Elsevier

Biomaterials

Volume 35, Issue 34, November 2014, Pages 9255-9268
Biomaterials

Chronic tissue response to carboxymethyl cellulose based dissolvable insertion needle for ultra-small neural probes

https://doi.org/10.1016/j.biomaterials.2014.07.039Get rights and content

Abstract

Implantable neural electrodes must drastically improve chronic recording stability before they can be translated into long-term human clinical prosthetics. Previous studies suggest that sub-cellular sized and mechanically compliant probes may result in improved tissue integration and recording longevity. However, currently these design features are restricted by the opposing mechanical requirements needed for minimally damaging insertions. We designed a non-cytotoxic, carboxymethylcellulose (CMC) based dissolvable delivery vehicle (shuttle) to provide the mechanical support for insertion of ultra-small, ultra-compliant microfabricated neural probes. Stiff CMC-based shuttles rapidly soften immediately after being placed ∼1 mm above an open craniotomy as they absorb vapors from the brain. To address this, we developed a sophisticated targeting, high speed insertion (∼80 mm/s), and release system to implant these shuttles. After implantation, the goal is for the shuttle to dissolve away leaving only the electrodes behind. Here we show the histology of chronically implanted shuttles of large (300 μm × 125 μm) and small (100 μm × 125 μm) size at discrete time points over 12 weeks. Early time points show the CMC shuttle expanded after insertion as it absorbed moisture from the brain and slowly dissolved. At later time points neuronal cell bodies populate regions within the original shuttle tract. The large CMC shuttles show that the CMC expansion can cause extended secondary damage. On the other hand, the smaller CMC shuttles show limited secondary damage, wound closure by 4 weeks, absence of activated microglia at 12 weeks, as well as evidence suggesting neural regeneration at the implant site. This shuttle, therefore, shows great promise facilitating the implantation of nontraditional ultra-small, and ultra-compliant probes.

Introduction

Penetrating neural probes have demonstrated feasibility and great potential in human brain computer interface applications [1], [2], [3]. However, current devices face the common issues of poor reliability and longevity, which have greatly limited the widespread clinical adoption of the technology [4], [5], [6], [7]. On a basic science level, limitations in single-unit neural recording reliability and stability have constrained longitudinal studies of memory and plasticity, particularly in awake, behaving animals [8].

The mechanisms of chronic recording failure are an active area of research with multiple identified or proposed biological and non-biological contributing factors. Non-biological factors mainly involve material failure including electrode corrosion, insulation leakage, and breaking of the lead and connectors [7]. On the other hand, the biological host tissue reactions could adversely affect the recording functionality. The host reaction begins with a penetration injury which damages the vasculature [6], [9], [10], tears and compresses the tissue and kills the neurons and glial cells along the penetration path. Activation of microglia cells occurs immediately after insertion as demonstrated by 2-photon in vivo imaging [11]. This acute injury may recover spontaneously to a significant degree if the implant is removed right away, so called “stab wound” injury [12]. The persistent presence of the device on the other hand triggers chronic inflammatory tissue reactions with a chronic glial scar formation and loss of neurons at the vicinity of the probe. Such consequences are likely to lead to degradation of neural recording [12], [13], [14], [15], [16].

Experimental and modeling evidences suggest that implant compliance (flexibility) and material softness significantly affect the degree of host tissue reaction to the implanted electrodes. Physiological motion (breathing and heartbeat), micromotion (normal movement), macromotion (falls and traumatic brain injuries) can exacerbate injuries surrounding the implants due to interfacial strain caused by the mechanical mismatch between probe and brain tissue. In addition, the friction induced stress resulting from the mismatch of mechanical properties leads to mechanosensitized inflammatory response in the surrounding tissues [17], [18], [19], [20], [21] as well as disruption of the blood-brain barrier (BBB) which could exacerbate tissue inflammation and neuronal degeneration [6], [9], [10], [11]. Computer models have indicated that soft and flexible materials help minimize mechanical strain at the probe tissue interface [22], [23].

Another major consideration is the size of the implant. Seymour et al. have shown that a neural probe with a subcellular size elicited minimal glial reaction with no neuronal loss around the implant [16]. In another study a 7 μm thick carbon fiber electrode produced much reduced tissue reaction than silicon probes (100 μm × 15 μm cross-section) and recorded stable neural signal chronically [14]. Besides the obvious benefit of reduced invasiveness, one proposed mechanism for size being an important factor is that smaller surface area leads to reduced inflammatory cell and molecule accumulation [14], [24]. It is also possible that reduction in size leads to increased device flexibility, which in turn reduces the mechanical irritation.

