chronic setting. Those of skill in the art know that devices that utilize materials such as polyimide accumulate scar tissue over time, leading to increases in electrode impedance and a narrowing of microchannels. This results in the inability to record small amplitude action potentials, an unfavorable environment for axon growth leading to axon death, and inevitable device failure in a chronic setting. The present use of PDMS circumvents this issue because it prevents scar tissue accumulation on its surface. Another patentable distinction between the conventional polyimide device and the present invention is the use of on-board electronics. This is a huge design advantage in that it allows for a significantly less invasive testing platform in an animal model. Recordings can be more reliably taken and stimulation can be performed while the rats are awake and over a chronic period of time.

Spontaneous signals generated from actual limb movement is recorded as opposed to just evoked signals by applying an electric current farther upstream to artificially stimulate the nerve and then recording the nerve's response. In terms of the design for human implantation, the onboard electronics allows for signal amplification directly after signal acquisition. This significantly increases the signal-to-noise ratio and greatly increases the ability to record specifically and selectively from axons. This platform also allows the direct integration of wireless components.

In an exemplary embodiment, the present invention is a regenerative neural interface to establish high-channel, bi-directional communication between an amputated nerve and a device comprising a thin-film polydimethylsiloxane (PDMS)/SU-8 based regenerative microchannel scaffold having microchannels, microelectrodes incorporated in the microchannels, and an integrated multiplexer/amplifier for large channel recordings. When in use, the amputated nerve can comprise groups of axons and the interface provides intimate contact between the microelectrodes and groups of axons. The microchannels confine axon growth to limit the volume of low impedance extracellular fluid and matrix surrounding the axon. The interface can further comprise a shield surrounding at least a portion of the interface, wherein when in use, the shield reduces noise and disruptive electromyogram (EMG) signals from axons. The scaffold can comprises a bottom layer comprising PDMS and walls comprising SU-8 polymer extending upward from the bottom layer, wherein the bottom layer and the walls form the microchannels. The bottom layer can have a thicknesses in the range of approximately ΙΟμιη and 40μιη. The walls can have a height of approximately ΙΟΟμιη. The walls can have a width of approximately 20μιη.

In an exemplary embodiment, the present invention is a regenerative neural interface to establish high-channel, bi-directional communication between a protoplasmic protrusion and a device comprising a bottom layer comprising PDMS, and a plurality of walls comprising SU-8 polymer extending upward from the bottom layer, the walls running approximately parallel to one another, forming microchannels, the bottom layer having a thicknesses in the range of approximately ΙΟμιη to 40μιη, the walls having a width of approximately 20μιη and a height of approximately ΙΟΟμιη, and the widths of at least a portion of the microchannels are in the range of approximately 50μιη to 150μιη. The interface can further comprise at least one microelectrode incorporated in at least a portion of the microchannels. The interface can further comprise an integrated multiplexer/amplifier. The interface can further comprise a shield surrounding at least a portion of the interface. The microchannels can have a length of approximately 3mm.

In an exemplary embodiment, the present invention is a rolled regenerative neural interface to establish high-channel, bi-directional communication between a protoplasmic protrusion and a device comprising a bottom layer comprising PDMS, and a plurality of walls comprising SU-8 polymer extending upward from the bottom layer, the layer and walls forming a layer of microchannels, that when rolled, form a scaffold with microchannels closed along their length. The bottom layer can have a thicknesses in the range of approximately ΙΟμιη to 40μιη, the walls can have a width of approximately 20μιη and a height of approximately ΙΟΟμιη, and the widths of at least a portion of the microchannels can be in the range of approximately 50μιη to 150μιη. The rolled regenerative neural interface can have a total radius of 1.5mm.

In an exemplary embodiment, the present invention is a method of fabricating a regenerative neural interface. The method can comprise providing a PDMS bottom layer and walls forming microchannels, and patterning gold electrodes into the PDMS bottom layer with exposed portions in the microchannels forming a conformable microelectrode array (cMEA). The method can further comprise connecting a channel amplifier/multiplexer chip to the interface.

In an exemplary embodiment, the present invention is a method of fabricating a regenerative neural interface comprising providing a bottom layer comprising PDMS, providing a top layer comprising PDMS, and providing walls comprising SU-8 polymer between the bottom and top layers, forming microchannels.

In an exemplary embodiment, the present invention is a neural interface comprising a base layer comprising a polymeric organosilicon compound, and at least two channel walls comprising phenol formaldehyde resin, the walls and base layer forming a channel, wherein the channel accommodates regenerating a protoplasmic protrusion. The channel can comprise an electrical conductor for communication with a protoplasmic protrusion. The polymeric organosilicon compound can comprise polydimethylsiloxane (PDMS). The phenol formaldehyde resin can comprise an epoxy-based negative photoresist. The phenol formaldehyde resin can comprise SU-8 polymer. The PDMS provides a flexible and stretchable scaffold, and the SU-8 polymer provides mechanical integrity to the microchannels.

The present interface can be used in a variety of environments and ways, including to control machines other than prosthetic limbs.

Other aspects and features of embodiments of the present invention will become apparent to those of ordinary skill in the art, upon reviewing the following detailed description in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF THE FIGURES

The various embodiments of the invention can be better understood with reference to the following drawings. The components in the drawings are not necessarily to scale, emphasis instead being placed upon clearly illustrating the principles of the various embodiments of the present invention. In the drawings, like reference numerals designate corresponding parts throughout the several views.

Fig. 1 is a schematic of a FINE peripheral nerve interface.

Fig. 2(a) is an micrograph of USEA, and Fig. 2(b) is an example of physical damage to nerve caused by USEA insertion (scale bar is 50μιη).

Fig. 3 is a schematic of a LIFE peripheral nerve interface.

Fig. 4 is a micrograph of a Regenerative Sieve peripheral nerve interface.

Fig. 5 is an illustration of isolated axons regenerating through microchannels.

Fig. 6 is an exploded view of the overall microchannel assembly of the present invention according to a preferred embodiment.

Fig. 7 is a top view of a portion of the microchannel assembly of Fig. 6.

Fig. 8 depicts an exemplary embodiment of the present scaffold after rolling. Figs. 9(a)-(c) illustrate an embodiment of the present invention functionally evaluated in vivo through the stimulation of and single unit recordings from regenerated axons in a rat sciatic nerve amputee model.

Fig. 10 illustrates the fabrication process (not to scale) for the bottom PDMS layer and SU-8 microchannel walls of the regenerative microchannel scaffold, according to a preferred embodiment of the present invention.

Fig. 11 is a micrograph of the bottom PDMS layer and SU-8 microchannels walls of the present invention according to exemplary embodiments.

Fig. 12 illustrates the fabrication process of adding the top PDMS layer to complete the regenerative microchannel scaffold.

Figs. 13(a)-(c) are images of the regenerative microchannel scaffold in a unrolled and rolled configuration showing how the top PDMS layer contacts the bottom PDMS layer once rolled.

