Scholarly article on topic 'Size-dependent long-term tissue response to biostable nanowires in the brain'

Size-dependent long-term tissue response to biostable nanowires in the brain Academic research paper on "Clinical medicine"

CC BY-NC-SA
0
0
Share paper
Academic journal
Biomaterials
OECD Field of science
Keywords
{Biocompatibility / Nanoparticle / Inflammation / Brain}

Abstract of research paper on Clinical medicine, author of scientific article — Lina Gällentoft, Lina M.E. Pettersson, Nils Danielsen, Jens Schouenborg, Christelle N. Prinz, et al.

Abstract Nanostructured neural interfaces, comprising nanotubes or nanowires, have the potential to overcome the present hurdles of achieving stable communication with neuronal networks for long periods of time. This would have a strong impact on brain research. However, little information is available on the brain response to implanted high-aspect-ratio nanoparticles, which share morphological similarities with asbestos fibres. Here, we investigated the glial response and neuronal loss in the rat brain after implantation of biostable and structurally controlled nanowires of different lengths for a period up to one year post-surgery. Our results show that, as for lung and abdominal tissue, the brain is subject to a sustained, local inflammation when biostable and high-aspect-ratio nanoparticles of 5 μm or longer are present in the brain tissue. In addition, a significant loss of neurons was observed adjacent to the 10 μm nanowires after one year. Notably, the inflammatory response was restricted to a narrow zone around the nanowires and did not escalate between 12 weeks and one year. Furthermore, 2 μm nanowires did not cause significant inflammatory response nor significant loss of neurons nearby. The present results provide key information for the design of future neural implants based on nanomaterials.

Academic research paper on topic "Size-dependent long-term tissue response to biostable nanowires in the brain"

Biomaterials

Contents lists available at ScienceDirect

Biomaterials

journal homepage: www.elsevier.com/locate/biomaterials

Size-dependent long-term tissue response to biostable nanowires in the brain

CrossMark

Lina Gällentoft a, Lina M.E. Pettersson a, Nils Danielsen a, Jens Schouenborg a, Christelle N. Prinz a,b' **, Cecilia Eriksson Linsmeiera' *

a Neuronano Research Center, Department of Experimental Medical Science, Medical Faculty, Lund University, BMCF10, Lund SE-221 84, Sweden b Division of Solid State Physics, The Nanometer Structure Consortium, Lund University, Box 118, Lund SE-221 00, Sweden

ARTICLE INFO

Article history: Received 1 September 2014 Accepted 25 November 2014 Available online 16 December 2014

Keywords:

Biocompatibility

Nanoparticle

Inflammation

ABSTRACT

Nanostructured neural interfaces, comprising nanotubes or nanowires, have the potential to overcome the present hurdles of achieving stable communication with neuronal networks for long periods of time. This would have a strong impact on brain research. However, little information is available on the brain response to implanted high-aspect-ratio nanoparticles, which share morphological similarities with asbestos fibres. Here, we investigated the glial response and neuronal loss in the rat brain after implantation of biostable and structurally controlled nanowires of different lengths for a period up to one year post-surgery. Our results show that, as for lung and abdominal tissue, the brain is subject to a sustained, local inflammation when biostable and high-aspect-ratio nanoparticles of 5 mm or longer are present in the brain tissue. In addition, a significant loss of neurons was observed adjacent to the 10 mm nanowires after one year. Notably, the inflammatory response was restricted to a narrow zone around the nanowires and did not escalate between 12 weeks and one year. Furthermore, 2 mm nanowires did not cause significant inflammatory response nor significant loss of neurons nearby. The present results provide key information for the design of future neural implants based on nanomaterials.

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-SA

license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

1. Introduction

Achieving long-term monitoring and interaction with neuronal circuits of the central nervous system (CNS) in conscious individuals will have great impact in neuroscience, both for research and clinical applications [1—4]. However, current neural interfaces usually show a decline in performance over time and typically do not provide stable communication with individual neurons. It is commonly assumed that these shortcomings are due to an encapsulating inflammatory tissue response to the implant and consequent displacement/loss of neurons nearby [5—8]. Notably, this occurs despite using nontoxic materials. There is accumulating evidence indicating that microforces between the brain, which constantly exhibits movements, and neural interfaces, which

* Corresponding author. BMCF10/F12, Lund SE-221 84, Sweden. Tel.: +46 (0)46 2224107.

** Corresponding author. Division of Solid State Physics, Lund University, Box 118, Lund SE-221 00, Sweden. Tel.: +46 (0)46 2224796.

E-mail addresses: Christelle.Prinz@ftf.lth.se (C.N. Prinz), Cecilia.Eriksson_ Linsmeier@med.lu.se (C.E. Linsmeier).

typically exhibit a poor mechanical compliance with the tissue, play an important part in fuelling the inflammatory tissue response and contribute to the instability of neural recordings [9—12].

A promising approach to locally improve the mechanical compliance of implants is to use flexible nanomaterials, such as nanorods or carbon nanotubes (CNTs), on the surface of the neural interfaces. Surfaces coated with CNTs have been reported to improve the electrical properties and lower the evoked inflammatory tissue response towards neural interfaces, both in vivo and in vitro [13,14]. Nanowire-modified substrates have been shown to promote neuronal growth and limit the proliferation of glial cells in vitro [15—17]. Furthermore, successful in vivo neural recordings have recently been achieved using nanowire-based electrodes [18].

