hyaluronic acid scaffold has a neuroprotective effect in ... · hyaluronic acid (ha) is a linear...

laboratory investigationJ neurosurg spine 25:114–124, 2016

Spinal cord injuries (SCIs) lead to immediate and usu-ally permanent lifestyle alterations. Most patients require lifelong medical care, and some patients

cannot live independently. They suffer from frequent com-plications, such as ulcers, urinary tract infections, pneu-monia, and fractures.30 The individual lifetime financial burden can exceed $2 million, and in the United States, $7 billion is spent annually to care for people with SCIs.33

Initial trauma to the spinal cord can sever neurons, which leads to a loss of signal from a certain part of the body to the brain; however, the full SCI occurs in 2 phas-es.12,14,17,20,27,29,33,40,46,53 The initial trauma is followed by in-flammation that leads to fibrotic scarring, glial scarring, and cavity formation. First, inflammatory cells infiltrate the spinal cord shortly after injury, destroying intact ax-ons and causing demyelination.20,40 Astrocytes quickly

abbreviations BBB = Basso, Beattie, and Bresnahan; DTP = 3,3´-dithiobis (propanoic acid); DTT = dithiothreitol; HA = hyaluronic acid; PBS = phosphate-buffered saline; PEGDA = poly(ethylene glycol)-diacrylate; R/L = right/left; SCI = spinal cord injury; SEM = scanning electron microscope; TEM = transmission electron microscopy.submitted May 28, 2015. accepted September 23, 2015.include when citing Published online March 4, 2016; DOI: 10.3171/2015.9.SPINE15628.

Hyaluronic acid scaffold has a neuroprotective effect in hemisection spinal cord injurysergiy v. Kushchayev, md,1 morgan b. giers, phd,1 doris hom eng, phd,2 nikolay l. martirosyan, md,1 Jennifer m. eschbacher, md,1 martin m. mortazavi, md,1 nicholas theodore, md,1 alyssa panitch, phd,3 and mark c. preul, md1

1Department of Neurosurgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix; 2School of Biological and Health Systems Engineering, Arizona State University, Tempe, Arizona; and 3Weldon School of Biomedical Engineering, Purdue University, West Lafayette, Indiana

obJective Spinal cord injury occurs in 2 phases. The initial trauma is followed by inflammation that leads to fibrous scar tissue, glial scarring, and cavity formation. Scarring causes further axon death around and above the injury. A reduction in secondary injury could lead to functional improvement. In this study, hyaluronic acid (HA) hydrogels were implanted into the gap formed in the hemisected spinal cord of Sprague-Dawley rats in an attempt to attenuate damage and regenerate tissue.methods A T-10 hemisection spinal cord injury was created in adult male Sprague-Dawley rats; the rats were as-signed to a sham, control (phosphate-buffered saline), or HA hydrogel–treated group. One cohort of 23 animals was followed for 12 weeks and underwent weekly behavioral assessments. At 12 weeks, retrograde tracing was performed by injecting Fluoro-Gold in the left L-2 gray matter. At 14 weeks, the animals were killed. The volume of the lesion and the number of cells labeled from retrograde tracing were calculated. Animals in a separate cohort were killed at 8 or 16 weeks and perfused for immunohistochemical analysis and transmission electron microscopy. Samples were stained using H & E, neurofilament stain (neurons and axons), silver stain (disrupted axons), glial fibrillary acidic protein stain (astrocytes), and Iba1 stain (mononuclear cells).results The lesions were significantly smaller in size and there were more retrograde-labeled cells in the red nuclei of the HA hydrogel–treated rats than in those of the controls; however, the behavioral assessments revealed no differ-ences between the groups. The immunohistochemical analyses revealed decreased fibrous scarring and increased retention of organized intact axonal tissue in the HA hydrogel–treated group. There was a decreased presence of inflam-matory cells in the HA hydrogel–treated group. No axonal or neuronal regeneration was observed.conclusions The results of these experiments show that HA hydrogel had a neuroprotective effect on the spinal cord by decreasing the magnitude of secondary injury after a lacerating spinal cord injury. Although regeneration and be-havioral improvement were not observed, the reduction in disorganized scar tissue and the retention of neurons near and above the lesion are important for future regenerative efforts. In addition, this gel would be useful as the base substrate in the development of a more complex scaffold.http://thejns.org/doi/abs/10.3171/2015.9.SPINE15628Key words hyaluronic acid; lesion size; neuroprotective; red nucleus; retrograde tracing; spinal cord injury; rat

©AANS, 2016J neurosurg spine Volume 25 • July 2016114

hyaluronic acid hydrogel is neuroprotective in spinal cord injury

J neurosurg spine Volume 25 • July 2016 115

become reactive and start to form a glial scar composed of dense astrocytes and altered extracellular matrix, such as increased chondroitin sulfate. The glial scar seals the blood-brain barrier and blocks infiltrating mononuclear cells and fibroblasts but also inhibits axon regeneration.20,27 The fibrotic and glial scars are incorporated together in a region of amorphous dense tissue. Furthermore, phagocy-tosis of dying cells and debris can form cavities inside the spinal cord.16 In addition to the inflammatory pathways, nonpathological pathways inhibit the ability of the cen-tral nervous system to regenerate, such as by expressing the neurite outgrowth inhibitor proteins.39,51 Therefore, to regenerate after injury in the central nervous system, ax-ons must overcome several formidable barriers, including dense scar tissue, fluid-filled gaps in nervous tissue, and nonpathological inhibitory pathways.7,16,20,27,39,43

Spinal cord injury involves some combination of contu-sion, crush, pinch, twist and torsion, partial tearing, and vascular injury. Although all types of SCI induce a sec-ondary-phase scarring process, there are some differences seen clinically in the rates of recovery, depending on the specific injury type.