Fig. 14 illustrates the fabrication process of patterning titanium and gold electrodes and traces using an NR4-8000P photoresist in a photolithography and lift-off based process.

Figs. 15(a)- (c) illustrates the actual electrode, trace, and bonding pad design for the present invention along with how two different electrode referencing paradigms are incorporated into the device. The electrode region is highlighted in orange, the bond pad region is highlighted in red, and the traces connect the two regions.

Fig. 16 is a micrograph of titanium/gold electrodes, traces, and bonding pads patterned on a base PDMS layer using the NR4-8000P photolithography and lift-off process.

Fig. 17 illustrates the fabrication process of insulating the titanium/gold traces using an NR4-8000P photoresist in a sacrificial-post based process.

Fig. 18 depicts a side view of the present invention after the complete fabrication process and bonding of electronic components including the chip for amplification and multiplexing.

Fig. 19 is a micrograph view from the top of the microchannels with incorporated titanium/gold electrodes and traces prior to being rolled. Fig. 20 is a micrograph of the present invention with after the microchannels have been rolled with incorporated titanium/gold electrodes and traces, and incorporated electronics for multiplexing and amplification.

Fig. 21 illustrates the present invention ready for implantation, after it has been encased in an outer tubing to aid with nerve suturing to the front of the microchannels, after being encased in an electrical shield comprising a cage or wiring to reduce noise and disruptive EMG signals, and after attaching signal extraction wires to the chip.

Figs. 22(a)-(c) are fluorescent micrographs of an in vitro dorsal root ganglia (DRG) culture on the PDMS/SU-8 microchannels. In these images axons are shown as red, and Schwann cells are shown as green. Sub-images a-c show the axons and supporting Schwann cells proliferated extremely well, were aligned and oriented within the microchannels, and even seemed to grow towards the microchannels.

Fig. 23 is an illustration of the rat sciatic nerve amputee animal model used for testing in the present invention. The animal model comprises first transecting the sciatic nerve. The device is sutured to the proximal and distal nerve stumps so that regenerating axons grow through the device in response to cues from the distal end. After suturing in the device, the distal nerve is again transected approximately 2mm distal to the end of the device and a portion of the nerve excised. This leaves what we term a 2mm 'distal nerve stump' on the distal end of the device. A portion of the nerve is excised in order to prevent regenerating axons from reinnervating their original targets. Signal extraction wires are then tunneled subcutaneously along the back of the rat to a percutaneous headcap where they can be connected to wires leading to an electrophysiology computer.

Fig. 24 is a characteristic fluorescent micrograph of the present invention comprising a cross-section of a regenerative microchannel scaffold with 150μιη wide microchannels after being implanted for eight weeks in a rat sciatic nerve amputee animal model. In this image, cell nuclei are shown in blue, axons are shown in red, and the support Schwann cells are shown in green. It is clear from the image that axons and Schwann cells populated the microchannels well.

Fig. 25 is a characteristic fluorescent micrograph of the present invention comprising a cross-section of a regenerative microchannel scaffold with ΙΟΟμιη wide microchannels after being implanted for eight weeks in a rat sciatic nerve amputee animal model. In this image, cell nuclei are shown in blue, axons are shown in red, and the support Schwann cells are shown in green. It is clear from the image that axons and Schwann cells populated the microchannels as well as in the 150μιη wide microchannels. This provides concrete proof that using this type of scaffold will greatly benefit a microchannel based neural interface.

Fig. 26 is a characteristic fluorescent micrograph of the present invention comprising a cross-section of a regenerative microchannel scaffold with 50μιη wide microchannels after being implanted for eight weeks in a rat sciatic nerve amputee animal model. In this image, cell nuclei are shown in blue, axons are shown in red, and the support Schwann cells are shown in green. It is clear from the image that axons and Schwann cells did not populate the microchannels well providing concrete proof that microchannels of this are not usable in a microchannel based neural interface.

Fig. 27 is a micrograph of the present invention, placed on a dime, used for a terminal experiment in a rat sciatic nerve animal model. In this case, the present invention was integrated with microwires serving as microelectrodes in regenerative microchannel scaffolds with 150μιη wide microchannels.

Fig. 28 shows spontaneous action potential recording data as obtained through the present invention integrated with microwires while the animal was at rest. The left column of images shows raw data collected using the microwires from two representative channels (1 and 7) showing spontaneous action potentials clearly visible over the background noise with a SNR of approximately 2: 1. The middle column shows the shape of individual single unit spontaneous action potentials isolated from the two representative channels. The right column shows the inter- spike-interval of these spontaneous action potentials showing that top action potential form channel 1 was highly periodic along with the action potential from channel 7. This is in contrast to the other two action potentials form channel 1 which had highly sporadic firing rates.

Fig. 29 shows the firing rate of action potentials in channel 7 at rest (top image) and in response to tactile stimulation (bottom image) of the 4th knuckle on the rat's paw. It can be clearly seen that the firing rat increased from approximately 3Hz to approximately 9Hz in response to the sensory stimulation. DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS Although preferred embodiments of the invention are explained in detail, it is to be understood that other embodiments are contemplated. Accordingly, it is not intended that the invention is limited in its scope to the details of construction and arrangement of components set forth in the following description or illustrated in the drawings. The invention is capable of other embodiments and of being practiced or carried out in various ways. Also, in describing the preferred embodiments, specific terminology will be resorted to for the sake of clarity.

It must also be noted that, as used in the specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless the context clearly dictates otherwise.

Also, in describing the preferred embodiments, terminology will be resorted to for the sake of clarity. It is intended that each term contemplates its broadest meaning as understood by those skilled in the art and includes all technical equivalents which operate in a similar manner to accomplish a similar purpose.

Ranges may be expressed herein as from "about" or "approximately" one particular value and/or to "about" or "approximately" another particular value. When such a range is expressed, another embodiment includes from the one particular value and/or to the other particular value.

By "comprising" or "comprising" or "including" is meant that at least the named compound, element, particle, or method step is present in the composition or article or method, but does not exclude the presence of other compounds, materials, particles, method steps, even if the other such compounds, material, particles, method steps have the same function as what is named.

It is also to be understood that the mention of one or more method steps does not preclude the presence of additional method steps or intervening method steps between those steps expressly identified. Similarly, it is also to be understood that the mention of one or more components in a device or system does not preclude the presence of additional components or intervening components between those components expressly identified.

Regenerative MicroChannel Scaffold

In an exemplary embodiment of the present invention, Figs. 6 and 7 illustrate a neural interface 10 comprising a base layer 14 comprising a polymeric organosilicon compound and at least two channel walls 16 comprising phenol formaldehyde resin, the walls 16 and base layer 14 forming a channel 20, wherein the channel 20 accommodates regenerating a protoplasmic protrusion 22.