In order to achieve long-term communication with the brain, these nanostructures need to be biostable. However, concerns have been raised about using biostable, long, and high-aspect-ratio nanoparticles (such as nanowires and CNTs) in the brain. In rodent abdominal tissue and lungs, it has been shown that biostable nanotubes, nanowires and nanorods of lengths comparable to the size of immune cells, can induce frustrated phagocytosis, resulting in a chronic inflammation which escalates over time, comparable to

http://dx.doi.org/10.1016/j.biomaterials.2014.11.051

0142-9612/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

asbestosis [19—21 ]. In the brain, the presence of biostable and high-aspect-ratio nanoparticles detached from implanted neural interfaces could pose a similar risk and give rise to an escalating inflammatory response as seen in asbestosis. At present, little is known concerning the brain tissue response after long-term exposure to biostable nanoparticles having morphological features similar to asbestos fibres.

To clarify the long-term risks involved in using nanostructured neural interfaces for stable communication with brain cells, we here examine the chronic brain tissue response to implantation of biostable nanorods of equal diameter but different lengths. Epitaxially grown nanowires were used as a model particle for high-aspect-ratio nanorods since their dimensions can be controlled precisely [22]. The glial response and neuronal survival after implantation of biostable nanowires of three different lengths in the rat brain were evaluated after up to one year implantation time, which corresponds to half of the animal's lifespan.

The results show that nanowire length has a significant influence on the inflammatory tissue response as well as on neuronal survival. Whereas no significant increase in glial response or loss of neurons was found for 2 mm long nanowires, the longer nanowires studied (5 and 10 mm) caused a persistent, but not escalating, glial response. Furthermore, for the 10 mm nanowires, a significant loss of neurons could be seen at one year post implantation.

2. Methods

2.1. Nanowire growth and coating

Metal organic vapour phase epitaxy (Aix 200/4, Aixtron, Germany) was used to grow gallium phosphide (GaP) nanowires from 40 nm gold aerosol particles randomly distributed on a (111)B GaP substrate (Girmet Ltd, Moscow, Russia) at an average density of 1/mm2, as previously described [22,23]. The temperature for nanowire growth was 470 °C and wire growth was initiated by supplying Ga(CH3)3 in addition to PH3. Precursor molar fractions were 4.3 x 10 6 and 8.5 x 10 2 for Ga(CH3)3 and PH3, respectively, in a hydrogen carrier gas flow of 6 L/min. The growth was conducted under low pressure (10 kPa). The growth duration was adjusted to produce nanowires of different lengths (2 ± 0.2 mm, 5 ± 0.2 mm and 10 mm ± 0.5 mm). The nanowire length was characterized by scanning electron microscope (SEM) imaging.

In order to obtain nanowires with a stable and inert surface chemistry, the GaP nanowires were coated with a 20 nm layer of hafnium oxide (HfOx) using atomic layer deposition (Savannah-100 system, Cambridge NanoTech Inc., USA), resulting in a final nanowire diameter of 80 nm ± 5 nm (±10 nm for the 10 mm long nanowires) which was characterized using SEM imaging (Fig. 1). Titanium is widely used as a biomaterial; hafnium belongs to the same group (IVB) as titanium in the periodic table of elements and has been shown to display similar biocompatible traits as titanium [24,25]. In addition, hafnium oxide has been shown to be biocompatible even in a nano configuration [26]. The substrates were cleaned with in an oxygen plasma chamber (Plasma Preen, Plasmatic Systems Inc., USA) and can therefore be expected to carry hydroxyl groups at the surface [27]. The nanowires were subsequently broken off from the substrate using ultra sonication and suspended in Hank's balanced salt solution (HBSS) to a final concentration of 70,000 nanowires/ mL. For further detailed description, see Eriksson Linsmeier et al., 2009 [28].

Table 1

Experimental setup; included and excluded animals listed (n = one unilateral implantation site). Total group number in bold.

Naïve Stab SW-control Control 2 mm 5 mm 10 mm

wound HfOx HfOx HfOx

Groups at 12 weeks 18 8 8 24 26 24 24

12 week included 18 8 8 20 20 18 16

12 week excludeda 0 0 0 4 6 6 8

Groups at one year 20 - - 20 20 20 20

One year included 18 - - 16 11 15 15

One year excludedb 2 - - 4 9 5 5

a At 12 weeks, animals were excluded for meeting predetermined histological exclusion criteria's.

b At one year, animals were excluded either for meeting ethical guideline exclusion criteria's (tumours) or predetermined histological exclusion criteria's.

2.2. Animals

Approval for the animal experiments described below was obtained from the Lund/Malmo local ethical committee on animal experiments. A total of 116 female Sprague-Dawley (SD) rats (Taconic, Denmark) were used in this study. All rats received food and water ad libitum and were kept in a 12-h day-night cycle. The rats weighed approximately 225 g at the beginning of the experiment. The rats followed a normal weight curve after surgery and up to the experimental end point.

For the 12 week time point, seven different experimental groups were used. Three groups received bilateral injections of 2, 5 or 10 mm long HfOx-coated nanowires in HBSS. One group received bilateral control vehicle-injections (HBSS only). One group was kept naïve (no surgical procedures). An additional set of animals received stab wounds (SW) in one hemisphere and control-injections (HBSS only) in the contralateral hemisphere (SW-control). For the one-year time point, five different experimental groups were used. Three groups received bilateral injections of 2, 5 or 10 mm long HfOx-coated nanowires in HBSS, and one group received bilateral control-injections (HBSS only). One group was kept naïve (no surgical procedures).

