SEARCH

SEARCH BY CITATION

Keywords:

  • Spinal cord injury;
  • Stem cell transplantation;
  • Neural stem cell;
  • Gene expression

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

The transplantation of neural stem/precursor cells (NSPCs) is a promising therapeutic strategy for many neurodegenerative disorders including spinal cord injury (SCI) because it provides for neural replacement or trophic support. This strategy is now being extended to the treatment of chronic SCI patients. However, understanding of biological properties of chronically transplanted NSPCs and their surrounding environments is limited. Here, we performed temporal analysis of injured spinal cords and demonstrated their multiphasic cellular and molecular responses. In particular, chronically injured spinal cords were growth factor-enriched environments, whereas acutely injured spinal cords were enriched by neurotrophic and inflammatory factors. To determine how these environmental differences affect engrafted cells, NSPCs transplanted into acutely, subacutely, and chronically injured spinal cords were selectively isolated by flow cytometry, and their whole transcriptomes were compared by RNA sequencing. This analysis revealed that NSPCs produced many regenerative/neurotrophic molecules irrespective of transplantation timing, and these activities were prominent in chronically transplanted NSPCs. Furthermore, chronically injured spinal cords permitted engrafted NSPCs to differentiate into neurons/oligodendrocytes and provided more neurogenic environment for NSPCs than other environments. Despite these results demonstrate that transplanted NSPCs have adequate capacity in generating neurons/oligodendrocytes and producing therapeutic molecules in chronic SCI microenvironments, they did not improve locomotor function. Our results indicate that failure in chronic transplantation is not due to the lack of therapeutic activities of engrafted NSPCs but the refractory state of chronically injured spinal cords. Environmental modulation, rather modification of transplanting cells, will be significant for successful translation of stem cell-based therapies into chronic SCI patients. STEM Cells 2013;31:1535–1547


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Due to the limited natural potential for repair and regeneration in the adult central nervous system (CNS), spinal cord injury (SCI) causes permanent severe motor/sensory dysfunction and thus significantly affects quality of life [1, 2]. Recent advances in acute management have drastically reduced mortality rates and contributed to an increased lifespan for SCI patients; at the same time, however, this progress in medicine has led to increased morbidity in patients who survive after SCI, because effective treatments have not been established. There is, therefore, a great demand for therapeutic developments for chronic SCI in clinical practice [3].

Stem cell-based therapies including neural stem/precursor cells (NSPCs) had shed light on the treatment for SCI [4–6]. NSPCs are somatic stem cells of CNS and characterized by the capacity for self-renewal and multilineage differentiation potential [7]. These cells can produce functional neural cells after transplantation [4, 8], demonstrating tremendous promise for stem cell-based therapies. In addition to neural replacement, NSPCs promote neurological recovery by providing trophic support and modifying the host environment to create a permissive environment for endogenous regeneration and repair [4, 5, 9]. Based on considerable experimental evidences, clinical trials for SCI have been performed or initiated [5, 6, 10], and stem cell-based therapies using NSPCs are now being translated to patients with chronic SCI [6, 11, 12]. However, while emerging evidence supports the therapeutic potential of NSPC transplantation for the acute and subacute phases of SCI [4, 5, 13], few studies have investigated NSPC transplantation for the chronic phase of SCI [14] and its therapeutic efficacy for chronic SCI thus remains controversial [3, 14, 15].

We recently reported that NSPCs are highly vulnerable to environmental factors and that the graft environment greatly affects engrafted NSPCs [16]. The dynamic change in the microenvironment of injured spinal cords with time [2] could have a considerable effect on the cellular properties of NSPCs transplanted at different phases after SCI. For the successful application of stem cell-based therapies in chronic SCI, it is crucial to have a more detailed understanding of the in vivo behavior of NSPCs transplanted into chronically injured spinal cords.