A channel is a passage or tunnel forming a conduit through which an axon can regenerate or grow. The channel is open at both ends to allow the entry and exit of regenerating nerve axons, and need not be closed along its length (have a top), but in an exemplary embodiment, the channel is closed to surround the regenerating nerve axons within the channel. The channel can include an number of cross-sectional shapes, including having a round (e.g. oval or circular) cross-section, or an angular (e.g. square or rectangular) cross- section.

The channel 20 can comprise an electrical conductor 18 for communication with the protoplasmic protrusion, the electrical conductor preferably being an electrode. The polymeric organosilicon compound can comprise polydimethylsiloxane (PDMS), and the phenol formaldehyde resin can comprise an epoxy-based negative photoresist, preferably SU-8 polymer.

In an exemplary embodiment, the neural interface 10 comprises microchannels 20 composed of top 12 and bottom 14 PDMS layers, and SU-8 polymer walls 16 with incorporated gold electrodes 18.

The PDMS top and bottom layer 12, 14 thicknesses Ti and T₂, respectively, can be approximately ΙΟμιη and 40μιη, respectively. The microchannels walls 16 can have a width W₂ and height He of approximately 20μιη and ΙΟΟμιη, respectively. The widths W3 of the microchannels 20 can range from approximately 50μιη to 150μιη, and with the optimal being ΙΟΟμιη as determined from in vivo regeneration studies. The microchannel 20 lengths Li + L₂ can range from approximately .95mm to 5mm.

Finally, the layer of microchannels formed as the interface 10 can be rolled to form an overall scaffold 30, with a total radius HR of approximately 1.5mm, as shown in Fig. 8.

The implantation setup shown in Figs. 9(a)-(c) involves mounting the proximal stump of a transected sciatic nerve to one end of the scaffold 30 and allowing the nerve 22 to disassemble and regenerate through the microchannels 20 as shown in Fig. 6. The regenerating axons in the nerve are expected to integrate with the microelectrodes 18 inside the microchannels 20. The microelectrodes 18 can be wired to the integrated Όη-board' electronics by a short PDMS cable. The Όη-board' electronics will help to reduce recording noise and power for stimulation, as well as allow for multiplexing that will reduce the number of wires needed. The VOs of the electronics can be wired to a percutaneous head stage through subcutaneous wires.

The one or more electrodes can, for example, detect extracellular electrical signals in the microchannel when an axon in the microchannel transmits an action potential or generate extracellular electrical signals in the microchannel that stimulate an action potential in the axon.

Extracellular electrical signals produced when an axon in the microchannel transmits an action potential may be detected by one or more electrodes exposed to the interior of the microchannel (i.e. a recording array). The one or more electrodes therefore allow information relating to action potentials in the axon to be gathered and analyzed. Action potentials from motor and sensory axons may be recorded and analyzed in this way. Information gathered from a motor axon may be amplified, recorded and processed and may be used for a range of purposes, for example, to drive a prosthetic limb, to supply information to a muscle stimulator downstream of the site of injury or to study nerve traffic in the regenerated nerve fiber. One or more electrodes coupled to the microchannel may be useful in transmitting artificial action potentials to an axon in the microchannel by means of capacitive coupling and/or the passage of Faradaic currents between the electrode surface and the axon (i.e. a stimulation array). Artificial action potentials may be produced, for example, in response to input from sensors, such as temperature or pressure sensors, and transmitted via the electrode to a sensory axon in the microchannel, to provide sensory input to the nervous system.

Preferably, the interior of the microchannels provides an environment that promotes and supports growth of one or more of axons, glial cells and blood vessels within the microchannel.

In some embodiments, the microchannels can contain a gel or other drug delivery vehicle for the sustained display or controlled release of chemical and/or biological factors. This provides a structure that encourages axon growth and glial invasion. A gel comprises a matrix of fibers contained in an interstitial electrically conductive medium. Suitable gels can be formed from fibrous materials such as Matrigel™ (BD Biosciences), collagen, agarose, haluronic acid, and spider silk using standard techniques. In some embodiments, the microchannels can comprise one or more biological factors that promote growth of one or more axons, glial cells and blood vessels. The biological factors can be incorporated in a gel within the microchannel or attached to the inner walls of the microchannel. Biological factors include promoters of axon growth and Schwann cell invasion, such as laminin, fibronectin, collagen type 4, and heparan sulphate proteoglycans, or molecules comprising the axon growth promoting domains of those molecules. The microchannels can contain a neurotrophic factor such as nerve growth factor, brain derived neurotrophic factor and/or neurotrophin 3. To promote Schwann cell division and migration, Schwann cell growth factors such as neuregulins can be employed. For oligodendrocyte precursor migration and division, growth factors such as fibroblast growth factor 1 and platelet derived growth factor can be employed. To promote the ingrowth of blood vessels, growth factors such as vascular endothelial growth factor (VEGF) cam be employed. Other biological factors, such as neutralizing antibodies to TGFbeta, decorin and corticosteroids, can be employed to limit inflammation and scarring. Additionally factors such as interleukin 4 (IL 4) that affect and control the type and amount of immune response after the injury can be incorporated.

In use, the microchannels accommodate growing axons in an extracellular fluid which is naturally produced by the surrounding tissue. This fluid provides an ionic conductive medium that contacts the one or more electrodes and allows the detection of action potentials propagating in the axon by the one or more electrodes. The microchannels can also accommodate glial cells, such as Schwann cells, and blood vessels.

Regenerative Microchannel Scaffold Fabrication

In a preferred fabrication of the base PDMS layer and SU-8 microchannel walls of the present regenerative microchannel scaffold, representatively shown in Fig. 10, a 40μιη PDMS (Sylgard 184, Dow Corning) base layer was first spun on glass, in which an MEA would be fabricated if electrodes were to be incorporated into the scaffold. The PDMS base layer was treated with oxygen plasma to increase the adhesion between PDMS and SU-8. A ΙΟΟμιη layer of SU-8 (SU-8 2100, MicroChem Corp) was then spun on top of the PDMS. The SU-8 was cured, exposed, and developed forming the patterned microchannel walls on the PDMS base layer. The width and length of the microchannel walls were approximately 20μιη and 10mm respectively. The width of the microchannels ranged from approximately 50-150μιη. An example of the resulting base PDMS layer with SU-8 microchannel walls forming open microchannels with a 75μιη width is depicted in Fig. 11. In this image, the SU-8 microchannel walls are reflecting on the clear base PDMS layer.

In a preferred fabrication process of adding the top PDMS layer, polyacrylic acid (PAA) is first spun on another glass slide and dried. This is done twice. A ΙΟμιη layer of PDMS is immediately spun on the PAA layers and partially cured. The bottom PDMS layer with the SU-8 microchannel walls is treated with oxygen plasma to increase the adhesion between the two layers. The two glass slides are place together with weight on top and baked until the top PDMS layer is fully cured. Finally, the glass slide sandwich is soaked in water until the PAA dissolves and allows an easy removal of the top glass slide. The simplified process is depicted in Fig. 12. An example of the resulting closed microchannels each with 150μιη widths is shown in Fig. 13(a) where the top PDMS layer, SU-8 microchannel walls, and bottom PDMS layer can be clearly seen. Bonding between the PDMS cover layer and the SU-8 channel walls can be seen by the lack of a visible division between the SU-8 and PDMS. This validates the present fabrication process design and its ability to produce microchannels in PDMS using SU-8 walls.