Therefore, many research groups have devoted efforts to fabricating probes with more flexibility and/or subcellular size [7], [8], [14], [22], [25], [26], [27], [28]. However increased flexibility and complex geometry of next generation electrodes introduces new challenges in inserting these devices into the brain. Ultra-compliant probes tend to buckle and bend prior to penetration into the brain [25]. Even when these probes penetrate the surface of the brain, they can deflect causing the electrodes to stray far from the intended brain region. Furthermore, subcellular sized device are usually not mechanically strong enough to penetrate the pia [25]. As a result, many strategies to implant flexible devices previously described elsewhere have been developed, each with varying design space limitations and tradeoffs [25]. Many of the emerging insertion strategies employ flexible polymer devices temporarily coupled to removable insertion shuttles [25], [28], [29], [30], [31]. Flexible devices are coupled to stiff shuttles through dissolvable adhesives, sleeve holes, or electrostatic interaction [25], [29], [30], [31]. This approach is generally limited to tethered flexible probes with planar geometry or simple 3D geometries and becomes more challenging with high channel count 3D multielectrode arrays. As these devices become smaller and more flexible (especially in the axial direction), it will be important to consider the surface chemistries of the flexible probe and shuttle to reduce hydrophobic adhesion in the aqueous environment of the brain and cerebral spinal fluid [25], [32], [33], [34].

Another approach is the use of an electromagnetic linear accelerator or coilgun [35]. While current efforts are focused on improving targeting precision, limitations also exist in compatible geometry, size, and the minimum size of ferromagnetic particles in the electrode necessary to generate velocity and overcome viscoelastic effects in tissue. Naturally, this also creates compatibility challenges with MRI/fMRI for the implanted patients. A separate approach uses thermal and water-sensitive polymers that are stiff at room temperature and dry environments, but become softer after insertion [33], [34]. While promising, these materials present challenges in fabrication of functional devices due to poor dimensional stability during microfabrication. With decreasing size and elasticity, cracking of thin-film conductors become a larger concern. It may become necessary to look to more durable composite thin-film conductors [28], [36], [37], [38]. An alternative approach that has been well explored employs materials to stiffen a compliant structure for cortical implantation [39]. The materials investigated include polyethylene glycol (PEG) [40] and poly(lactide-co-glycolide) (PLGA) [41] tyrosine-derived polycarbonate [42], and biodegradable silk polymer [43]. These strategies have relied on dip coating of flexible devices, which requires a certain amount of minimum stiffness and present challenges in tip sharpness and coating uniformity desired for implantation [42], [43], [44]. We have proposed an innovative platform technology to enable the fabrication and delivery of lithographically patterned ultra-small and complex neural probe via a micromolded dissolvable shuttle made by carboxymethyl cellulose (CMC) [25], [28]. CMC is a natural occurring polysaccharide which is soluble in water and leaves no harmful dissolution by-products behind, thereby making it a suitable material choice for our application. The molding process allows precise design of the CMC delivery vehicle (shuttle) geometry and size for optimal outcomes. As a necessary first step, a comprehensive brain tissue reaction analysis to the CMC shuttles is reported here.

Section snippets

Fabrication

The probes are fabricated using a solvent-based spin-casting process Fig. 1a. The spin-casting fabrication method was previously described in Ref. [45] for fabrication of microneedle arrays for transdermal drug delivery. This method enables molding of solvent-based polymers. In general, the mold is placed inside one of the buckets of a centrifuge, and the dissolvable polymer mixed with the solvent (commonly in a gel form) is loaded on the mold. Centrifuging provides both the required force to

Results

Small and large CMC shuttles and control metal microwires of two diameters were implanted through the dura into the motor cortex. 6–7 animals were sacrificed at discrete time points of 1 day, 1 week, 4 weeks, or 12 weeks for histological analysis (Fig. 2, Fig. 3). Control images were taken 500–700 microns away from the implants in each hemisphere.