One of the best way of elucidating cellular activities is to reveal the transcriptome, which is the complete set of transcripts in a cell [17]. For transcriptome analysis, expression microarrays have been most widely used. However, these hybridization-based methods have several limitations, including the restriction of gene expression measurement using designed probes to only a few regions of the gene; the narrow dynamic range of detection, because of both high background and saturation of fluorescence intensity and the difficulty in comparing expression levels across different experiments owing to semiquantitative issues [17–19]. Ultra high-throughput RNA sequencing (RNA-Seq) is a recently developed evolutional technology which overcomes these limitations of microarrays by providing both single-base resolution for annotation and absolute gene expression levels [17, 18]. In particular, one of great advantages of RNA-Seq is their capability to capture even low expressed transcripts, including differentiation-associated genes and secretory molecules, in a digital manner [17]. Due to these advantages, RNA-Seq is particularly useful for elucidating cellular activities such as differentiation and secretion.

In this study, NSPCs were transplanted at 3 months after incomplete contusion injury as chronic phase of SCI [14]. To identify the cellular and molecular properties of NSPCs transplanted into chronically injured spinal cords, we combined flow-cytometric isolation and RNA-Seq. We found that NSPCs transplanted into chronically injured spinal cords have a sufficient capacity to produce therapeutic molecules and differentiate into neurons/oligodendrocytes [14]. However, chronic transplantation of NSPCs did not improve locomotor function after SCI. These results indicate that failure of chronic transplantation in SCI is not due to the cellular therapeutic activities of engrafted NSPCs, but due to the refractory state of chronically injured spinal cords. Environmental modulation, rather modification of transplanting cells, will be significant for successful translation of stem cell-based therapies into chronic SCI patients. Our findings and database will provide a conceptual framework for the design of strategies to treat chronic SCI patients.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Mice

Female C57BL/6J mice (8–10 weeks old) were used in this study. All mice were housed in a temperature- and humidity-controlled environment with a 12 hour light-dark cycle. All surgical procedures and experimental manipulations were approved by the Committee on Ethics in Animal Experiments in the Faculty of Medicine, Kyushu University (A21-056-0). Experiments were conducted in accordance with the guidelines for animal experiments.

Spinal Cord Injury

In brief, mice were anesthetized by an intraperitoneal injection of sodium pentobarbital (Somnopentyl; 50 mg/kg). Moderate contusion SCI was induced using an Infinite Horizons Impactor Precision Systems Instrumentation, LLC, Lexington, KY, http://www.presysin.com/ at 70 kdyn at the 10th thoracic level (Th10) [16]. This is experimental model for nonambulatory patients after incomplete SCI.

Flow Cytometry

After SCI, animals were reanesthetized and perfused with phosphate-buffered saline (PBS). The spinal cord (4.0 mm, centered around Th10) was dissected from the vertebral column and then prepared for flow cytometry [16]. Briefly, spinal cord samples were dissociated in collagenase type I (175 U/mL; Invitrogen, Carlsbad, CA, http://www.invitrogen.com) for 15 minutes at 37°C, and the cells were washed with Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. After centrifugation, the resulting suspension was pelleted by centrifugation at 3,000 rpm for 5 minutes at 4°C, and then incubated on ice with fluorescent antibodies for 30 minutes. All antibodies used are described in the supporting information Materials and Methods. The single cell suspension was collected and resuspended in 500 μL of fluorescence-activated cell sorting (FACS) buffer (Hanks balanced salt solution, 2.5% fetal calf serum, 0.1% NaN3) and then analyzed using a FACSAria II flow cytometer (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com) and FACSDiva software (BD Biosciences). Each population of inflammatory cells was analyzed by flow cytometry from a pool of CD45posi inflammatory cells obtained from the injured spinal cords. Neutrophils were defined as CD11bhigh/Gr-1posi cells, monocytes/macrophages were defined as CD45high/CD11bposi/Gr-1nega-int cells, and microglia were defined as CD45int/CD11bposi/Gr-1nega-int cells.

Quantitative Reverse Transcription-Polymerase Chain Reaction

Total RNA was isolated from the spinal cord (4 mm long) using an RNeasy Lipid Tissue kit (Qiagen, Hilden, Germany, http://www1.qiagen.com) [20]. For cDNA synthesis, the reverse transcriptase reaction was carried out using a PrimeScript first strand cDNA synthesis kit (TaKaRa Bio., Shiga, Japan, http://www.takara.co.jp). Quantitative polymerase chain reaction (qPCR) was performed using primers specific to the genes of interest (supporting information Table 1) and SYBR Premix Dimmer Eraser (TaKaRa Bio.). The data were normalized to glyceraldehydes-3-phosphate dehydrogenase [16].