Clinically, SCI is managed by stabilizing the spine, de-compressing the spinal cord, and repairing the dura mater where necessary. Although clinical trials on SCI are cur-rently underway, there are no devices or treatments ap-proved by the US Food and Drug Administration that are available for the reduction of secondary injury or restora-tion of function after SCI.36,45

Hyaluronic acid (HA) is a linear glycosaminoglycan ex-tracellular matrix component found in all tissues.21 It has been used in many medical applications and is frequently investigated as a potential biomaterial for tissue engineer-ing.1,3,11,13,18,19,28,37,54 Hyaluronic acid is highly biocompat-ible, immunologically and chemically inert, and nontoxic; thus, using it prevents the problem of neural tissue rejec-tion. Furthermore, high-molecular-weight HA promotes neuroprotective cell-signaling cascades.2 Hyaluronic acid is often used in a modified hydrogel because native HA is not mechanically durable. These HA hydrogels can adhere firmly to host tissue and bridge the whole spinal cord le-sion. We previously reported that the use of a cross-linked HA scaffold promotes robust neurite outgrowth in vitro.25 Khaing et al.28 implanted an HA hydrogel in a rat hemi-section model and found that the hydrogel decreased the number of reactive astrocytes at short and long time points and decreased a glial scar component (chondroitin sulfate proteoglycan) at a short time point. They also found that the HA hydrogel decreased the presence of microglia and macrophages up to 10 days after injury; however, Khaing et al.28 did not study the effect of this decreased inflam-mation on neuron survival. Austin et al.3 studied an HA/methylcellulose composite hydrogel in vivo by using a compression-injury model. They found that at 6 weeks, the composite hydrogel reduced fibrotic and glial inflamma-tion and scarring. Furthermore, their results showed that the composite gel improved some aspects of neural and behavioral outcomes.

Because HA hydrogels have shown regeneration-pro-moting and immunomodulating potential, they could be useful in treating laceration-type SCIs. In this study, HA

hydrogels were used to fill the hemisection spinal cord le-sions of adult rats to evaluate their neuroprotective and neuroregenerative potential at time points up to 16 weeks.

methodsHyaluronic acid (molecular mass 110 kD) was pur-

chased from Genzyme. Poly(ethylene glycol)-diacrylate (PEGDA) (molecular mass 3400 Da) was purchased from Nektar Therapeutics. Phosphate-buffered saline (PBS), sy-ringe filters (0.2 μm), syringes, and sodium hydroxide were purchased from VWR International LLC. 1-Ethyl-3-[3-(dimethylamino)propyl]carbodiimide, 3,3´-dithiobis (pro-panoic acid) (DTP), dithiothreitol (DTT), and hydrazine hydrate were obtained from Sigma Chemical Co. Dextran microruby (D7162), wheat germ agglutinin-Oregon green 488 (W6748), and Prolong gold antifade reagents were pur-chased from Invitrogen Corp. Hamilton syringes and nee-dles were purchased from Hamilton Company. Sprague-Dawley rats were supplied by Charles River Laboratories International, Inc.

hydrogel preparationThe hydrogel was prepared according to the method

outlined by Shu et al.44 In brief, this method created a 2-liquid component system that, when mixed together, formed a hydrogel. In this experiment, the components re-mained separated until the time of surgery, when the gels were allowed to form in situ. The first component was a thiolated HA created by adding DTP to the glucuronate carboxyl group of HA and removing the disulfide bond us-ing DTT. The second component was a PEGDA solution, which cross-links the thiol groups of HA.

The thiolation of HA was performed in a 2-step reac-tion. First, we performed a 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide–mediated reaction in which DTP and HA reacted at pH 4.75 for 1 hour. The reaction was terminated by raising the pH to 7.0. The disulfide bond in DTP was broken, which formed a thiol group by add-ing the reducer DTT at pH 8.5. The solution was dialyzed against a buffer (100 mM NaCl, 0.3 mM HCl) at pH 3.5 to remove reactants and by-products.

Solutions of 0.52%, 0.64%, 0.75%, 1%, and 1.25% (w/v) thiolated HA and a 2.25% (w/v) solution of PEGDA were prepared in PBS at pH 7.4. All solutions were fil-tered through a 0.2-μm syringe filter to sterilize them. At the time of surgery, the HA and PEGDA solutions were mixed together at a 0.45 PEGDA/thiolated HA ratio. After mixing, the PEGDA acted as a bifunctional electrophilic cross-linker, which reacted with the nucleophilic thiol groups in a Michael-type addition. The cross-linking cre-ated a stable hydrogel.

hydrogel analysisThe morphologies of the HA hydrogels were visualized

with a scanning electron microscope (SEM). Fully cross-linked HA gels of 0.64%, 0.75%, 1%, and 1.25% (w/v) thio-lated HA were flash frozen by immersion in liquid nitro-gen, fractured with a scalpel, and placed immediately on a lyophilizer for freeze-drying. The HA gels were then gold coated (10 μm), and images were acquired with the SEM

s. v. Kushchayev et al.

J neurosurg spine Volume 25 • July 2016116

(XL30 ESEM-FEG, FEI Co.) at 15 kV and a spot size of 2. The SEM images were analyzed for pore size by using MATLAB (MathWorks, Inc.).

surgical procedureA total of 89 adult male Sprague-Dawley rats (weigh-

ing 250–350 g) were used in 2 separate cohorts of ani-mals. The experimental design of each cohort is shown in Fig. 1 and described below; in brief, each cohort contained a sham surgery group (negative control), a positive control group, and an HA hydrogel–treated group.

Each animal was anesthetized with an intramuscu-lar injection of ketamine/xylazine/acepromazine (50/7/1 mg/g of body weight). Fluoroscopy was used to identify the T9–10 level, which was marked with a gold bead in-serted subcutaneously. Under sterile conditions, a T-10 laminectomy was performed, and the dura was opened. In each rat in the control and HA hydrogel–treated groups, a lateral hemisection of the left spinal cord was performed at T9–10 using an operating microscope (Zeiss OPMI Pentero, Carl Zeiss Meditec, Inc.). The hemisection was carefully verified by placing a dissector in the gap. Figure

2 shows the spinal cord before, during, and after the hemi-section through the operating microscope. Adequacy of the hemisection was confirmed by 3 independent observ-ers, including the surgeon using the operating microscope. High-magnification digital images of each completed hemisection were obtained. Next, either the experimental HA hydrogel or PBS was injected into the hemisection lesion until the lesion was filled but not expanded. This methodology led to a variable HA hydrogel volume that was approximately 20–30 μl in all cases. The surgeon was blinded to the choice of material for injections, although he could possibly feel and see the difference between the PBS and HA hydrogel, depending on the viscosity of the substances. Finally, the muscular fascia was sutured with Vicryl 4.0 (Ethicon, Inc.), and the skin was sutured with Ethilon 3.0 (Ethicon). The dura was not closed. The sur-geries were identical in the sham group except that hemi-section of the cord was not performed.