Fig. 13(b) depicts a microchannel scaffold with a ΙΟΟμιη width after it has been rolled to form the implantable tubular construct. Fig. 13(c) shows a close up of the microchannels and clearly shows where the top PDMS layer of one of the microchannel layers meets the bottom PDMS layer of another microchannel layer.

Based on these results, the present invention has been successfully fabricated as an regenerative microchannel scaffold using PDMS as the base and cover layers, and SU-8 as the microchannel walls. The devices capability of being rolled to form a three-dimensional scaffold has also been verified.

Titanium and Gold Electrode, Trace, and Bond Pad Fabrication

A preferred fabrication process for incorporating the titanium and gold electrodes, traces, and bond pads into the bottom PDMS layer is illustrated in Fig. 14. Starting with the base PDMS layer of approximately 40μιη in thickness, a photoresist that is preferably NR4-8000P is spun on. This is cured by baking in an oven before exposing and developing the NR4-8000P according to the electrode, trace, and bond pad design. This design is illustrated in Fig. 15 and also depicts how two different referencing paradigms are incorporated into the device. These referencing paradigms are bi-polar and tri-polar and allow a user to choose in real-time which to use in order to best reduce noise and improve action potential recordings.

In Fig. 15 the electrode region is highlighted in orange, the bond pad region is highlighted in red, and the traces connect the two regions. After developing the NR4-8000P according to the electrode, trace, and bond pad design, titanium is first deposited using preferably an E-Beam Evaporator to serve as an adhesion layer between the gold and PDMS base. Gold is then deposited preferably using the method on top of the titanium. Finally, the remaining NR4-8000P is stripped away leaving behind the titanium/gold electrodes, traces, and bond pads. An example of this is shown in Fig. 16 where three micrograph images of the electrodes, traces, and bond pads have been stitched together to give a complete high-resolution image.

Titanium/Gold Trace Insulation

A preferred fabrication for insulating the titanium/gold traces is illustrated in Fig. 17 and starts with a base PDMS layer with electrodes, traces and bond pads already patterned on it. First the sample is treated with oxygen plasma to increase the adhesion with the NR4-8000P which is then spun on immediately. The NR4-8000P is then cured, exposed, and developed so that all that remains are posts covering the electrodes and bond pads on the sample. At this point, the sample is again treated with oxygen plasma to increase the bonding of the base layer with the PDMS insulation layer. After oxygen plasma treatment, the PDMS insulation layer, approximately ΙΟμιη thick, is immediately spun on and cured. The sample is then plasma etched using preferably oxygen and carbon-tetrafluoride to remove the possible thin layer of PDMS covering the NR4-8000P sacrificial posts. Finally, the sacrificial posts are stripped away using preferably Resist Remover 41 leaving behind titanium/gold PDMS insulated traces with openings in the insulation corresponding to the electrodes and bond pads.

After the PDMS insulation of the traces is completed, the SU-8 microchannel walls and top PDMS layer can be added to the sample. A micrograph of the resulting structure is shown in Fig. 18. This image has been taken as viewed through the top PDMS layer where one can clearly see the SU-8 microchannel walls, microchannels, and titanium/gold electrodes and traces. Incorporation of Electronics for Signal Amplification and Multiplexing

A preferred fabrication for incorporating the electronics of the present invention involves placing conductive epoxy on the bonding pads of the electronics, whether that is a PCB, chip or some other electrical component. This is laid with the bonding pads facing up. The PDMS sample is then flipped so its bonding pads are facing down. The bonding pads are lined up under a microscope and placed gently together so the conductive epoxy is connecting the bonding pads on the PDMS substrate to the bonding pads on the electronic components. Finally the sample is cured at length to ensure the conductive epoxy has firmly bonded to both sets of bond pads.

A schematic of a device after this step is shown in Fig. 19 where the image is a cross- section of the device showing the electronic components, conductive epoxy, titanium/gold, and microchannels prior to being rolled. A micrograph of the device after these fabrication steps is shown in Fig. 20. In this image, the microchannels have been rolled and the chip for signal amplification and multiplexing has been bonded to the bonding pads on the PDMS substrate using conductive epoxy which is not visible.

Final Device Preparation for Implantation

A preferred process for preparing the device for implantation, illustrated in Fig. 21, first involves placing the rolled microchannels inside an outer tube that is longer than the microchannels by preferably at least 1mm on both ends. This tube serves allows the nerve to be sutured into the tube so that it is directly facing the opening of the microchannels. In addition to the outer tube, a grounded electrical shield preferably comprising a cage or wire is placed around the implant to reduce noise and local EMG signals in the same way a Faraday cage helps reduce noise. Finally, signal extraction wires are connected to the electronics.

Dorsal Root Ganglia Culturing In Vitro Experiment

Regenerative microchannel scaffolds lacking the PDMS cover layer were used for the in vitro experiments. Once fabricated, the open regenerative microchannel scaffolds were placed at the bottom of tissue culture wells. Dorsal root ganglia (DRG's) were explanted from the spinal cords of PI rat pups. The nerve roots were removed and the DRG's were placed on the open scaffolds at the entrance to the microchannels. For the first several hours, the DRG's were incubated with only a thin layer of DMEM/F12 media with approximately 10% FBS and 50 ng/mL nerve growth factor (NGF) (Roche). Afterwards, the wells were fully covered with the same media. The media, including NGF, were replaced every two days for a total of seven days. After seven days the DRG's were fixed with approximately 4% paraformaldehyde in PBS for approximately 20 minutes and washed three times with IX PBS.

To visualize neurite outgrowth and non-neuronal cell migration, axons and Schwann cells were stained overnight at approximately 4°C with the primary neurofilament 160 kDa (NF160, 1:500, mouse IgGl, Sigma) and primary S-100 respectively. The secondary antibodies goat anti- mouse IgGl Alexa 488/594 and goat anti-rabbit IgG Alexa 488/594 were used respectively. Cell nuclei were labeled with DAPI (10μΜ, Invitrogen). The fluorescently labeled cells and nuclei were visualized using a Zeiss upright microscope and the images were captures with an Olympus digital camera.

A micrograph of the scaffolds lacking the PDMS cover layer fabricated for in vitro DRG culturing is shown in Fig. 11. To reiterate, this open scaffold has microchannel widths of 75μιη. The cultured DRGs adhered well to these open scaffolds.