NSPC Primary Culture and Lentiviral Transduction

The striatum of C57BL/6J mice on embryonic day 14 was dissociated, and concentrated lentiviruses were added to the culture medium [16]. Lentiviral vector expressing both luciferase and green fluorescent protein (GFP) was generated in 293T cells by transient transfection of the transfer vectors pCSII-EF-Luc-IRES2-Venus, pLP1, pLP2, and pLP/VSVG (Invitrogen) using GeneJuice Transfection Reagent (Novagen), followed by ultracentrifugation and titration [16]. After purification of GFP-positive transfected cells and passaging twice, neurospheres were used for transplantation. For RNA-Seq experiments, NSPCs were harvested from the embryonic striatum of commercially available green mice (purchased from Japan SLC Inc., Hamamatsu, Shizuoka, Japan) and expanded in the same manner.

Cell Transplantation

Using a stereotaxic injector (KDS 310, Muromachi-kikai, Tokyo, Japan), a 25 μL Hamilton syringe (Hamilton, Reno, NV) with a glass tip was inserted into injured or naïve spinal cords and a total volume of 2 μL of cell suspension (2.5 × 105 viable cells per microliter) was injected at a rate of 0.5 μL/minute [16]. Two intraspinal injections were made bilaterally at a point 1 mm rostral and 1 mm caudal to the injury site. In chronic transplantation experiments, an equal volume of medium was injected in the control group.

Bioluminescence Imaging

Bioluminescence imaging was performed using a Xenogen-IVIS 50 cooled CCD optical macroscopic imaging system (SC BioScience, Tokyo, Japan) [16]. Mice were given an intraperitoneal injection of d-luciferin (150 mg/kg body weight), and serial images were acquired from 15 to 40 minutes after the administration until the maximum intensity was obtained with the field-of-view set at 7.2 cm. All images were analyzed with Igor (WaveMetrics, Lake Oswego, OR) and Living Image software (Xenogen, Alameda, CA), and optical signal intensity was expressed as photon flux, in units of photons/second per cm2 per steradian.

Immunohistochemistry

Animals were reanesthetized and transcardially perfused with 4% paraformaldehyde in PBS. The spinal cord was then removed and immersed in the same fixative for 24 hours at 4°C. The spinal cord samples were immersed in 10% sucrose in PBS for 24 hours and 30% sucrose in PBS for another 24 hours [16]. Frozen sections of spinal cord were cut on a cryostat in the sagittal plane at 14 μm. For immunostaining, spinal cord sections were permeabilized and then primary antibodies were applied to the sections at 4°C overnight, followed by incubation with secondary antibodies. All antibodies used are described in supporting information Materials and Methods. For quantification of GFP and Hu or GSTπ-positive cells, one series of sagittal sections containing seven representative levels, each separated by 140 mm, was analyzed. After visual identification, GFP and Hu or GSTπ-double-positive cells were manually counted.

Isolation of Transplanted NSPCs

For the isolation of engrafted NSPCs, dissected spinal cord samples (4 mm long) were minced with scissors, then digested in 0.5% Trypsin with 0.5 mg/ml DNase for 5 minutes at 37°C [16]. The digests were further dissociated by gentle pipetting with a plastic pipette. The cell suspension was washed with DMEM and filtered through a 40-μm nylon cell strainer. The single-cell suspension was collected by centrifugation and resuspended in FACS buffer, and propidium Iodide was added to determine cell viability and then analyzed with FACSAria II.

RNA-Seq for Transplanted NSPCs

RNA was extracted from FACS-purified NSPCs using the RNeasy Micro Kit (Qiagen) following manufacturer's instructions [16]. RNA-Seq library preparation methods are described in supporting information Materials and Methods. All libraries were sequenced as 36-base-length reads using the Genome Analyzer (GAIIX; Illumina). Mapping of fragments to the genome, transcript assembly, abundance estimation, and data analysis are described in supporting information Materials and Methods. All sequencing data from this study are available at the DNA Data Bank of Japan Sequence Read Archive under submission ID DRA000584.