We found spontaneous tumours in eight out of 60 animals (13.33%) in the one year group (rats approximately 425 days old). Upon tumour detection, these animals were immediately terminated and excluded from the study. A summary of the animal group history is shown in Table . The prevalence of the tumours was distributed evenly over all groups, including the naïve group (no surgical procedures). The occurrence of spontaneous age-related tumours in SD rats has previously been well documented with a reported tumour incidence of 57-58% in female SD rats kept until day 540 or allowed to live out their lifespan [29,30]. Due to the prevalence of tumours in our naïve and control rats and the vast literature demonstrating the incidence of spontaneous tumours in aged rats, the tumours observed in our study were therefore most likely age-related, and not linked to the presence of nanowires in the brain [29-32].

2.3. Surgery

The rats were deeply anaesthetized by intraperitoneal (i.p.) injections of Fen-tanyl (0.3 mg/kg body weight) and Domitor vet (metedetomidin hydrochloride, 0.3 mg/kg body weight). The surgical procedures have been described in detail previously [28]. In short, the anaesthetized animals were prepared for surgery, i.e. their heads were shaved and they were placed in a stereotactic frame (KOPF instruments, USA) set under a stereomicroscope (Leica Microsystems, Germany). The scalp was disinfected using 70% ethanol solution and local anaesthetic, 0.25% Mar-caine (Bupivacaine, 0.33 mg/kg body weight) in sterile water, was administered. To expose the skull a 2 cm midline incision was made. Tissue attached to the skull was

Fig. 1. SEM images of nanowires. Representative SEM images of vertical GaP nanowires coated with HfOx, at low (a) and high (b) magnification. SEM images were used to determine the length and diameter of the nanowires. Stage tilt 20°. Scale bars: 1 mm (a), and 200 nm (b).

carefully removed and blood was cleansed away. Bilateral craniotomies (0 1 mm2) were drilled at 1.0 mm anterior and 2.5 mm lateral to bregma under stereotactic control. The dura mater was incised and deflected using fine forceps. Bilateral stereotactic injections at the above-mentioned coordinates were made using a 2 mL Hamilton syringe with a glass microcapillary (tip 0 ~ 130 mm) attached. The suspension was injected into the striatum at two different depths; i) 5 mm (1 mL) and ii) 4 mm (1 mL); in total 2 mL/hemisphere over a total time of 2 x 2 min. The amount of nanowires injected corresponds to an estimation of the number of nanowires that could detach from a future nanostructured neural interface. Stab wounds were performed in an identical manner. After injections or stab wound, the skin was closed using surgical clips. The surgeries for the different groups took place in different sessions. In order to confirm that the nanowires were still individually suspended in the HBSS and had not assembled into larger aggregates, a drop of the suspension containing nanowires was ejected from the syringe onto a microscope slide and examined using a Nikon eclipse 80i microscope, before and after each injection series.

The animals received subcutaneous injections of Temgesic (buprenorphine, 50 mg/kg body weight) to reduce postoperative pain, as well as an antidote to the anaesthesia (Antisedan, atipamezole hydrochloride, 0.5 mg/kg body weight), and were awakened under supervision.

2.4. Histology

The animals were killed by an i.p. overdose of pentobarbital and transcardially perfused with 200 mL of ice-cold saline solution (sodium chloride 0.9% in distilled water) followed by 125 mL of ice-cold 4% paraformaldehyde (PFA) in 0.1 m phosphate buffer (pH 7.4). The brains were carefully removed and post-fixed in 4% PFA overnight (4 °C). In addition, lymph nodes, liver, heart and vessels from all experimental groups (12 weeks and one year) (n = 24) were dissected and post-fixed in 4% PFA.

The tissues were cryoprotected in 25% sucrose solution until equilibrated and were subsequently attached to the sectioning block using Tissue Tek O.C.T. compound (Sakura Finetek, USA). Coronal serial sections of the brains were cut (6 series) at 10 mm thickness onto Super Frost® plus slides (Menzel-Gläser, Germany) using a cryostat (Microm, Germany). The markers used to visualize activated microglia and macrophages (CD68/ED1), astrocytes (glial fibrillary acidic protein (GFAP)), and neuronal nuclei (NeuN) are summarized in Table 2. All sections were counterstained using the cell nuclei marker 4',6-diamidino-2-phenylindole (DAPI).

The lymph nodes, liver, heart and vessels were sectioned in the same manner as mentioned above, at 10 mm thickness and labelled with ED1 and DAPI. They were subsequently screened for presence of nanowires using a Nikon eclipse 80i microscope and a laser scanning confocal microscope (Zeiss LSM 510). Nanowires scatter confocal laser light and can therefore be visualized using the laser-reflection mode.

Tissue sections were hydrated in phosphate buffered saline (PBS) and blocked with 5% normal goat serum and 0.25% Triton X-100 in PBS (blocking solution). Incubation with primary antibodies (in blocking solution) was made at room temperature (RT) overnight. Sections were subsequently rinsed in PBS followed by incubation with DAPI, goat anti-rabbit IgG Alexa 594 and goat anti-mouse IgG Alexa 488 (in blocking solution) in the dark (RT) for 2 h (Table 2). Sections were rinsed and coverslipped using PVA-DABCO (polyvinyl alcohol, Fluka/Sigma—Aldrich, Switzerland).

Prior to NeuN staining an antigen retrieval method was performed. In short, after hydration, the sections selected for NeuN staining were immersed in a 10 mM sodium citrate buffer (0.05% Tween 20, pH 6) and microwaved for 3 x 5 min at 500 W. Tris-buffer (Sigma—Aldrich, Germany) was used instead of PBS for all steps and Triton X-100 was omitted from the blocking solution, otherwise all steps were made according to the staining protocol above.

Table 2

Summary of primary antibodies, secondary antibodies and nucleic acid stain.

Name Characteristics Host Working Source dilution

ED1 Activated Mouse 1:250 Cat. Nr.