Fig. 22, a fluorescent micrograph, shows an example DRG (the large bright circle) cultured on a scaffold with 50μιη microchannel widths. The SU-8 microchannel walls auto- fluoresce while the microchannels themselves appear as dark horizontal lines and are positioned to the left of the DRG. The main image shows two overlaid images of axons (shown in red) and non-neuronal Schwann cells (shown in green), while the subsets show the separate fluorescent images. When the axons and Schwann cells overlap in the main image, they appear as orange. As seen from the main image, both the neurites and Schwann cells grew and proliferated in a robust manner.

Fig. 22(a) and Fig. 22(b) show that axon extension and Schwann cell migration were aligned and oriented along the direction of the microchannels. Fig. 22(c) shows that the axons and Schwann cells actually extended processes and migrated towards the microchannels in some cases. It should also be noted that there does appear to be more neurite extension and Schwann cells growth on the SU-8 channel walls as opposed to on the PDMS. However, this will not negatively affect the scaffold in vivo because in a preferred embodiment, there is a top PDMS layer and once inside the channel, it does not matter which wall the axons and cells grow on.

Overall these results illustrate that the substrate and structural materials of the present microchannel scaffold are biocompatible and can support the growth of multiple cells types, DRG axons and non-neuronal cells. Furthermore, the capability of the microchannel design to guide and direct DRG axon outgrowth and non-neuronal cell migration along and through the microchannels has been verified. These results show that the present microchannel design provides an extremely robust method to guide regenerating axons and can be used as a fundamentally sound platform to incorporate electronics for chronic recording and stimulation of axons.

In Vivo Regeneration Experiment in a Rat Sciatic Nerve Amputee Animal Model

Regenerative microchannel scaffolds shown in Fig. 13(b) were used for the in vivo experiments. These scaffold had widths of approximately 50, 100 or 150μιη. This experiment was to validate and assess levels of regeneration axons and supporting cell type regeneration through the regenerative microchannel scaffolds. Additionally, this experiment was to identify the microchannel dimension that hold the fewest number of regenerating axons while ensuring that a majority of microchannels actually contain axons when the varying parameter is the microchannel width.

The three variations of widths tested were approximately 50, 100, or 150μιη. The scaffolds were implanted in a rat sciatic nerve amputee model for approximately eight weeks. The specific animal model comprises first transecting the sciatic nerve. The device was sutured to the proximal and distal nerve stumps so that regenerating axons grow through the device in response to cues from the distal end. After suturing in the device, the distal nerve was again transected approximately 2mm distal to the end of the device and a portion of the nerve excised. This leaves what we term a 2mm 'distal nerve stump' on the distal end of the device. A portion of the nerve was excised in order to prevent regenerating axons from reinnervating their original targets. The implantation model is illustrated in Fig. 23.

After the regenerative microchannel scaffolds were explanted, they were fixed for approximately two hours in approximately 4% paraformaldehyde. To prepare the scaffolds for cryosectioning, they were transferred to an approximately 30% sucrose in PBS solution and incubated at approximately 4°C for one-two days until saturation. Finally, the samples were embedded in O.C.T. gel and frozen for cryosectioning with a Leica CM30505 cryostat. Cross- sections of the scaffold were taken in the middle of the scaffold corresponding to the region that holds electrodes. This region was analyzed for axons numbers per microchannel. The present invention needed to be sectioned at very thick intervals (approximately ΙΟΟμιη) in order to maintain sample integrity. This is due mainly to the flexibility of PDMS and the fact that this flexibility is not reduced even at temperatures as low as -50°C. When the blade of the cryosectioner contacts the PDMS, instead of getting cut, the PDMS flexed out of the way causing inconsistencies in the sectioned sample. The easiest way to mitigate these effects was to section thicker samples. Another method developed was sectioning under liquid nitrogen.

In order to characterize the regeneration through the microchannels, the sections taken were immunohistologically stained for markers of axons (NF-160), Schwann cells, (S-100), and cell nuclei (DAPI). The basic staining process involved first incubating the sections for approximately one hour at room temperature in a blocking solution of goat serum in PBS, then incubating overnight at approximately 4°C in a mixture of primary antibody and blocking solution. Next, the sections were washed and incubated for approximately one hour at room temperature in a solution of secondary antibody (goat anti-rabbit IgG Alexa 488/594, and goat anti-mouse IgGl Alexa 488/594) mixed in 0.5% triton in PBS. Finally, the sections were be washed, dried, and cover slipped for evaluation under a Zeiss LSM 510 NLO confocal microscope.

Fluorescent micrographs of representative sections of the regenerative microchannel scaffolds with microchannels widths of approximately 150, 100, or 50μιη are shown in Fig. 24, Fig. 25, and Fig. 26, respectively. In these images, axons are shown in red, Schwann cells are shown in green, and cell nuclei are shown in blue. For the most part, the scaffold itself is black, however in some cases the SU-8 auto-fluoresces and appears a faint orange or blue depending on the figure. It can be clearly seen in both Fig. 24 and Fig. 25 that the 150 and ΙΟΟμιη microchannels support the robust growth and integration of tissue including importantly axons and Schwann cells. A majority of microchannels in both microchannel widths have axon regeneration within them to the benefit of neural interfacing. This is in contrast to the 50μιη microchannel width which clearly does not support regenerating axons and Schwann cells as seen in Fig. 26.

This experiment concretely validates that ΙΟΟμιη wide microchannel act as reliable and robust vehicles to guide, house, and support regenerating axons in a chronic setting to the benefit of a neural interface with electrodes integrated into the microchannels. Finally, this experiment also point towards the fact that the ΙΟΟμηι wide microchannel is an optimal channel width to use because it most effectively reduces the low impedance extracellular fluid while still maintaining axon regeneration in a majority of microchannels. In Vivo Spontaneous and Sensory Evoked Action Potentials Recordings

The purpose of this experiment was to record spontaneous and sensory induced single unit action potentials from axons regenerated through the microchannels of the present invention. The present invention integrated with eight microwires, shown in Fig. 27, approximately 60μιη in diameter serving as microelectrodes was implanted in a rat sciatic nerve animal model for 12 weeks. The microwires were insulated except at the very tip of the wire which resided in approximately the middle of the length of the microchannel. The microwires extended from the device implanted in the sciatic nerve to lie just under the skin of the rat for the duration of the implantation. Given the diameter of the microwires, the 150μιη wide microchannels were used as the net area was closest to the ΙΟΟμιη width shown to be optimal in the previous in vivo study.

After 12 week the rat was anesthetized using anesthetic cocktail (Acepromazine, 0.5 mg/ml, ketamine HCl, 65 mg/ml and Xylazine, 7.5 mg/ml). The skin of the rat was then opened exposing the ends of the microwires leading away from the device. The wires were connected to a Cerebus 128 channel data acquisition system (Cyberkinetics Inc.) for signal acquisition and amplified using a sampling rate of 30 kHz for neural spike data. Once the wires were connected, resting signals were recorded for the purpose of proving that the present invention capably records spontaneous single unit action potentials from regenerated axons. Secondly, sensory evoked action potentials were evoked using tactile stimulation of the rat foot pad to further prove the ability of the present invention to record single unit action potentials in response to a normal environmental stimulus. The raw signals were filtered at 300 and 8000 Hz. Thresholding and K- means semi-automatic offline sorting were used to identify individual action potential waveforms.