Each animal received ceftriaxone for 3 days after sur-gery. Manual bladder expression was performed until the animals recovered bladder control. All animal experi-ments were performed with approval from the Institution-

Fig. 1. Schematic representation of the experimental design.



al Animal Care and Use Committee at St. Joseph’s Hospi-tal and Medical Center.

retrograde tracingTwenty-three adult male Sprague-Dawley rats (weigh-

ing 250–350 g) were randomly assigned to 1 of 3 groups using Excel (Microsoft Corp.): sham (n = 3), control (n = 5), or HA hydrogel treated (n = 15). Twelve weeks after surgery, each animal received a retrograde tracer injec-tion caudal to the lesion site. The tracer traveled up the axons and labeled cells in the corresponding area of the red nucleus.

For the retrograde tracer injection, each rat first under-went a laminectomy at L-2, localized using fluoroscopy. Surgifoam was placed around the injection site to prevent any leakage of the tracer. An incision was made in the pia mater using a 32-gauge needle. An automatic nanoinjec-tor (World Precision Instruments, Inc.) with a 33-gauge Hamilton needle was positioned in the spinal cord using a Kopf stereotaxic system. Care was taken to avoid damag-ing the spinal cord with excess pressure from the injec-tion. For this reason, the 33-gauge needle was first posi-tioned 2 mm deep in the left paramedian gray matter and then was retracted 0.5 mm to allow space for the tracer to be injected. Over a period of 60 seconds, 100 nl of 2% Fluoro-Gold (Fluorochrome, LLC), a retrograde tracer, in PBS was injected to allow for pressure equilibration in the spinal cord. Over the next 60 seconds, the needle was retracted 0.5 mm. Then, a second 100 nl of Fluoro-Gold was injected over 60 seconds. After a 60-second delay, the needle was removed from the spinal cord.

Two weeks after labeling, each animal underwent induction of anesthesia and was killed by transcardial infusion of 200 ml of saline followed by 250 ml of 4% paraformaldehyde in 0.1 M PBS, pH 7.4. The brains and spinal cords were dissected and placed in 4% parafor-maldehyde for 24 hours and then in 30% sucrose for 24 hours. Thereafter, the samples were sent to NeuroScience Associates for cryoprotection (overnight in 20% glycerol and 2% dimethyl sulfoxide), gelatin embedding, and sec-tioning to a 40-μm thickness. Every other section was mounted. Slides were studied under a fluorescent micro-scope (Olympus BX61, Olympus America, Inc., Scientific Equipment Group) with a chroma Fluoro-Gold filter set (excitation/emission 350/515 nm) and a 400-nm dichroic beam splitter (50 nm wide with 80% transmission).

Cells present on each side of the red nucleus were counted in every slide and multiplied by 2 because only

every other slide was mounted. The cells in the right red nucleus corresponded to the lesion side because the rubro-spinal tracts cross.

lesion sizeMicroscope images of each of the histological slides

from the retrograde tracing experiment were collected. The lesions were assumed to be rectangular in shape. Le-sion sizes were approximated using the following equa-tion:

Vlesion = N × L × W × (T + S),

where N is the number of histological sections, L is the vertical length of the lesion, W is one-half the width of the thoracic spinal cord, T is the histological slice thickness, and S is the distance between histological slices (which is the same as T). The volume of the HA hydrogel was excluded from lesion-size calculations.

behavioral testingThe motor function of each hind limb was assessed by

using the Basso, Beattie, and Bresnahan (BBB) scale. In brief, during 4 minutes of free activity, we used a 21-point scale to evaluate joint movements, paw placement, weight support, and forelimb-to–hind limb coordination. As-sessments were performed weekly for 12 weeks.5 Scores were averaged between 2 independent observers who were blinded to the treatment group.

histochemical analysisSixty-six adult male Sprague-Dawley rats (weighting

250–350 g) were assigned to 1 of 3 groups: sham (n = 12), control (n = 12), or HA hydrogel treated (n = 42). Either 8 or 16 weeks after surgery, the animals were split evenly into 1 of the 2 time point groups and anesthetized and killed as described above (Fig. 1). Samples were cyroprotected in 30% sucrose for 24 hours before extraction of a 1.0-cm length of spinal cord at the injury site. The section of spi-nal cord was embedded in paraffin and serially sectioned parasagittally in multiple 20-μm slices. The sections were stained with H & E for general morphology. Immunohis-tochemistry was performed to study specific biomarkers. Samples were stained with a neurofilament stain to study neurons and axons (mouse antineurofilament 200, 1:500; Sigma), the Nauta method of silver staining to selectively stain disrupted axons (performed by NeuroScience As-

Fig. 2. Pictures of the hemisection surgery through the operative microscope showing a dissector applied to the anterior aspect of the spinal cord to prevent damage of the anterior dural venous plexus during hemisectioning (a), a hemisection performed with a scalpel (b), and instruments removed after the hemisection (c).



sociates),48 an astrocyte stain (mouse antiglial fibrillary acidic protein, 1:200; Boehringer-Mannheim Chemicon), and an Iba1 stain for mononuclear cells such as macro-phages and microglia (antibody against ionized calcium-binding adaptor molecule 1, performed by NeuroScience Associates). All microscopic analyses were performed by a professional pathologist.