(CD68) microglia/macrophages MCA341R,

AbD Serotec

GFAP Glial fibrillary acidic protein Rabbit 1:5000 Cat. Nr. Z0334,

NeuN Neuronal nuclei, Mouse 1:100 Cat. Nr.

neuronal marker MAB377,

Millipore

Alexa Goat anti-rabbit lgG (H + L) Goat 1:500 Cat. Nr.A11005,

594 Invitrogen

Alexa Goat anti-mouse lgG (H + L) Goat 1:500 Cat. Nr.A11001,

488 Invitrogen

DAPI Nucleic Acid Stain 1:1000 Cat. Nr. D3571,

(4',6-diamidino-2-phenylindole) Invitrogen

2.5. Image acquisition and analysis for quantitative assessment

In order to quantitatively evaluate the density of neurons and activation state of glial cells in defined regions of interest (ROls) in relation to the injection site, we quantified the binding of selected markers to cell specific proteins/antigens in the tissue. ln short, the sections were screened to detect the location of the scar where the ED1- and GFAP-positive area was seen at its maximum. At this location, photographs of the fluorescence of the ED1- and GFAP-positive cells and cell nuclei (DAPI) were taken at each injection site, using a DS-2MV digital camera (Nikon, Japan) mounted on a Nikon eclipse 80i microscope with a 10x objective. Image capture and analysis were performed using the NlS-Elements 3.1 software (Nikon lnstruments, Japan). The adjacent brain sections were stained using NeuN, GFAP and DAPl and photographs of the injection sites were taken.

The quantification analysis was based on a previously described method [28]. In short, rectangular shaped ROls, inner and outer ROls (100 x 800 mm and 300 x 800 mm, respectively), were centred on the injection tract (Supplementary Fig. 1). The different cell quantifications were made by measuring the proportion of immunoreactive area (for ED1 and GFAP) and the number of cells (for NeuN and DAPI) in the total screened area for each marker and injection site for all experimental groups [28]. Due to variability in the binding specificity of ED1 and GFAP to their respective antigen, thresholds were set for each individual image at a fixed ratio of the mean background intensity for each marker. The thresholds were set to ensure that only positively stained antigens were quantified, whereas the nonspecific background staining was not. The threshold for the signal to background ratio was set to 5.5 for ED1 immunofluorescence and to 4.5 for GFAP immunofluores-cence. The fraction of the area above this threshold in each ROl was quantified. Neuronal nuclei were counted manually by matching NeuN-positive cells with a co-stained DAPl-positive nucleus within the ROls. Cell nuclei were counted manually by counting the number of DAPl-positive nuclei (above 0 3 mm to avoid counting debris/ artefacts) within the ROl.

Confocal images of the scar were photographed with a laser scanning confocal microscope (Zeiss LSM 510) using a 63 x oil-immersion objective (N.A. 1.4) and the ZEN software (Zeiss). Nanowires were visualized using the laser-reflection mode.

2.6. Statistical analyses

Kruskal Wallis with Dunn's multiple comparison test was used to compare the experimental groups. Wilcoxon matched-pairs signed rank test was used when paired comparisons of two experimental groups were performed (SW vs. SW-control). p-values < 0.05 (*) were considered significant. All values are presented as median values, together with 25 and 75 percentiles, as well as, minimum and maximum values. All analyses were performed using the GraphPad Prism 6.0 software (GraphPad Software lnc., USA).

3. Results

To determine the effects of the length of biostable high-aspect-ratio nanowires on the evoked inflammatory tissue response, we injected nanowires of three different lengths (2, 5 and 10 mm) into the rat striatum. The animals were killed and perfused either after 12 weeks or one year. In order to assess the degree of reactive astrocytosis and the evoked inflammatory response, brain sections from all groups were stained for ED1- (microglial cells and macrophages) and GFAP-positive cells (astrocytes). To assess the neuronal survival and the number of total cell nuclei in the scar area, NeuN (neuronal nuclei) and DAPI (cell nuclei) were used. The different markers were quantified in two regions of interest (ROls) surrounding the injection tract (see method for detailed experimental protocol).

3.1. General observations

At both time points, areas positive for ED1 and GFAP were significantly larger for control and experimental groups, when compared to the naïve group (inner ROl) (Supplementary Table 1 ). The neuronal and cell nuclei density (inner ROl) in control and experimental groups was significantly lower as compared to the naïve group at both time points (Supplementary Table 1).

Excluding brain tissue, no nanowires were detected in additionally dissected and investigated tissues, i.e. in the lymph nodes, liver, heart and vessels, in any of the experimental groups.

3.2. Effect of the suspension media

In order to investigate the effect of the control, (i.e. vehicle solution, Hank's balanced salt solution (HBSS)), we performed bilateral surgery on eight animals; stab wound (SW) and a stab wound with control injection of HBSS (SW-control) in opposing hemispheres, respectively. No significant difference in ED1-positive area, neuronal density or cell nuclei density (inner or outer ROI) could be found when comparing the responses in the two groups (SW vs. SW-control) after 12 weeks. However, GFAP-positive area was significantly higher (p = 0.023) in the inner ROI for the SW group (mean 17.4% ± 2.3%) as compared to SW-control (mean 10.9% ± 1.3%) (Fig. 2). These results indicate that HBSS may lower the astrocytic inflammatory response in the brain towards the stab wound caused by the injection needle.

3.3. Effect of nanowire length on the brain tissue response after 12 weeks

We injected nanowires of three different lengths (2, 5 and 10 mm) and assessed the brain tissue response and cell density 12 weeks after the injections. Representative fluorescence microscopy images of the scar tissue area after injections of control solution and nanowires of different lengths are shown in Fig. 3.