An example of spontaneous action potential recordings is shown in Fig. 28. The left column of images shows raw data collected using the microwires from two representative channels (1 and 7) showing spontaneous action potentials clearly visible over the background noise with a SNR of approximately 2: 1. The middle column shows the shape of individual single unit spontaneous action potentials isolated from the two representative channels. The right column shows the inter- spike-interval of these spontaneous action potentials showing that top action potential form channel 1 was highly periodic along with the action potential from channel 7. This is in contrast to the other two action potentials form channel 1 which had highly sporadic firing rates. Fig. 29 representatively shows how the firing rate of axons in channel 7 increased by approximately three fold in response to tactile stimulation of the 4th toe on the rat paw.

Together the data collected from this experiment clearly shows the ability of the present invention to be used to record action potentials from regenerated axons to the benefit of a neural interface. Not only were sensory evoked single unit action potentials recorded but a significant amount of spontaneous activity was also recorded. Finally, the action potentials are clearly visible over the background noise in the raw signal without needing any processing besides online filtering. As a first step towards a more capable device incorporating upwards of a hundred electrodes, this experiment has validated the approach and concept behind the present invention.

Methods For Fabricating Closed Microchannels

The present invention further comprises methods for fabricating closed microchannels. A preferred fabrication protocol comprises SU-8 microchannel walls fabrication after electrode insulation, as follows:

1. Treat the PDMS substrate with 0₂ plasma for 60 seconds to activate the surface and make hydrophilic. This allows the photoresist in the next step to adhere well to the PDMS surface.

2. Immediately spin on SU-8 photoresist

a. 3000rpm; lOOOrpm/sec; 30 seconds

3. Pre-exposure bake on a hotplate

a. 60°C for 5 minutes

b. 95°C for one hour

c. Take off of hotplate and let cool to room temp. 4. Expose SU-8 using mask aligner

a. Measure power

b. Exposure energy=520mJ/cm2 (experimentally determined)

c. Exposure time = Energy/Power

d. Expose for calculated exposure time

5. Post-exposure bake on a hotplate

a. 65 °C for 5 minutes

b. 95°C for 15 minutes

c. Take off of hotplate and let cool to room temperature

6. Develop the sample in SU-8 Developer (NO agitation, carefully and slowly place in baths)

a. Develop for approximately 25 minutes and then carefully and slowly remove b. Carefully and slowly dunk in fresh developer for additional few minutes to help remove remaining residue

A preferred fabrication protocol comprises the addition of top PDMS layer to the microchannels, as follows:

1. Spin PAA (Polyacrylic Acid, water soluble polymer) on a clean glass slide

a. 1500rpm; 500rpm/sec; 30 seconds

2. Bake PAA covered glass at 60°C for five minutes

3. Repeat steps 1 and 2 immediately (2 PAA layers)

4. Immediately spin PDMS (10 microns) onto the PAA covered glass slide

a. 2000rpm; 500rpm/sec; 400sec

5. Partially bake the sample @ 65 °C for four minutes.

6. Remove edge beads from sample and from the sample with the bottom PDMS sample 7. Treat the bottom PDMS substrate with 0₂ plasma for 60 seconds to activate the surface and make hydrophilic. This allows the partially baked top PDMS layer to adhere well to the top of the SU-8 microchannel walls.

8. Placed the top PDMS layer glass slide on top of the bottom sample so the PDMS is contacting the top of the SU-8 microchannel walls

9. Place a 250 gram weight on top of the glass slide sandwiched structure and bake on a hot plate

a. 60°C for 30 minutes

b. 90°C for one hour

10. Place glass slide sandwiched structure in water under vacuum to draw out all air

a. Allow water to dissolve the PAA layers until the top glass slide can be easily removed

Another preferred fabrication protocol comprises the addition of top PDMS layer to the microchannels, as follows:

1. Spin PAA (Polyacrylic Acid, water soluble polymer) on a clean glass slide

a. 1500rpm; 500rpm/sec; 30 seconds

2. Bake PAA covered glass at 60°C for 5 minutes

3. Repeat steps 1 and 2 immediately (2 PAA layers)

4. Immediately spin PDMS (5 microns) onto the PAA covered glass slide

a. 5000rpm; lOOOrpm/sec; 5 minutes

5. Bake the sample @ 95°C for 10 minutes.

6. Allow sample to cool to room temperature

7. Spin a second layer of PDMS (5 microns) onto the sample

a. 5000rpm; lOOOrpm/sec; 5 minutes

8. Remove edge beads from sample and from the sample with the bottom PDMS sample 9. Placed the top PDMS layer glass slide on top of the bottom sample so the PDMS is contacting the top of the SU-8 microchannel walls

10. Place a 250 gram weight on top of the glass slide sandwiched structure and bake on a hot plate

a. 90°C for one hour

11. Place glass slide sandwiched structure in water under vacuum to draw out all air

Methods For Patterning Electrodes On PDMS Using Photolithography And Lift-Off Processes

The present invention further comprises methods for patterning electrodes on PDMS using photolithography and lift-off processes. A preferred fabrication protocol comprises electrode patterning using 30μιη NR4-8000 photolithography and lift-off, as follows:

1. Spin PDMS base layer (40μιη) onto gold coated glass slide.

a. lOOOrpm; 500rpm/sec; 120sec

b. Immediately bake @ 95C for 25 minutes

c. Let cool to room temperature

2. Treat PDMS substrate with 0₂ plasma for 60 seconds to activate the surface and make hydrophilic. This allows the photoresist in the next step to adhere well to the PDMS surface.

3. Immediately spin on the photoresist NR4-8000P

a. 800rpm; 200rpm/sec; 30sec

4. Pre-exposure bake in an oven

a. 120°C for 10 minutes

b. Take out of oven and let cool to room temperature

5. Expose NR4-8000P using mask aligner a. Measure power under transparency, glass plate, and filter

b. Exposure energy=630mJ

c. Exposure time = Energy/Power

d. Expose for slightly less than the calculated exposure time in order to create an undercut.

i. This undercut can be quite important and will really only be present in the low feature density areas. High feature density areas will have straighter side walls instead of an undercut. If the sample is exposure for the calculated time, the high density areas will take longer to develop and result in delamination of the photoresist from the PDMS substrate.

Post-exposure bake in an oven

a. 90°C for 30 minutes

b. Turn oven off, open crack door open and allow to slowly cool to room temperature

i. If the sample cools too quickly cracking in the photoresist will occur.