transmission electron microscopyRegenerating axons were identified by preparing the

spinal cord from 1 rat in each histochemical analysis group for transmission electron microscopy (TEM). Tis-sues were fixed overnight in 3.0% glutaraldehyde buffered to a pH of 7.4 with Millonig phosphate buffer and post-fixed for 2 hours with 2.0% osmium tetroxide buffered to a pH of 7.4 with Millonig phosphate buffer. The samples were then dehydrated with ethanol and propylene oxide and embedded in Epon substitute (LX-112). Blocks were sectioned to 80 nm using a diamond knife (DiATOME Ltd), mounted on 150-mesh copper grids (Ted Pella, Inc.), and stained with lead citrate and uranyl acetate. Transmis-sion electron microscopy images were taken with a Phil-ips CM100 transmission electron microscope (FEI Co.). Images were captured with a Gatan 697 or Gatan Orius (SC1000 1) camera. Image contrast and brightness were adjusted using Adobe Photoshop.

statistical analysisAll parametric data were analyzed with an ANOVA

test in Statistica (StatSoft, Inc.). The differences in number of cells were analyzed using repeated-measures ANOVA. All other data were analyzed with 1-way ANOVA. In all cases, the normality assumption was checked with the normal probability plot. Post hoc t-tests were performed in Excel (Microsoft Corp.) to determine specific differences between groups; p values less than 0.05 were considered significant. Error bars shown in the figures represent the SD. In Results, mean values are presented ± SD.

resultsproperties of the ha hydrogels

Pore size and channel formation are important for cell infiltration.9,10 The SEM images showed that the pores formed interconnected channels in all the HA hydrogels (Fig. 3). Furthermore, the pore sizes (mean 131 ± 87 to 210 ± 59 μm) were not significantly different between the various weight-percent gels listed in “Hydrogel Prepara-tion” (Fig. 4). Because there were no apparent differences between the different weight-percent gels, data from all HA hydrogel–treated groups were analyzed together for the remainder of the experiment.

exclusions due to surgeryA total of 5 animals from the retrograde tracing co-

hort were excluded from analysis. Two animals in the HA hydrogel–treated group died, and 1 animal in the control group was killed because of extensive chewing of a hind paw during the course of the study. Two other animals were excluded from the study after the complete follow-up

period: 1 in the sham group that showed signs of a vascu-lar lesion and 1 in the control group that seemed to have an incomplete microscopic hemisection. Therefore, the study included data from 18 rats: 2 in the sham group, 3 in the control group, and 13 in the HA hydrogel–treated group. A total of 5 animals in the histology cohort died. One ani-mal was in the control group, and the other 4 were in the HA hydrogel–treated group. Therefore, the study included histology data from 61 animals: 12 in the sham group, 11 in the control group, and 38 in the HA hydrogel–treated group.

Fig. 3. Scanning electron microscope images of HA hydrogels at con-centrations of 0.64% (a), 0.75% (b), 1% (c), and 1.25% (d) HA (w/v). Bar = 50 µm.

Fig. 4. Quantification of pore size in different HA weight-percent gels using SEM images.



retrograde tracingThe right red nucleus was contralateral and corre-

sponded to the left spinal hemisection. The total numbers of labeled cells in the red nuclei did not vary among the HA hydrogel–treated, control, and sham groups (p = 0.16); however, the numbers of cells on the right side were sig-nificantly lower than the numbers of cells on the left side in both the control group (mean 267 ± 108 cells on the left, 106 ± 125 cells on the right; p = 0.006) and the HA hydro-gel–treated group (mean 698 ± 259 cells on the left, 568 ± 250 cells on the right; p < 0.0001) (Fig. 5A). Furthermore, the ratio of number of cells on the right to the number of cells on the left (R/L) was significantly lower (p = 0.0005) in the control group (mean 0.32 ± 0.28) than in the HA hy-drogel–treated group (0.78 ± 0.12) (Fig. 5B). The improved ratio in the HA hydrogel–treated group indicated a greater percentage of functioning neurons in that group. The num-ber of labeled cells in the sham group was not significantly different between sides (p = 0.70), and the R/L ratio (mean 0.89 ± 0.04) was not significantly different from either the control group (p = 0.07) or the HA hydrogel–treated group (p = 0.33). These results indicate that there were more functional axons in the HA hydrogel–treated group than in the control group. In a fully regenerated spinal cord, the number or ratio of axons in the treated group would not be different from that in the sham surgery group; however, the sham group did not differ from the control group, so the significance of this result is unclear.

lesion sizeThe lesions had a significantly lower (p = 0.03) mean

volume in the HA hydrogel–treated group (1292 ± 334 mm3) than in the control group (2884 ± 1753 mm3) (Fig. 5C). The significant difference in lesion size might indi-cate either retention of axons or axon regeneration. Data from the sham group were not included in the analysis be-cause the lesion volume was 0.

behavioral testingAll the groups showed some recovery of motor func-

tion, assessed according to the BBB scale,5 during the 12-week follow-up period (Fig. 5D). Compared with the left side, the right side in both the control and HA hydro-gel–treated groups had decreased function (p = 0.002), but there was no significant difference between the control and HA hydrogel-treated groups (p = 0.95).

microscopic analysisThe H & E stain from the control samples revealed

dense scar tissue filling the lesions and disorganization of the tissue cranial to the lesion site (Fig. 6A). The HA hydrogel–treated group not only lacked scar tissue in the lesion but also retained more structural alignment above the lesion site (Fig. 6B). The HA hydrogel in the lesion site showed slight degradation 8 weeks after injection (Fig. 6B). The Nauta silver stain revealed more disrupted axons

Fig. 5. a: Mean number of cells in each side of the red nucleus labeled by the gold retrograde tracer. Asterisks indicate a signifi-cant difference (p < 0.05, post hoc t-test) from other groups within the bracket. b: Bar graph of the mean ratio of number of cells in the right red nucleus to number of cells in the left red nucleus labeled by the retrograde tracer. Asterisks indicate a p < 0.05 (1-way ANOVA). c: The mean lesion volume was significantly lower in the HA hydrogel–treated group than in the control group. Asterisks indicate a p < 0.05 (unpaired t-test). d: Mean BBB scores for the motor function assessment.



above the lesion site in the control group (Fig. 6C) than in the HA hydrogel–treated group (Fig. 6D). Similarly, the neurofilament stains in the controls (Fig. 6E) revealed a pronounced loss of axons cranial to the lesion site, which was absent in the HA hydrogel–treated group (Fig. 6F). The neurofilament stain adjacent to the lesion showed that axons in the control group were poorly aligned in the le-sion scar (Fig. 6G), whereas the HA hydrogel–treated group maintained axon alignment along the edge of the lesion (Fig. 6H). Together, these results show that there was less scar tissue and subsequently more axons in the HA hydrogel–treated group than in the control group, as evident by the neurofilament stain, retrograde tracing, and decreased mean lesion size.