3.3.1. Microglia/macrophage response

A significant increase in ED1-positive area, in the inner ROI, was found for the group receiving 10 mm long nanowires compared to both the control group (p = 0.0074) and to the group receiving 2 mm long nanowires (p = 0.018), with a mean percentage of fluorescent area of 3.75% ± 0.78% (for 10 mm nanowires), 1.01% ± 0.13% (for control), and 1.11% ± 0.15% (for 2 mm nanowires) (Fig. 4). In the outer ROI, no significant difference in ED1-positive area was found when comparing the groups (data not shown). Moreover, the overall immunoreactive area for ED1 in the outer ROI was close to zero in all groups, indicating that the microglial and macrophage activity is largely limited to the inner ROI 12 weeks after implantation.

3.3.2. Astrocytic reactivity

In the inner ROI, we observed a significant increase in GFAP-positive area in the group receiving 10 mm long nanowires compared to the control group (p = 0.035), with a mean percentage of fluorescent area of 18.3% ± 2.3% and 9.34% ± 1.3%, respectively (Fig. 4). In the outer ROI, there was a significant increase in GFAP-positive area for the group receiving 5 mm long nanowires compared to the control group (p = 0.049), with a mean percentage of fluorescent area of 5.75% ± 1.1% and 2.30% ± 0.38%, respectively (Fig. 4). Furthermore, a trend towards an increase in GFAP-positive area in the outer ROI was

Fig. 2. Analysis of SWand SW-control after 12 weeks. (a—d) Quantification in the inner ROI (0—50 mm) of ED1-positive area (a), GFAP-positive area (b), NeuN density (c), and cell nuclei density (d) at 12 weeks for stab wound (SW) and stab wound with injection of vehicle solution, HBSS (SW-control). The boxes correspond to median values with indication of the 25 and 75 percentiles, and the whiskers show the minimum and maximum values. Wilcoxon matched-pairs signed rank test was used and the horizontal lines indicate statistical differences (*p < 0.05, **p < 0.01, ***p < 0.001).

seen for the group receiving 10 mm long nanowires (4.45% ± 0.52%) compared to the control group (p = 0.056) (Fig. 4).

3.3.3. Neuronal density

There was no significant difference in the number of neuronal nuclei (NeuN), in the inner (Fig. 4) or outer ROI (data not shown) between the different groups 12 weeks after the injections. This suggests that no nanowire-induced neurotoxic effect was present at this time point.

3.3.4. Cellular density

A significant increase in the number of cell nuclei (DAPI) was found in the group receiving 10 mm long nanowires compared to the control group (p = 0.0071), with a mean nuclei density of 0.00265 ± 7.7e-005 mm~2 and 0.00227 ± 4.7e-005 mm~2, respectively (Fig. 4). This corresponds to a 16.7% increase in cell nuclei in the 10 mm nanowire-injected group compared to control, which may be explained by the increase of ED1 and GFAP- positive cells in this group.

Fig. 4. lnflammatory response and neuronal density after 12 weeks. (a—d) Quantification in the inner ROl (0—50 mm) of ED1-positive area (a), GFAP-positive area (b), NeuN density (c), and cell nuclei density (d) at 12 weeks for nanowire- injected animals and control group. (e) Quantification in the outer ROl (50—150 mm) of GFAP-positive area at 12 weeks for nanowire-injected animals and control group. Hf corresponds to hafnium oxide coated nanowires. The boxes correspond to median values with indication of the 25 and 75 percentiles, and the whiskers show the minimum and maximum values. Kruskal Wallis with Dunn's multiple comparison test was used and the horizontal lines indicate statistical differences (*p < 0.05, **p < 0.01, ***p < 0.001).

3.4. Effect of nanowire length on the brain tissue response after one year

Representative fluorescence microscopy images of the scar tissue area after injections of control solution and nanowires of different lengths one year after surgery are shown in Fig. 5.

p = 0.0003) and 5 mm long nanowires (mean area 1.65% ± 0.32%, p = 0.029) compared to the control group (mean area 0.690% ± 0.081%) (Fig. 6). In the outer ROI, however, there was no significant difference in ED1-positive area comparing the groups (data not shown) and the overall immunoreactive area for ED1 in this ROI was close to zero in all groups, comparable to the findings at 12 weeks.

3.4.1. Microglia/macrophage response

In the inner ROI a significant increase in ED1-positive area was found in the group receiving 10 mm (mean area of 2.82% ± 0.59%,

3.4.2. Astrocytic reactivity

In the inner ROI, there was no significant difference in GFAP-positive area between any of the groups (Fig. 6). In the outer ROI,

Control 2 pm HfOx 5 pm HfOx 10 |jm HfOx

Fig. 5. ¡n vivo tissue response after one year. Representative fluorescent images of the tissue response to injections of control and 2, 5 and 10 mm long HfOx (hafnium oxide) -coated nanowires at one year. ED1-positive cells (green), GFAP-positive cells (red), and cell nuclei (blue) and merge. Scale bars: 100 mm.

however, a significant increase of GFAP-positive area was found in the group receiving 10 mm long nanowires (7.31% ± 1.1%) compared to the group receiving 5 mm long nanowires (3.61% ± 0.46%) (p = 0.042) (Fig. 6).