Develop the sample in Resist Developer 6 (NO agitation, carefully and slowly place in baths)

a. Develop for approximately four minutes and then carefully and slowly remove. b. Carefully and slowly dunk in water to wash off developer and residue.

c. Carefully and slowly dunk in fresh developer for an additional 10 seconds to help remove remaining residue

d. Carefully and slowly dunk in water again to wash off developer and residue. e. Let completely dry at room temperature

As a last measure to make sure there is no NR4-8000P remaining in the developed regions of the sample, etch the sample using 0₂ plasma for two minutes.

Create gold electrodes using an E-Beam Evaporator a. Deposit 100A titanium (Ti) @ 1 angstrom/sec that serves as an adhesion layer between gold and the PDMS substrate.

b. Deposit 750A gold (Au) @ langstrom/sec

10. Perform NR4-8000P lift-off using Resist Remover 41 (RR41)

a. Place sample in RR41 for approximately 30 minutes

b. Allow all photoresist coated with Ti/Au to (completely) lift-off

c. Wash in water and allow to dry at room temperature

Another preferred fabrication protocol comprises insulation of electrode traces except at electrode active sites using NR4-8000p sacrificial posts, as follows:

1. Treat electrode/PDMS substrate with 0₂ plasma for 60 seconds to activate the surface and make hydrophilic. This allows the photoresist in the next step to adhere well to the PDMS surface.

2. Immediately spin on the photoresist NR4-8000P

a. 800rpm; 200rpm/sec; 30sec

3. Pre-exposure bake in an oven

a. 120°C for 10 minutes

b. Take out of oven and let cool to room temperature

4. Expose NR4-8000P using mask aligner

a. Measure power under transparency, glass plate, and filter

b. Exposure energy=630mJ

c. Exposure time = Energy/Power

Post-exposure bake in an oven

a. 90°C for 30 minutes

i. If the sample cools too quickly cracking in the photoresist will occur.

f. This will leave posts covering the electrode active sites for protection in the next step

Treat the sample with 0₂ plasma for 60 seconds to activate the surface and make hydrophilic.

Immediately spin the PDMS insulation layer (ΙΟμιη) onto sample.

a. 5000rpm; lOOOrpm/sec; 120sec

b. Place sample on hot plate for one hour @ room temp

c. Place sample on hot plate for one hour @ 60 °C

d. Put sample in oven for one hour (minimum) @ 75 °C

Etch the sample to remove the possible thin layer of PDMS covering the NR4-8000P posts with 25% 0₂ and 75 % CF4 for 10 minutes. 10. Perform NR4-8000P lift-off using Resist Remover 41 (RR41) to remove the sacrificial posts.

a. Place sample in RR41 for approximately 30 minutes

b. Allow (all) photoresist coated with Ti/Au to completely lift-off

c. Wash in water and allow to dry at room temperature

11. As a last measure to make sure there is no NR4-8000P remaining in the electrode actives sites of the sample, etch the sample using 0₂ plasma for two minutes.

Integration of Electronics for Multiplexing and Amplification

The present policy further comprises integration of electronics for multiplexing and amplification. A preferred integration protocol comprises:

1. Purchase a printed circuit board (PCB) with all off-chip components and Intan Tech chip assembled from a third party PCB vendor

2. Place conductive epoxy on bonding pad regions of the PCB

3. Take microchannel scaffold electrode array previously fabricated and turn over so you are looking through the device and the electrode bonding pads are facing down

4. Line electrode bonding pads on the microchannel scaffold electrode array with the bonding pads on the PCB and gently place together

5. Allow to cure in an oven for at least three hrs @ 120°C.

6. Roll the microchannel region of the microchannel scaffold electrode array to form the final device configuration.

The present invention further comprises another exemplary integration of electronics for multiplexing and amplification protocol comprising:

1. Place conductive epoxy on bonding pad regions of the off chip components and lie upside down.

2. Take microchannel scaffold electrode array previously fabricated and turn over so you are looking through the device and the electrode bonding pads are facing down. 3. Line electrode bonding pads on the microchannel scaffold electrode array with the bonding pads on the off-chip components and gently place together

4. Allow to cure in an oven for at least three hrs @ 120°C.

5. Repeat steps 1-4 with the Intan tech chip

*Note that this method allow one to incorporate electronics directly onto a PDMS substrate without an intermediary PCB. However, the PCB approach allows one to save overall space by reducing the footprint of the device.

A prototype regenerative microchannel scaffold using PDMS as the base and cover layers and SU-8 as the microchannel walls has been successfully fabricated. Additionally, a first step in developing an integration method for on-board electronics to aid in signal extraction has been made and further validated the present invention's capability of being rolled to form a three- dimensional scaffold.

The substrate and structural materials of the scaffold have been shown to be non-toxic by supporting the growth of multiple cells types, DRG neurites and non-neuronal cells. Furthermore, the capability of the microchannel design to guide and direct DRG neurite outgrowth and non-neuronal cell migration along and through the microchannels has been verified. These results show that the microchannel design provides a method to guide regenerating axons, and can be used as a novel platform to incorporate electronics for chronic recording and stimulation from small specific groups of axons.

The neural interface has an integrated microelectrode in each microchannel to form a high-throughput electrode array. Such a microchannel-scaffold electrode array has potential to significantly enhance the efficacy and reliability of peripheral nerve interfacing. This is based upon the rational that confining an axon to a microchannel limits the volume of the low impedance extracellular fluid and matrix surrounding the axon.

Limiting the extracellular volume effectively increases the extracellular resistance and following from Ohm's Law, increases the extracellular potential. Using an approximately 3mm microchannel will also ensure that a node of Ranvier is somewhere in the channel allowing the incorporated electrodes to be spatially independent of the nodes, which is a major advantage. Furthermore, encouraging the axons to regenerate in small numbers through numerous microchannels allows for highly selective recording and stimulation based upon electrical and spatial isolation.

With the success of this invention, chronic recording of AP's from single to small sets of axons from an innovative, regenerative, bi-directional neural interface will be established. The present invention itself will house upwards of 100 microchannels and each of these channels can hold an incorporated electrode in future generations of the device. Importantly, the channels of the device in future generations can be used for different purposes where many can be used towards the control of a neural prosthetic and many can be used towards providing sensory feedback. With human implantation, each microchannel could be stimulated and the individual asked if they feel anything. This would allow the sensory mapping of the microchannels. This type of analysis is already conducted in deep brain stimulation surgeries where surgeons ask patients what they are feeling during the surgery. Then the microchannels could record patient induced motor axon AP's allowing the motor mapping of the microchannels. This technology would close the prosthetic control loop allowing amputees to "feel" with their prosthetic limb. Furthermore, the On board' electronics can provide a simple direct connection between the present invention and any prosthetic device a user would desire. The invention can also be used as a research tool significantly advancing knowledge of nerves because this device will provide an ability to "map" nerves based on active sets of axons during movements and sensations. In conclusion, the success of this invention paves the way for a high degree of control, proprioception, tactile feedback, and other sensation in future neural prostheses, positively affecting amputees and individuals with disabilities.

While the invention has been disclosed in its exemplary forms, it will be apparent to those skilled in the art that many modifications, additions, and deletions can be made therein without departing from the spirit and scope of the invention and its equivalents, as set forth in the following claims.