Inflammation was present in both the HA hydrogel–treated and control groups, although it appeared qualita-tively better in the HA hydrogel–treated group. The Iba1 stain (Fig. 7A and B) revealed the presence of mononuclear cells in both the control and HA hydrogel–treated groups, respectively, although there were fewer mononuclear cells in the HA hydrogel–treated group. Reactive astrocytes were noted near the lesion site (Fig. 7C). Transmission electron microscopy confirmed the presence of inflam-matory cells, including both mononuclear macrophages, some of which had lysosomes filled with myelin debris, and multinucleated macrophages (Fig. 7D and E, respec-tively). Neither neutrophils nor necrosis was found with TEM in the HA hydrogel–treated group.

Transmission electron microscopy also showed periph-eral-type myelinated nerve fibers in the spinal cord tissue

above the injured area (Fig. 8). Many myelinated cells possessed features similar to myelinating Schwann cells with basal lamina, as opposed to central nervous system myelinating cells, oligodendrocytes, which do not have the basal lamina. These findings indicate that Schwann cells may migrate to the area of SCI to myelinate outgrowing axons. Schwann cells were not apparent in TEM images of the spinal cord 8 weeks after surgery, but they were abun-dant by 16 weeks after surgery.

discussionMultiple aspects of SCI have been studied in terms

of pathophysiology and treatment strategies. Despite the identification of multiple approaches for decreasing post-traumatic ischemia and inflammation, and for boosting the initial neurite outgrowth in the central nervous system, the prevailing cavitation limits the possibility of bridging the gap.16,43 Therefore, multiple scaffolding strategies have been explored. Autotransplantation of stem cells, Schwann cells, and olfactory ensheathing cells in the damaged spi-nal cord, at both short and long time periods after injury, and implantation of artificial materials have been per-formed with various results.6,8,15,24,32,34,35,37,42,51,52 Hyaluronic acid is especially interesting in light of the fact that not only is it a natural part of the extracellular matrix and is nonimmunogenic, but also it has been used commercially in the skin and joints and intraabdominally.26,31,38,49 It has also shown strong neurite-promoting characteristics in vi-tro.25 In vivo, HA has been shown to have neuroprotective

Fig. 6. Spinal cord injury sites in samples from the control and HA hydrogel–treated groups at 8 weeks show differences in immu-nohistochemistry. The injured side is on the left, and the uninjured side is on the right. The top of the image is the cranial direction. H & E stain in a control sample shows scar tissue (arrow) in the lesion site (a), whereas an HA hydrogel–treated group sample does not (b) (arrow indicates the edge of the lesion site, where little scar tissue is seen). Silver staining reveals denser staining of the injury side in a control sample (c) than in an HA hydrogel–treated group sample (d); arrows indicate dark stained disrupted neural tissue. Neurofilament stain in a control sample shows loss of axons cranial to the lesion site, as indicated by the lighter color (arrow) on the left side of the cord (e), and neurofilament stain in axons in an HA hydrogel–treated group sample shows intact axons (arrow) above the lesion on the injury side (F). Neurofilament stain of a control lesion site shows axons mixed with scar tissue (arrow) in the lesion (g), whereas an HA hydrogel–treated group sample retains axon alignment (arrow) at the edge of the lesion (h). Bar = 100 µm (in all panels).



and antiinflammatory functions.3,28 Khaing et al.28 found an antiinflammatory effect of an HA hydrogel in a hemi-section model, and Austin et al.3 found that a composite hydrogel that included HA had a neuroprotective effect in a crush-injury model.

In this study, HA hydrogels were placed in a hemi-sectioned spinal cord lesion to promote regeneration. Structural changes in the spinal cord were studied using histochemistry, functional changes were assessed using a retrograde tracer, and the effects of those changes on mo-tor function were observed. The hemisection model is not as rigorous as a full transection model; more often, human SCI involves some combination of contusion, crush, pinch, twist and torsion, partial tearing, and vascular injury. We chose a hemisection model because of the relatively con-trolled injury environment in which only part of the spinal cord was damaged. This model had the added benefit of making the animal model much easier to maintain than other models, which can include extensive morbidity, such as is seen in full transection or crush-injury models, specifically with bladder function and ability to feed and groom. Furthermore, the hemisection model had the ben-efit of easy implantation of the HA hydrogel; in contrast, in crush-injury models, the spinal cord is closed, when not torn by the impact, and the injection site is difficult to de-termine. In addition, crush injuries are highly variable in terms of the characteristics and features of injury. There-fore, the controlled injury environment of the hemisection model made the HA effect more easily assessed in a small number of animals than the crush-injury models, in which there is greater animal-to-animal variability. The choice

of a hemisection model limited the study in that hemi-section injuries are far less common than contusion inju-ries; however, the previous success of an HA crush-injury model that includes hydrogel3 makes it likely that the HA hydrogels are neuroprotective in multiple types of SCI. The lesions were significantly smaller in the HA hydro-gel–treated group than in the control group; in addition, there were noticeable changes between the HA hydrogel–treated and control groups in the quality of the lesions. The HA hydrogel–treated group had increased longitudi-nal alignment of the tissue above and adjacent to the lesion site. Furthermore, there was increased retention of intact axons immediately cranial to the lesion site and decreased mononuclear cells present in the treated groups. A major function of the neuroinflammatory pathway is to restore the blood–brain barrier.20,27 In this experiment, the blood–brain barrier was partially restored by the HA hydrogel as a result of simple mechanical blocking of exogenous inflammatory cells. The artificial barrier might have led to the decreased inflammatory response by the spinal cord and, subsequently, to decreased scar tissue formation in the spinal cord and lesion site. The decreased scar tissue correlated with the retention of functional axons adjacent to the lesion in the HA hydrogel–treated group, as indi-cated by the increased retrograde tracer uptake. There was no neural tissue visible in the hydrogel, indicating that the hydrogel was not functioning as a scaffold; therefore, it is probable that the increased function was a result of the neuroprotection.