3.4.3. Neuronal density

In the inner ROI, a significant decrease in the neuronal nuclei density was found in the group receiving 10 mm long nanowires

(66.6e-005 ± 4.9e-005 mm 2) compared to the control group (88.0e-005 ± 5.2e-005 mm~2) (p = 0.032) (Fig. 6). This corresponds to a 24.2% loss of neuronal nuclei in the 10 mm nanowire group compared to control. In the outer ROI, however, no significant difference in neuronal density was found when comparing the different groups (data not shown). This finding suggests that the reduction of neurons observed in the 10 mm group in the inner ROI,

Fig. 6. Inflammatory response and neuronal density after one year. (a—d) Quantification in the inner ROI (0—50 mm) of ED1-positive area (a), GFAP-positive area (b), NeuN density (c), and cell nuclei density (d) at one year for nanowire-injected animals and control group. (e—f) Quantification in the outer ROI (50—150 mm) of GFAP-positive area (e) and NeuN density (f) at one year for nanowire-injected animals and control group. Hf corresponds to hafnium oxide coated nanowires. The boxes correspond to median values with indication of the 25 and 75 percentiles, and the whiskers show the minimum and maximum values. Kruskal Wallis with Dunn's multiple comparison test was used and the horizontal lines indicate statistical differences (*p < 0.05, **p < 0.01, ***p < 0.001).

was not due to a displacement of neurons from the inner ROI to the outer ROI.

3.4.4. Cellular density

There was no significant difference in the number of cell nuclei when comparing the different groups one year after the injections (Fig. 6).

3.5. Confocal analysis of the nanowire distribution

At both time points, confocal examination of the groups receiving nanowires showed that ED1-positive cells had engulfed most of the 2 mm long nanowires. However, some of the 5 mm and 10 mm long nanowires could be seen free in the tissue, i.e. not engulfed by any type of labelled cell, (Figs. 7—8). We also observed

nanowires lining blood vessel walls at both time points, free or engulfed by ED1-positive cells (Figs. 7—8). Furthermore, in the centre of the injection tract, small aggregations of co-localized ED1 and GFAP staining could be observed. These aggregates resembled cell debris or cell aggregates, and the dominating majority contained nanowires, see Fig. 9. However, they were regularly devoid of DAPl-positive nuclei, suggesting that they were not viable cells. The aggregates were mostly observed in the groups with longer nanowires (5 and 10 mm) and were rarely found at the early time point, i.e. at 12 weeks.

4. Discussion

To assess possible risks involved in using future nanowire-structured neural interfaces, we analyzed the tissue response and neuronal survival after injection of non-degradable, high-aspect-ratio nanoparticles into the brain, for a period corresponding to approximately half the lifespan of a rat, i.e. one year. lmportantly, while the inflammatory tissue response depends on the nanowire length, it did not escalate between 12 weeks and one year. However, for the longest nanowires, the increased microglial activation found at 12 weeks did persist for one year and the neuronal density was reduced as compared to control after one year.

At 12 weeks post implantation, ED1 and GFAP immunoreactivity was increased only in the 10 mm nanowire group as compared to control. After one year, both 5 and 10 mm nanowires groups exhibited an increase in ED1 immunoreactivity compared to the control group. The increased ED1 response observed towards longer nanowires might reflect a size dependent limit of the phagocytic capacity of macrophages and microglial cells. lndeed, the fibre pathogenicity paradigm predicts that long, thin and biostable fibres will lead to frustrated phagocytosis and elicit a long lasting/chronic inflammatory response. Such asbestos-like patho-genicity has been observed after implantation of long CNTs in the mice abdominal tissue [20]. More recently, the same group used silver nanowires with narrow length distributions to investigate the length threshold for asbestos-like pathogenicity in the lung and the pleura, which was determined to be around 5—10 mm [21,33]. The present results are consistent with these findings. The observed cell debris/cell aggregates, containing 5 and 10 mm long nanowires (Fig. 9), could be compared to granulomas, which supports the hypothesis that the phagocytosis of the longer nanowires cannot be completed normally. lt may be speculated that longer nanowires exert a higher strain force on the membrane of the phagocytizing cells since these nanowires are close to the size of the microglial cells [34]. This strain might impede the engulfment of nanowires or lead to membrane deformation or even piercing of the cell membrane during engulfment or migration, which could lead to cell death.

ln the present study, we found a higher number of 5 and 10 mm long nanowires that were free in the brain tissue (i.e. not engulfed by any type of labelled cells) compared to 2 mm long nanowires (which were almost always engulfed by ED1-positive cells). This suggests that when unable to phagocytize long nanowires, cells either undergo apoptosis without signalling to

Fig. 7. Confocal images 12 weeks after nanowire injection. Representative laser scanning confocal microscopy images of the scar after injection of 2 (a), 5 (b), and 10 mm (c) long nanowires at 12 weeks. Free nanowires, not engulfed by any labelled cell type, were observed in the scar area. ED1-positive cells (green), GFAP-positive cells (red), cell nuclei (blue), nanowires (white), free nanowires (/), and blood vessels (*). Scale bars: 30 mm.

b * ^ if * — — $ * w v £

V * < -

V > »3 ffm r

• * v * * • ^ A- * SÄ > H » J

recruit new cells, or the immune cells do not detect some of the long nanowires.

Importantly, our results showed no indication of an escalation of the inflammatory tissue response over time. However, the neuronal density in the immediate vicinity of the injection tract for the 10 mm nanowire-injected group was reduced as compared to control, at the one year follow up, indicating that neurons in the 10 mm group are slowly dying over time. This might be related to the increased glial response to 10 mm nanowires as compared to shorter nanowires. It has been reported that activated glial cells may continuously secret neurotoxic substances, which could explain the neuronal cell loss one year post nanowire injection [35-37].