Behavioral Analysis

Motor function was evaluated with a locomotor open-field rating scale on the Basso Mouse Scale (BMS). A team of two experienced examiners evaluated each animal for 4 minutes and assigned an operationally defined score for each hind limb [20]. For the grip walk test [21], a grid walk was constructed for mice using two parallel pieces of wood (1 m long) to hold 100 rungs (round wooden steps; 2 mm in diameter, 10 cm in length) spaced 1 cm apart. After 3 days of training, each mouse was allowed to cross the grid walk three consecutive times and the number of grips was counted. A grip was defined as placing the toes on the rung while pushing the hind limb off to move to the next step. The mice were evaluated using 50 cm of the grid with three patterns: easy (50 steps, 1 cm apart), medium (every third step was removed), and hard (every other step was removed). The sum of the number of the grips for all three patterns was used in the grip walk analysis. For footprint analysis, the fore limbs and hind limbs of the mice were dipped in red and green dyes, respectively [20]. Footprints were made on a paper-covered narrow runway (80 cm length and 4 cm width) as the animals walked across. The stride length was defined as the distance from the start of a step with the rear paw to the end of that step with the same paw. The stride width was defined as the distance from the left outermost hind paw digit to the right outermost hind paw digit. Paw rotation was defined as the angle between the axis of the rear paws and the midline axis of the body. All measurements were taken on each side for three consecutive steps and averaged. Each test was double-blinded.

Statistical Analysis

Statistical analysis was performed with a two-tailed unpaired or paired Student's t test or the Mann–Whitney U test for single comparisons, and one-way factorial ANOVA or two-way repeated measures ANOVA followed by the post hoc Bonferroni or Dunnett's correction for multiple comparisons. Correlation analysis was performed with the Pearson correlation coefficient. In all statistical analyses, significance was accepted at p < .05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Cellular and Molecular Dynamics of Traumatic Injured Spinal Cords

The success of stem cell-based strategies is strongly influenced by the timing of transplantation following SCI [14, 15], which may be attributable to the changing environment of the injured spinal cord and the adaptation capacity of engrafted cells. Many researchers focused on elucidating pathophysiological processes during the acute and subacute phases of SCI [11, 14], but the microenvironment of chronic SCI (more than 42 days after SCI) is still poorly understood. Therefore, we first examined the time course of the change in the number of inflammatory cells until 3 months after SCI using flow cytometry. Consistent with previous reports [16, 20], neutrophil infiltration peaked 12 hours after SCI and decreased immediately thereafter (Fig. 1A). Unlike the neutrophil infiltration pattern, the number of microglia gradually increased and peaked 6 weeks after SCI (Fig. 1A). Macrophages/monocytes showed a biphasic pattern with a first peak at 12 hours and a second peak 6 weeks after SCI (Fig. 1A). Three months after SCI, large numbers of macrophages/monocytes and microglia, but few neutrophils, were detected. These results demonstrate a biphasic cellular inflammatory response in the acute and chronic phases of SCI.

thumbnail image

Figure 1. Time-dependent changes of injured spinal cords. (A): Time course of the number of infiltrated inflammatory cells in injured spinal cords by flow cytometry (n = 7 per group). (B–H): Time course of gene expression in injured spinal cords. Gene expression levels of proinflammatory cytokines (B), anti-inflammatory cytokines (C), CXCL chemokines (D), CCL chemokines (E), and growth factors and neurotrophic factors (F–H) were determined by quantitative polymerase chain reaction (n = 6–9 per group). Gene expression levels were normalized to naïve spinal cords. Data are represented as the mean ± SEM. Abbreviations: BMP, bone morphogenetic protein; CXCL, CXC chemokine ligand; CCL, CC chemokine ligand; CNTF, ciliary neurotrophic factor; EGF, epidermal growth factor; FGF, fibroblast growth factor; GDNF, glial cell-derived neurotrophic factor; HGF, hepatocyte growth factor; HDGF, hepatoma-derived growth factor; IL, interleukin; IGF-1, insulin growth factor-1; LIF, leukemia inhibitory factor; NGF, nerve growth factor; NT-3, neurotrophin 3; PDGF, platelet derived growth factor; TNF-α, tumor necrosis factor α; TGF, transforming growth factor; VEGF, vascular endothelial growth factor.