There are multiple ways in which gel formation in situ could provide neuroprotection. Because the gel forms a

Fig. 7. Iba1 staining shows more mononuclear cells (arrow) present in a control group sample (a) than in an HA hydrogel–treated group sample (b), and glial fibrillary acidic protein staining reveals reactive astrocytes (arrows) in an HA hydrogel–treated group sample (c). Sixteen weeks after surgery, TEM shows a few macrophages above the traumatic area in HA hydrogel–treated group samples in which some macrophages contain lysosomes filled with myelin debris (arrows) (d) and some were multinucleated macrophages (arrow) (e). Bar = 50 µm (A–C) or 2 µm (D and E).



tight seal over the lesion site, one possible neuroprotec-tive mechanism is the mechanical blocking of exogenous inflammatory cells and fibroblasts. This mechanical blocking would prevent scar tissue from forming in the lesion site and decrease infiltration of cells into the spi-nal cord. HA itself has also been shown to have neuro-protective properties;2 therefore, the neuroprotective effect could be a result of molecular interaction of the HA with the surrounding cells. However, the gel might also have neuroprotective effects not directly tested in this experi-ment, such as prevention of hematoma formation and a corresponding decrease in toxic effect from hemoglobin degradation,41 prevention of cerebrospinal fluid leakage, and mechanical support of the spinal cord near the lesion.

The neuroprotective effects of the HA hydrogel will need to be tested for their time-dependent effect. In this study, the HA hydrogel was placed in the surgical site im-mediately after injury for ease of the surgical procedure; however, clinically, the hydrogel would not be placed until hours after an injury. The time delay could hinder the neu-roprotective effect of the HA hydrogel.

Although neuroprotection led to a decreased lesion volume and greater retention of functional axons, motor function was not significantly improved according to the BBB scale scores. In hemisection models, especially in rats, there is expected recovery due to collateral recruit-

ment.4,23 Therefore, with a larger sample size or a different animal model, an HA hydrogel–treated group could show some functional motor improvement compared to that of control animals.

One limitation of this study was the small number of animals in the behavioral testing group, with 18 animals included in total and only 2 animals in the sham group. Furthermore, no data on sensation or bowel and urinary function were collected, so it is unknown if there were nonmotor functional improvements. However, even if the HA hydrogel did not cause any motor or sensory reten-tion, it could serve an important function in future or con-current regenerative efforts. The retention of functioning axons near the edge of the lesion, decreased scar tissue, and decreased cavitation all serve to decrease the gap that regenerating nerves need to cross.

Even in its basic form, the HA hydrogel shows poten-tially useful neuroprotective effects, but this study could also be considered a basis for further development. For instance, results could be improved by adding growth-promoting signaling factors (i.e., neural growth factor), growth-inhibitory signaling suppressors (i.e., neurite out-growth inhibitor 66 receptor antibody), cell-attachment moieties, and various types of progenitor cells. Much of this work has already been performed or is underway by other research groups;22,47,48,50,51 however, the work de-

Fig. 8. a: Sixteen weeks after injury, TEM reveals myelinated nerve fibers near the HA hydrogel. b and c: On closer examina-tion, bundles of collagen (arrow) are visible around the myelinated nerve fibers, which indicates a basal lamina. d: Schwann cells (boxes) are seen in multiple other images as well. Bar = 10 µm (A and D) or 2 µm (B and C).



scribed here highlights the importance of base material choices in the development of complex scaffolding sys-tems.

conclusionsHyaluronic acid hydrogel protected the spinal cord from

inflammation and reduced secondary injury in an animal model. The HA hydrogel–treated group had decreased le-sion sizes and significant retrograde uptake of the tracer. Although these results could indicate regeneration, they more likely indicate decreased scar tissue formation. The immunohistochemistry results showing decreased inflam-matory cell presence, decreased scarring, and increased axon retention in the HA hydrogel–treated group support this theory. Although results of the behavior assessment in this experiment did not indicate functional improvement in the treated group, the decrease in lesion size and the re-tention of structural alignment of the spinal cord near the lesion could be important for future regenerative efforts.

acknowledgmentsThis work was supported by a grant from the Arizona Bio-

medical Research Commission as well as funds from the Newsome Family Chair Endowment for Neurosurgery Research, the Beverly and Millard Seldin Family Disc and Spinal Regeneration Fund, and the Barrow Neurological Foundation. The authors gratefully acknowledge the use of facilities within the Center for Solid State Science at Arizona State University.

references 1. Alini M, Li W, Markovic P, Aebi M, Spiro RC, Roughley

PJ: The potential and limitations of a cell-seeded collagen/hyaluronan scaffold to engineer an intervertebral disc-like matrix. Spine (Phila Pa 1076) 28:446–454, 2003

2. Austin JW, Gilchrist C, Fehlings MG: High molecular weight hyaluronan reduces lipopolysaccharide mediated microglial activation. J Neurochem 122:344–355, 2012

3. Austin JW, Kang CE, Baumann MD, DiDiodato L, Sat-kunendrarajah K, Wilson JR, et al: The effects of intrathecal injection of a hyaluronan-based hydrogel on inflammation, scarring and neurobehavioural outcomes in a rat model of severe spinal cord injury associated with arachnoiditis. Bio-materials 33:4555–4564, 2012

4. Ballermann M, Fouad K: Spontaneous locomotor recovery in spinal cord injured rats is accompanied by anatomical plas-ticity of reticulospinal fibers. Eur J Neurosci 23:1988–1996, 2006

5. Basso DM, Beattie MS, Bresnahan JC: A sensitive and reli-able locomotor rating scale for open field testing in rats. J Neurotrauma 12:1–21, 1995