The nanowires used in this study are non-degradable by design and therefore cannot be cleared by phagocytic activity. However, we observed nanowires lining blood vessel walls near the injection site, both free and engulfed by EDl-positive cells (Figs. 7-8). This may suggest a clearance mechanism by translocation, where ED1-positive cells migrate and transport nano-wires to blood vessels in their vicinity. Similar observations have been made several decades ago after injection of colloidal carbon into the brain [38]. Our previous study, on short and biodegradable nanowires in the brain, also suggested a nanowire clearance mechanism through the blood vessels [28]. It is known that the CNS lacks conventional lymphatic drainage. Recent studies have shown that the cerebrospinal fluid drains mainly from the subarachnoid space through the cribriform plate into nasal lymphatics and to deep cervical lymph nodes, and infiltrating monocytes are also capable of migrating via the cribriform plate [39,40]. Drainage of interstitial fluid and solutes, on the other hand, has been suggested to mainly occur through the basement membrane in the walls of capillaries, ultimately ending up in the cervical lymph nodes [39]. Hence, it is possible that the microglial cells and macrophages containing nanowires follow these routes as well. In our case, no nanowires could be detected in the dissected lymph nodes, suggesting that, either the injected nanowires could not leave the brain, or that too small amounts of nanowires were cleared out of the brain to be detected.

Vehicle injections of HBSS led to a smaller astrocytic reactivity compared to stab wound without vehicle injection. This suggests that HBSS may reduce the adverse effects caused by the stab wound. Possible explanations for this could be that the injected solution dilutes the concentration of inflammatory cytokines locally, thereby, indirectly reducing the ensuing inflammatory response at the injection site. Further studies are needed in order to determine the mechanisms behind this observation.

5. Conclusions

In conclusion, we show that the brain inflammatory response to the implantation of biostable nanorods and the loss of nearby neurons are strongly length dependent. The results indicate that, similar to the lungs and abdomen, a persistent, albeit not escalating, inflammatory response occurs in the brain when high-aspect-ratio particles, with a length equal to or greater than the size of immune cells, are present in the tissue. The obtained results

Fig. 8. Confocal images one year after nanowire injection. Representative laser scanning confocal microscopy images of the scar after injection of 2 (a), 5 (b), and 10 mm (c) long nanowires at one year. Free nanowires, not engulfed by any labelled cell type, were observed in the scar area for the longer nanowires (5 and 10 mm). EDl-positive cells (green), GFAP-positive cells (red), cell nuclei (blue), nanowires (white), free nanowires (/), and blood vessels (*). Scale bars: 30 mm.

Fig. 9. Confocal image of an aggregate at one year after nanowire injection. Laser scanning confocal microscopy image of the scar area after injection of 5 mm long nanowires at one year showing the aggregates of cell debris or cells observed in the tissue at one year. ED1-positive cells (green), GFAP-positive cells (red), cell nuclei (blue), and nanowires (white). Scale bars: 10 mm.

stress the importance of using nanorods that are shorter than the size of brain immune cells, when modifying the surface of neural interfaces in order to improve their performance.

Acknowledgements

The authors thank Suzanne Rosander-Jonsson and Agneta SanMartin for excellent technical assistance. The confocal microscopy was performed at the Microscopy Facility at the Department of Biology, Lund University.

This work was funded by the The Knut and Alice Wallenberg Foundation (project number: KAW 2004-0119), a Linnaeus grant (project number: 600012701) from the Swedish Research Council, The Crafoord Foundation and The Royal Physiographic Society in Lund.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2014.11.051.

References

[1] Awan NR, Lozano A, Hamani C. Deep brain stimulation: current and future perspectives. Neurosurg Focus 2009;27:E2.

[2] Daly JJ, Wolpaw JR. Brain-computer interfaces in neurological rehabilitation. Lancet Neurol 2008;7:1032—43.

[3] Hochberg LR, Serruya MD, Friehs GM, Mukand JA, Saleh M, Caplan AH, et al. Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 2006;442:164—71.

[4] Kipke DR, Shain W, Buzsaki G, Fetz E, Henderson JM, Hetke JF, et al. Advanced neurotechnologies for chronic neural interfaces: new horizons and clinical opportunities. J Neurosci 2008;28:11830—8.

[5] Biran R, Martin DC, Tresco PA. Neuronal cell loss accompanies the brain tissue response to chronically implanted silicon microelectrode arrays. Exp Neurol 2005;195:115—26.

[6] Marin C, Fernandez E. Biocompatibility of intracortical microelectrodes: current status and future prospects. Front Neuroeng 2010;3:8.

Szarowski DH, Andersen MD, Retterer S, Spence AJ, Isaacson M, Craighead HG, et al. Brain responses to micro-machined silicon devices. Brain Res 2003;983: 23—35.

Williams JC, Hippensteel JA, Dilgen J, Shain W, Kipke DR. Complex impedance spectroscopy for monitoring tissue responses to inserted neural implants. J Neural Eng 2007;4:410—23.

Biran R, Martin DC, Tresco PA. The brain tissue response to implanted silicon microelectrode arrays is increased when the device is tethered to the skull. J Biomed Mater Res A 2007;82:169—78.

Gilletti A, Muthuswamy J. Brain micromotion around implants in the rodent somatosensory cortex. J Neural Eng 2006;3:189—95.

Lind G, Eriksson Linsmeier C, Schouenborg J. The density differences between tissue and neural probes is a key factor for glial scarring. Sci Rep 2013;3:2942.

Thelin J, Jorntell H, Psouni E, Garwicz M, Schouenborg J, Danielsen N, et al. Implant size and fixation mode strongly influence tissue reactions in the CNS. Plos One 2011;6:e16267.

Cellot G, Cilia E, Cipollone S, Rancic V, Sucapane A, Giordani S, et al. Carbon nanotubes might improve neuronal performance by favouring electrical shortcuts. Nat Nanotechnol 2009;4:126—33.