Claims

CLAIMS We Claim:

1. A regenerative neural interface to establish high-channel, bi-directional communication between an amputated nerve and a device comprising:

a thin-film polydimethylsiloxane (PDMS)/SU-8 based regenerative microchannel scaffold having microchannels;

microelectrodes incorporated in the microchannels; and

an integrated multiplexer/amplifier for large channel recordings.

2. The regenerative neural interface of Claim 1, wherein when in use, the amputated nerve comprises groups of axons and the interface provides intimate contact between the microelectrodes and groups of axons.

3. The regenerative neural interface of Claim 1, wherein when in use, the microchannels confine axon growth to limit the volume of low impedance extracellular fluid and matrix surrounding the axon.

4. The regenerative neural interface of Claim 1 further comprising a shield surrounding at least a portion of the interface.

5. The regenerative neural interface of Claim 4, wherein when in use, the shield reduces noise and disruptive electromyogram (EMG) signals from axons.

6. The regenerative neural interface of Claim 1, wherein the scaffold comprises:

a bottom layer comprising PDMS; and

walls comprising SU-8 polymer extending upward from the bottom layer;

wherein the bottom layer and the walls form the microchannels.

7. The regenerative neural interface of Claim 6, the bottom layer having a thicknesses in the range of approximately ΙΟμιη and 40μιη.

8. The regenerative neural interface of Claim 6, the walls having a height of approximately ΙΟΟμπι.

9. The regenerative neural interface of Claim 6, the walls having a width of approximately 20μιη.

10. A regenerative neural interface to establish high-channel, bi-directional communication between a protoplasmic protrusion and a device comprising:

a bottom layer comprising PDMS; and

a plurality of walls comprising SU-8 polymer extending upward from the bottom layer; the walls running approximately parallel to one another, forming microchannels;

the bottom layer having a thicknesses in the range of approximately ΙΟμιη to 40μιη;

the walls having a width of approximately 20μιη and a height of approximately ΙΟΟμιη; and

the widths of at least a portion of the microchannels are in the range of approximately

11. The regenerative neural interface of Claim 10 further comprising at least one microelectrode incorporated in at least a portion of the microchannels.

12. The regenerative neural interface of Claim 10 further comprising an integrated multiplexer/amplifier.

13. The regenerative neural interface of Claim 10 further comprising a shield surrounding at least a portion of the interface.

14. The regenerative neural interface of Claim 10, the microchannels having a length of approximately 3mm.

15. A rolled regenerative neural interface to establish high-channel, bi-directional communication between a protoplasmic protrusion and a device comprising:

a bottom layer comprising PDMS; and

a plurality of walls comprising SU-8 polymer extending upward from the bottom layer; the layer and walls forming a layer of microchannels, that when rolled, form a scaffold with microchannels closed along their length.

16. The rolled regenerative neural interface of Claim 15, the bottom layer having a thicknesses in the range of approximately ΙΟμιη to 40μιη, the walls having a width of approximately 20 μιη and a height of approximately ΙΟΟμιη, and the widths of at least a portion of the microchannels are in the range of approximately 50μιη to 150μιη.

17. The rolled regenerative neural interface of Claim 15, having a total radius of 1.5mm.

18. The rolled regenerative neural interface of Claim 15 further comprising at least one microelectrode incorporated in at least a portion of the microchannels.

19. The rolled regenerative neural interface of Claim 15 further comprising an integrated multiplexer/amplifier.

20. The rolled regenerative neural interface of Claim 15 further comprising a shield surrounding at least a portion of the interface.

21. The regenerative neural interface of Claim 15, the microchannels having a length of approximately 3mm.

22. A method of fabricating the regenerative neural interface of Claim 1.

23. A method of fabricating the regenerative neural interface of Claim 10.

24. A method of fabricating the rolled regenerative neural interface of Claim 15.

25. A method of fabricating a regenerative neural interface comprising:

providing a PDMS bottom layer and walls forming microchannels; and

patterning gold electrodes into the PDMS bottom layer with exposed portions in the microchannels forming a conformable microelectrode array (cMEA).

26. The method of fabricating according to Claim 25 further comprising connecting a channel amplifier/multiplexer chip to the interface.

27. The method of fabricating according to Claim 25 further comprising fabricating the PDMS-based cMEA first with bonding pads available for mounting the cMEA.

28. A method of fabricating a regenerative neural interface comprising:

providing a bottom layer comprising PDMS;

providing a top layer comprising PDMS; and providing walls comprising SU-8 polymer between the bottom and top layers, forming microchannels.

29. The method of fabricating according to Claim 28 further comprising patterning gold electrodes into the PDMS bottom layer with exposed portions in the microchannels forming a cMEA.

30. A neural interface comprising:

a base layer comprising a polymeric organosilicon compound; and

at least two channel walls comprising phenol formaldehyde resin, the walls and base layer forming a channel;

wherein the channel accommodates regenerating a protoplasmic protrusion.

31. The neural interface of Claim 30, the channel comprising an electrical conductor for communication with a protoplasmic protrusion.

32. The neural interface of Claim 30, the polymeric organosilicon compound comprising polydimethylsiloxane (PDMS).

33. The neural interface of Claim 30, the phenol formaldehyde resin comprising an epoxy- based negative photoresist.

34. The neural interface of Claim 30, the phenol formaldehyde resin comprising SU-8 polymer.

35. A neural interface comprising:

a base layer comprising polydimethylsiloxane (PDMS); and

a plurality of channel walls comprising SU-8 polymer, the walls and base layer forming microchannels;

wherein the microchannels accommodate regenerating a protoplasmic protrusion.

36. The neural interface of Claim 35, a channel comprising an electrical conductor for communication with a protoplasmic protrusion.

37. The neural interface of Claim 35, wherein the PDMS provides a flexible and stretchable scaffold, and wherein the SU-8 polymer provides mechanical integrity to the microchannels.

38. A neural interface comprising:

a base layer comprising polydimethylsiloxane (PDMS);

a plurality of channel walls comprising SU-8 polymer; and

a top layer comprising PDMS;

wherein the walls and layers form closed microchannels; and

wherein the microchannels accommodate regenerating nerves.

39. The neural interface of Claim 38, each channel comprising an electrode for communication with a regenerating nerve.

40. The neural interface of Claim 38, each channel comprising an electrode atop at least a portion of the base layer for communication with a regenerating nerve.

41. A tubular neural interface comprising:

a base layer comprising polydimethylsiloxane (PDMS);

a plurality of channel walls comprising SU-8 polymer;

a top layer comprising PDMS; and

a plurality of electrodes;

wherein the walls and layers form closed microchannels having at least one electrode; and

wherein the microchannels accommodate regenerating nerves.

42. The neural interface of Claim 41, at least one of the at least one electrode of the closed microchannels located atop at least a portion of the base layer for communication with a regenerating nerve.