6. Baumann MD, Kang CE, Tator CH, Shoichet MS: Intrathecal delivery of a polymeric nanocomposite hydrogel after spinal cord injury. Biomaterials 31:7631–7639, 2010

7. Blight AR: Cellular morphology of chronic spinal cord injury in the cat: analysis of myelinated axons by line-sampling. Neuroscience 10:521–543, 1983

8. Boido M, Garbossa D, Fontanella M, Ducati A, Vercelli A: Mesenchymal stem cell transplantation reduces glial cyst and improves functional outcome after spinal cord compression. World Neurosurg 81:183–190, 2014

9. Brown BN, Barnes CA, Kasick RT, Michel R, Gilbert TW, Beer-Stolz D, et al: Surface characterization of extracellular matrix scaffolds. Biomaterials 31:428–437, 2010

10. Bryant SJ, Cuy JL, Hauch KD, Ratner BD: Photo-patterning

of porous hydrogels for tissue engineering. Biomaterials 28:2978–2986, 2007

11. Caicco MJ, Zahir T, Mothe AJ, Ballios BG, Kihm AJ, Tator CH, et al: Characterization of hyaluronan-methylcellulose hy-drogels for cell delivery to the injured spinal cord. J Biomed Mater Res A 101:1472–1477, 2013

12. Choo AM, Liu J, Dvorak M, Tetzlaff W, Oxland TR: Second-ary pathology following contusion, dislocation, and distrac-tion spinal cord injuries. Exp Neurol 212:490–506, 2008

13. Collins MN, Birkinshaw C: Hyaluronic acid based scaf-folds for tissue engineering—a review. Carbohydr Polym 92:1262–1279, 2013

14. Donnelly DJ, Popovich PG: Inflammation and its role in neuroprotection, axonal regeneration and functional recovery after spinal cord injury. Exp Neurol 209:378–388, 2008

15. Du BL, Zeng CG, Zhang W, Quan DP, Ling EA, Zeng YS: A comparative study of gelatin sponge scaffolds and PLGA scaffolds transplanted to completely transected spinal cord of rat. J Biomed Mater Res A 102:1715–1725, 2014

16. Ellis-Behnke RG, Liang YX, You SW, Tay DK, Zhang S, So KF, et al: Nano neuro knitting: peptide nanofiber scaffold for brain repair and axon regeneration with functional return of vision. Proc Natl Acad Sci U S A 103:5054–5059, 2006

17. Eng DH: Hyaluronan Hydrogels for Spinal Cord Regen-eration [dissertation]. Phoenix: Arizona State University, 2009

18. Eng D, Caplan M, Preul M, Panitch A: Hyaluronan scaffolds: a balance between backbone functionalization and bioactiv-ity. Acta Biomater 6:2407–2414, 2010

19. Fakhari A, Berkland C: Applications and emerging trends of hyaluronic acid in tissue engineering, as a dermal filler and in osteoarthritis treatment. Acta Biomater 9:7081–7092, 2013

20. Fitch MT, Silver J: CNS injury, glial scars, and inflammation: Inhibitory extracellular matrices and regeneration failure. Exp Neurol 209:294–301, 2008

21. Fraser JR, Laurent TC, Laurent UB: Hyaluronan: its nature, distribution, functions and turnover. J Intern Med 242:27–33, 1997

22. Führmann T, Obermeyer J, Tator CH, Shoichet MS: Click-crosslinked injectable hyaluronic acid hydrogel is safe and biocompatible in the intrathecal space for ultimate use in regenerative strategies of the injured spinal cord. Methods 84:60–69, 2015

23. Gulino R, Dimartino M, Casabona A, Lombardo SA, Percia-valle V: Synaptic plasticity modulates the spontaneous recov-ery of locomotion after spinal cord hemisection. Neurosci Res 57:148–156, 2007

24. Haggerty AE, Oudega M: Biomaterials for spinal cord repair. Neurosci Bull 29:445–459, 2013

25. Horn EM, Beaumont M, Shu XZ, Harvey A, Prestwich GD, Horn KM, et al: Influence of cross-linked hyaluronic acid hydrogels on neurite outgrowth and recovery from spinal cord injury. J Neurosurg Spine 6:133–140, 2007

26. Iannitti T, Morales-Medina JC, Coacci A, Palmieri B: Ex-perimental and clinical efficacy of two hyaluronic acid-based compounds of different cross-linkage and composition in the rejuvenation of the skin. Pharm Res [epub ahead of print], 2014

27. Kawano H, Kimura-Kuroda J, Komuta Y, Yoshioka N, Li HP, Kawamura K, et al: Role of the lesion scar in the response to damage and repair of the central nervous system. Cell Tissue Res 349:169–180, 2012

28. Khaing ZZ, Milman BD, Vanscoy JE, Seidlits SK, Grill RJ, Schmidt CE: High molecular weight hyaluronic acid limits astrocyte activation and scar formation after spinal cord in-jury. J Neural Eng 8:046033, 2011

29. Koyama Y: Signaling molecules regulating phenotypic con-versions of astrocytes and glial scar formation in damaged nerve tissues. Neurochem Int 78:35–42, 2014



30. Levi R, Hultling C, Nash MS, Seiger A: The Stockholm spi-nal cord injury study: 1. Medical problems in a regional SCI population. Paraplegia 33:308–315, 1995

31. Li C, Long X, Deng M, Li J, Cai H, Meng Q: Osteoarthritic changes after superior and inferior joint space injection of hyaluronic acid for the treatment of temporomandibular joint osteoarthritis with anterior disc displacement without reduc-tion: a cone-beam computed tomographic evaluation. J Oral Maxillofac Surg 73:232–244, 2015

32. Madigan NN, Chen BK, Knight AM, Rooney GE, Sweeney E, Kinnavane L, et al: Comparison of cellular architecture, axonal growth, and blood vessel formation through cell-load-ed polymer scaffolds in the transected rat spinal cord. Tissue Eng Part A 20:2985–2997, 2014

33. McDonald JW, Sadowsky C: Spinal-cord injury. Lancet 359:417–425, 2002

34. Mortazavi MM, Verma K, Tubbs RS, Theodore N: Cellular and paracellular transplants for spinal cord injury: a review of the literature. Childs Nerv Syst 27:237–243, 2011