Keefer EW, Botterman BR, Romero MI, Rossi AF, Gross GW. Carbon nanotube coating improves neuronal recordings. Nat Nanotechnol 2008;3:434—9. Hallstrom W, Martensson T, Prinz C, Gustavsson P, Montelius L, Samuelson L, et al. Gallium phosphide nanowires as a substrate for cultured neurons. Nano Lett 2007;7:2960—5.

Piret G, Perez M-T, Prinz CN. Neurite outgrowth and synaptophysin expression of postnatal CNS neurons on GaP nanowire arrays in long-term retinal cell culture. Biomaterials 2013;34:875—87.

Prinz C, Hallstrom W, Martensson T, Samuelson L, Montelius L, Kanje M. Axonal guidance on patterned free-standing nanowire surfaces. Nanotech-nology 2008;19:345101.

Suyatin DB, Wallman L, Thelin J, Prinz CN, Jäorntell H, Samuelson L, et al. Nanowire-based electrode for acute in vivo neural recordings in the brain. PLoS One 2013;8:e56673.

Donaldson K, Murphy F, Duffin R, Poland C. Asbestos, carbon nanotubes and the pleural mesothelium: a review of the hypothesis regarding the role of long fibre retention in the parietal pleura, inflammation and mesothelioma. Part Fibre Toxicol 2010;7:5.

Poland C, Duffin R, Kinloch I, Maynard A, Wallace W, Seaton A, et al. Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like pathogenicity in a pilot study. Nat Nanotechnol 2008;3:423—8. Schinwald A, Murphy FA, Prina-Mello A, Poland CA, Byrne F, Movia D, et al. The threshold length for fiber-induced acute pleural inflammation: shedding light on the early events in asbestos-induced mesothelioma. Toxicol Sci 2012;128:461—70.

Suyatin DB, Hallstrom W, Samuelson L, Montelius L, Prinz CN, Kanje M. Gallium phosphide nanowire arrays and their possible application in cellular force investigations. J Vac Sci Technol B 2009;27:3092—4. Hallstrom W, Lexholm M, Suyatin DB, Hammarin G, Hessman D, Samuelson L, et al. Fifteen-piconewton force detection from neural growth cones using nanowire arrays. Nano Lett 2010;10:782—7.

Mohammadi S, Esposito M, Cucu M, Ericson LE, Thomsen P. Tissue response to hafnium. J Mater Sci Mater Med 2001;12:603—11.

Matsuno H, Yokoyama A, Watari F, Uo M, Kawasaki T. Biocompatibility and

osteogenesis of refractory metal implants, titanium, hafnium, niobium,

tantalum and rhenium. Biomaterials 2001;22:1253—62.

Maggiorella L, Barouch G, Devaux C, Pottier A, Deutsch E, Bourhis J, et al.

Nanoscale radiotherapy with hafnium oxide nanoparticles. Future Oncol

2012;8:1167—81.

Loo Y-L, Willett RL, Baldwin KW, Rogers JA. Additive, nanoscale patterning of metal films with a stamp and a surface chemistry mediated transfer process: applications in plastic electronics. Appl Phys Lett 2002;81:562—4. Eriksson Linsmeier C, Prinz CN, Pettersson LME, Caroff P, Samuelson L, Schouenborg J, et al. Nanowire biocompatibility in the brain — looking for a needle in a 3D stack. Nano Lett 2009;9:4184—90.

Davis RK, Stevenson GT, Busch KA. Tumor incidence in normal sprague-dawley female rats. Cancer Res 1956;16:194—7.

Prejean JD, Peckham JC, Casey AE, Griswold DP, Weisburger EK, Weisburger JH. Spontaneous tumors in Sprague-Dawley rats and swiss mice. Cancer Res 1973;33:2768—73.

Nakazawa M, Tawaratani T, Uchimoto H, Kawaminami A, Ueda M, Ueda A, et al. Spontaneous neoplastic lesions in aged Sprague-Dawley rats. Exp Anim 2001;50:99—103.

Thompson SW, Huseby RA, Fox MA, Davis CL, Hunt RD. Spontaneous tumors in the Sprague-Dawley rat. J Natl Cancer Inst 1961;27:1037—57. Schinwald A, Chernova T, Donaldson K. Use of silver nanowires to determine thresholds for fibre length-dependent pulmonary inflammation and inhibition of macrophage migration in vitro. Part Fibre Toxicol 2012;9:47.

Agarwal R, Singh V, Jurney P, Shi L, Sreenivasan SV, Roy K. Mammalian cells preferentially internalize hydrogel nanodiscs over nanorods and use shape-specific uptake mechanisms. Proc Natl Acad Sci U S A 2013;110: 17247—52.

Hanisch UK. Microglia as a source and target of cytokines. Glia 2002;40: 140—55.

[36] Streit W, Mrak R, Griffin WS. Microglia and neuroinflammation: a pathological perspective. J Neuroinflammation 2004;1:14.

[37] Verkhratsky A, Butt A. General pathophysilogy of glia. In: Verkhratsky A, Butt A, editors. Glial neurobiology: a textbook. Chichester, UK: John Wiley & Sons, Ltd; 2007. p. 153-65.

[38] McKeever PE, Balentine JD. Macrophages migration through the brain parenchyma to the perivascular space following particle ingestion. Am J Pathol 1978;93:153-64.

[39] Weller RO, Djuanda E, Yow HY, Carare RO. Lymphatic drainage of the brain and the pathophysiology of neurological disease. Acta Neuropathol 2009;117: 1-14.

[40] Kaminski M, Bechmann I, Pohland M, Kiwit J, Nitsch R, Glumm J. Migration of monocytes after intracerebral injection at entorhinal cortex lesion site. J Leukoc Biol 2012;92:31-9.