Discussion

Dissolvable CMC shuttles with base cross-section areas of 12,500 μm2 and 37,500 μm2 were implanted into motor cortex and compared with tapered parylene-C insulated microwires with base cross-sectional area of 12,685 μm2 and 45,239 μm2. It should be noted that the tapering for the CMC shuttle and microwires were different. The small CMC was tapered at the tip over ∼150 μm while the small microwire was tapered over ∼650 μm. For the large CMC, the tip was tapered over ∼450 μm, while the large

Conclusion

CMC insertion shuttles are non-cytotoxic and may be a viable method for long-term implantation of ultra-small, ultra-compliant, complex-structured implants into neural tissue. It may also be a viable way to deliver drugs, scaffolds, or drug eluting matrices into soft tissue. However, it should be noted that the size (volume) of the shuttle plays a large role in the ability and duration the tissue takes to completely heal the wound. Other strategies may be applied to increase the dissolution

Acknowledgments

The authors would like to thank Gregory J. Brunette for assistance with cell counting and Center for Biological Imaging for confocal microscopy. This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) MTO under the auspices of Dr. Jack Judy through the Space and Naval Warfare Systems Center, Pacific Award No. N66001-11-1-4025. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not

References (61)

  • J.N. Israelachvili et al.

    Measurement of the hydrophobic interaction between two hydrophobic surfaces in aqueous electrolyte solutions

    J Colloid Interface Sci

    (1984)
  • D.B. Jaroch et al.

    Magnetic insertion system for flexible electrode implantation

    J Neurosci Methods

    (2009)
  • S.L. Chorover et al.

    A sweet new multiple electrode for chronic single unit recording in moving animals

    Physiol Behav

    (1972)
  • D. Lewitus et al.

    Ultrafast resorbing polymers for use as carriers for cortical neural probes

    Acta Biomater

    (2011)
  • E. Azemi et al.

    The surface immobilization of the neural adhesion molecule L1 on neural probes and its effect on neuronal density and gliosis at the probe/tissue interface

    Biomaterials

    (2011)
  • J.J. Bernstein et al.

    Sequestering of immunoglobulins by astrocytes after cortical lesion and homografting of fetal cortex

    Int J Dev Neurosci

    (1993)
  • J.J. Bernstein et al.

    Injury-related spinal cord astrocytes are immunoglobulin-positive (IgM and/or IgG) at different time periods in the regenerative process

    Brain Res

    (1987)
  • T. Masuda et al.

    Increase in neurogenesis and neuroblast migration after a small intracerebral hemorrhage in rats

    Neurosci Lett

    (2007)
  • J.L. Collinger et al.

    High-performance neuroprosthetic control by an individual with tetraplegia

    Lancet

    (2012)
  • L.R. Hochberg et al.

    Reach and grasp by people with tetraplegia using a neurally controlled robotic arm

    Nature

    (2012)
  • L.R. Hochberg et al.

    Neuronal ensemble control of prosthetic devices by a human with tetraplegia

    Nature

    (2006)
  • T.D.Y. Kozai et al.

    Reduction of neurovascular damage resulting from microelectrode insertion into the cerebral cortex using in vivo two-photon mapping

    J Neural Eng

    (2010)
  • T. Kozai et al.

    Nanostructured coatings for improved charge delivery to neurons

  • G. Guitchounts et al.

    A carbon-fiber electrode array for long-term neural recording

    J Neural Eng

    (2013)
  • C.S. Bjornsson et al.

    Effects of insertion conditions on tissue strain and vascular damage during neuroprosthetic device insertion

    J Neural Eng

    (2006)
  • T.D.Y. Kozai et al.

    In vivo two photon microscopy reveals immediate microglial reaction to implantation of microelectrode through extension of processes

    J Neural Eng

    (2012)
  • T.D. Kozai et al.

    Ultrasmall implantable composite microelectrodes with bioactive surfaces for chronic neural interfaces

    Nat Mater

    (2012)
  • D.G. Harrison et al.

    Endothelial mechanotransduction, nitric oxide and vascular inflammation

    J Intern Med

    (2006)
  • T. Lund et al.

    Shear stress regulates inflammatory and thrombogenic gene transcripts in cultured human endothelial progenitor cells

    Thromb Haemost

    (2010)
  • S. Uhlig

    Ventilation-induced lung injury and mechanotransduction: stretching it too far?

    Am J Physiol Lung Cell Mol Physiol

    (2002)
  • Cited by (147)

    View all citing articles on Scopus
    View full text