35. Mothe AJ, Tam RY, Zahir T, Tator CH, Shoichet MS: Repair of the injured spinal cord by transplantation of neural stem cells in a hyaluronan-based hydrogel. Biomaterials 34:3775–3783, 2013

36. Nagoshi N, Fehlings MG: Investigational drugs for the treat-ment of spinal cord injury: review of preclinical studies and evaluation of clinical trials from Phase I to II. Expert Opin Investig Drugs 24:645–658, 2015

37. Park J, Lim E, Back S, Na H, Park Y, Sun K: Nerve regenera-tion following spinal cord injury using matrix metallopro-teinase-sensitive, hyaluronic acid-based biomimetic hydrogel scaffold containing brain-derived neurotrophic factor. J Biomed Mater Res A 93:1091–1099, 2010

38. Prestwich GD, Marecak DM, Marecek JF, Vercruysse KP, Ziebell MR: Controlled chemical modification of hyaluronic acid: synthesis, applications, and biodegradation of hydrazide derivatives. J Control Release 53:93–103, 1998

39. Prinjha R, Moore SE, Vinson M, Blake S, Morrow R, Chris-tie G, et al: Inhibitor of neurite outgrowth in humans. Nature 403:383–384, 2000

40. Raposo C, Schwartz M: Glial scar and immune cell involve-ment in tissue remodeling and repair following acute CNS injuries. Glia 62:1895–1904, 2014

41. Regan RF, Guo Y: Toxic effect of hemoglobin on spinal cord neurons in culture. J Neurotrauma 15:645–653, 1998

42. Reginensi D, Carulla P, Nocentini S, Seira O, Serra-Picamal X, Torres-Espín A, et al: Increased migration of olfactory ensheathing cells secreting the Nogo receptor ectodomain over inhibitory substrates and lesioned spinal cord. Cell Mol Life Sci 72:2719–2737, 2015

43. Rooney GE, Endo T, Ameenuddin S, Chen B, Vaishya S, Gross L, et al: Importance of the vasculature in cyst forma-tion after spinal cord injury. J Neurosurg Spine 11:432–437, 2009

44. Shu XZ, Liu Y, Luo Y, Roberts MC, Prestwich GD: Disulfide cross-linked hyaluronan hydrogels. Biomacromolecules 3:1304–1311, 2002

45. Silva NA, Sousa N, Reis RL, Salgado AJ: From basics to clinical: a comprehensive review on spinal cord injury. Prog Neurobiol 114:25–57, 2014

46. Silver J, Miller JH: Regeneration beyond the glial scar. Nat Rev Neurosci 5:146–156, 2004

47. Suri S, Schmidt CE: Cell-laden hydrogel constructs of hyal-uronic acid, collagen, and laminin for neural tissue engineer-ing. Tissue Eng Part A 16:1703–1716, 2010

48. Switzer RC III: Application of silver degeneration stains for neurotoxicity testing. Toxicol Pathol 28:70–83, 2000

49. ten Broek RPG, Stommel MWJ, Strik C, van Laarhoven CJHM, Keus F, van Goor H: Benefits and harms of adhe-sion barriers for abdominal surgery: a systematic review and meta-analysis. Lancet 383:48–59, 2014

50. Wang X, He J, Wang Y, Cui FZ: Hyaluronic acid-based scaf-fold for central neural tissue engineering. Interface Focus 2:278–291, 2012

51. Wei YT, He Y, Xu CL, Wang Y, Liu BF, Wang XM, et al: Hyaluronic acid hydrogel modified with nogo-66 receptor antibody and poly-L-lysine to promote axon regrowth after spinal cord injury. J Biomed Mater Res B Appl Biomater 95:110–117, 2010

52. Williams RR, Henao M, Pearse DD, Bunge MB: Permissive Schwann cell graft/spinal cord interfaces for axon regenera-tion. Cell Transplant 24:115–131, 2015

53. Yuan YM, He C: The glial scar in spinal cord injury and re-pair. Neurosci Bull 29:421–435, 2013

54. Zheng Shu X, Liu Y, Palumbo FS, Luo Y, Prestwich GD: In situ crosslinkable hyaluronan hydrogels for tissue engineer-ing. Biomaterials 25:1339–1348, 2004

disclosuresThe authors report no conflict of interest concerning the materi-als or methods used in this study or the findings specified in this paper.

author contributionsConception and design: Preul, Kushchayev, Hom Eng, Mar-tirosyan, Mortazavi, Theodore, Panitch. Acquisition of data: Kushchayev, Hom Eng, Martirosyan, Mortazavi. Analysis and interpretation of data: all authors. Drafting the article: Preul, Kushchayev, Giers, Hom Eng, Martirosyan, Mortazavi, Panitch. Critically revising the article: Preul, Kushchayev, Giers, Esch-bacher, Mortazavi. Reviewed submitted version of manuscript: all authors. Statistical analysis: Giers. Administrative/technical/mate-rial support: Preul, Panitch. Study supervision: Preul, Panitch.

supplemental informationPrevious PresentationsPortions of this work were presented at annual meetings of the American Association of Neurological Surgeons, San Diego, CA, May 6, 2009, and Washington, DC, May 4, 2015. Portions of this work were also published in 2009 in the doctoral dissertation “Hyaluronan hydrogels for spinal cord regeneration” by Doris Hom Eng, Arizona State University.

Current AffiliationsDr. Kushchayev: Department of Radiology, Mercy Catholic Med-ical Center, Philadelphia, PA; Dr. Hom Eng: Ventana Medical Systems, Inc., Tucson, AZ; Dr. Mortazavi: California Neurologi-cal Institute, Los Robles Hospital and Medical Center, Thousand Oaks, CA.

correspondenceMark C. Preul, c/o Neuroscience Publications, Barrow Neurologi-cal Institute, St. Joseph’s Hospital and Medical Center, 350 W. Thomas Rd., Phoenix, AZ 85013. email: [email protected].

hyaluronic acid scaffold has a neuroprotective effect in ... · hyaluronic acid (ha) is a linear...

Documents