Download figure to PowerPoint

Next, to elucidate the long-term molecular changes occurring after SCI, we examined the gene expression of extracellular molecules in injured spinal cords during 3 months after SCI. qPCR analyses revealed that interleukin (IL)-6 and IL-1β gene expression dramatically increased during the acute phase of SCI and immediately decreased to the normal range (Fig. 1B). Tumor necrosis factor-α gene expression was upregulated at the acute and chronic phase of SCI (Fig. 1B). Regarding anti-inflammatory cytokines, IL-10 gene expression was upregulated only during the acute phase of SCI, whereas IL-4 gene expression was gradually upregulated from the subacute phase of SCI (Fig. 1C). Transforming growth factor (TGF)-β1 gene expression exhibited a biphasic pattern of upregulation in the acute and chronic phases of SCI (Fig. 1C). Expression of all CXC chemokine ligand and CC chemokine ligand (CCL) genes was strongly upregulated during the acute phase of SCI, and gene expression of CCL3 and CCL5 was also elevated in the chronic phase of SCI (Fig. 1D, 1E). Regarding growth and neurotrophic factors, expression of the genes for nerve growth factor (NGF), glial cell-derived neurotrophic factor, brain-derived neurotrophic factor (BDNF), hepatoma-derived growth factor, vascular endothelial growth factor (VEGF)-α, VEGF-β, bone morphogenetic protein (BMP)-2, and leukemia inhibitory factor (LIF) was transiently upregulated in the acute phase of SCI (Fig. 1F). In, contrast, gene expression of insulin-like growth factor (IGF)-1, hepatocyte growth factor, fibroblast growth factor (FGF)-2, ciliary neurotrophic factor, platelet-derived growth factor (PDGF)-β, neurotrophin-3, and BMP-4 gradually increased from the subacute phase of SCI and was secularly upregulated during the chronic phase of SCI (Fig. 1G). Gene expression of PDGF-α and epidermal growth factor increased continuously during the subacute and chronic phases of SCI (Fig. 1H). These results reveal that chronically injured spinal cords provide a growth factor-enriched environment, whereas acutely injured spinal cords provide an environment enriched in inflammatory cytokine/chemokines and neurotrophic factors, indicating that the surrounding microenvironment of transplanted cells varies greatly during the different phases of SCI, particularly in the acute and chronic phases of injury.

Engraftment of NSPCs Transplanted into Chronically Injured Spinal Cords

Although transplanted NSPCs have been demonstrated to survive and migrate when they were transplanted into acutely or subacutely injured spinal cords [14, 22], little is known about these information for chronic SCI. To investigate their survival in chronically injured spinal cords, NSPCs lentivirally labeled with both luciferase and GFP reporter genes were transplanted 3 months after SCI. Bioluminescence imaging analysis, which detects luciferase photon signals only from living cells [22], demonstrated long-term cell viability and revealed an approximate graft survival rate of 16.9% ± 6.9% at 42 days after transplantation (DAT; Fig. 2A, 2B). At 42 DAT, engrafted NSPCs occupied the entire spared rim around the lesion center and had migrated extensively as far as 4 mm rostral or caudal to the graft site (Fig. 2C), and most engrafted cells had extended fine cellular processes (Fig. 2D). These results demonstrated the successful survival and migration of NSPCs transplanted at 3 months after SCI.

thumbnail image

Figure 2. Engraftment and migration of neural stem/precursor cells (NSPCs) transplanted into chronically injured spinal cords. (A): Bioimaging images of a representative mouse transplanted with NSPCs 3 months after spinal cord injury (SCI). (B): Time course of bioluminescence intensity relative to the initial value over a period of 42 DAT (n = 7 per group). (C): Images of representative sagittal sections of chronically injured spinal cords transplanted with NSPCs 3 months after SCI showing the great extent of rostral and caudal migration of NSPCs from the injection sites. Asterisk indicates the lesion center. Scale bar = 500 μm. (D): Magnification of the boxed area in (C). Scale bar = 50 μm. Data are represented as the mean ± SEM. Abbreviations: DAT, days after transplantation; GFP, green fluorescent protein.

Download figure to PowerPoint

Chondroitin sulfate proteoglycans (CSPGs) have recently been reported to be a major cause of failure during the chronic phase of transplantation [12] and, therefore, CSPGs may decrease at 3 months after SCI. To confirm this hypothesis, we assessed the temporal and spatial patterns of CSPGs after SCI using qPCR and immunohistochemistry. CSPGs were widely distributed around the lesion site between 2 and 6 weeks after SCI (supporting information Fig. 1A, 1B), and the gene expression of each CSPG peaked at 1 or 6 weeks after SCI (supporting information Fig. 1C). At 3 months after SCI, all types of CSPG expression returned to a near normal range (supporting information Fig. 1C). Most CSPGs were limited to the glial scar (supporting information Fig. 1A), whereas a few CSPGs were observed around the lesion area (supporting information Fig. 1B). After chronic transplantation at 3 months after SCI, no NSPCs were double-positive for CSPGs at 42 DAT, but they were widely distributed in the injured spinal cord, except for the lesion area (Fig. 2C; supporting information Fig. 1D). Because CSPGs are potent inhibitors of transplanted cell migration and survival, the reduction in CSPGs would permit the survival and migration of NSPCs transplanted at 3 months after SCI.

RNA-Seq for NSPCs Transplanted at Different Phases of SCI

Because the microenvironments of injured spinal cords have distinct pathological features among the acute, subacute, and chronic periods (Fig. 1), transplanted NSPCs will be greatly altered by specific environmental cues. To investigate how these environmental differences affect cellular properties, we transplanted NSPCs around the lesion area during the acute, subacute, and chronic phases of SCI (acute, immediately after SCI; subacute, 7 days after SCI; chronic, 3 months after SCI), and then selectively isolated them at 7 DAT using flow cytometry [16]. As a control, NSPCs transplanted into naïve spinal cords were isolated at 7 DAT (naïve). Engrafted NSPCs were successfully dissociated and purified even from chronically injured spinal cords, as shown in Figure 3A–3C. Sequencing libraries were constructed from these FACS-purified NSPCs and sequenced on an Illumina GA IIX Sequencer; the sequenced fragments were mapped to the mouse reference genome using TopHat and assembled with Cufflinks [23]. The median expression level of the reconstructed mRNA was estimated by fragments per kilobase of exonic sequences per million aligned reads [24]. Biological replicates were examined by comparing current samples with previous samples [16]. Very strong correlations were observed between replicates for NSPCs transplanted into naïve spinal cords (r = .984, p < .01) and for NSPCs transplanted into acutely injured spinal cords (r = .996, p < .01) (Fig. 3D). These results confirmed the high biological reproducibility of our experimental procedure. The Venn diagram (Fig. 3E) summarizes the overlapping transcripts detected in each sample (Fig. 3E).

thumbnail image

Figure 3. RNA sequencing (RNA-Seq) for neural stem/progenitor cells (NSPCs) transplanted in naïve and injured spinal cords. (A): Images of chronically transplanted NSPCs after the dissociation of injured spinal cord at 7 days after transplantation (DAT). Scale bar = 100 μm. (B): FACS analysis for homogenate of chronically injured spinal cord tissue at 7 DAT. The GFP positive population was observed in the homogenate of NSPC transplanted spinal cords. (C): Images of chronically transplanted NSPCs isolated from the injured spinal cord before and after FACS purification. Scale bar = 500 μm. (D): Biological replicates of RNA-Seq for NSPCs transplanted into naïve and injured spinal cords. (E): Venn diagram showing the number of overlapping transcripts detected in each context (FPKM >1). (F): Scatter plots of the gene expression levels, showing significant differences between NSPCs transplanted into naïve and injured spinal cords. Genes with at least a twofold difference from NSPCs transplanted into naïve spinal cords are indicated in red (upregulated genes) or blue (downregulated genes). (G): Validation of transcriptional repression in NSPCs transplanted into acutely injured spinal cords and transcriptional activation in those transplanted into chronically injured spinal cords by quantitative polymerase chain reaction using 96 randomly selected genes. (H): Principal component analysis score plot of PC1 versus PC2. This score plot indicates the similarities and differences among NSPCs transplanted in different phases of SCI. Data are represented as the mean ± SEM. Abbreviations: FACS, fluorescence activated cell sorting; FPKM, fragments per kilobase of exonic sequences per million aligned reads; GFP, green fluorescent protein; NSPCs, neural stem/progenitor cells; PC, principal component.

Download figure to PowerPoint

Transcriptional Activities of Engrafted NSPCs

We previously reported that transcriptional activity is generally suppressed in acutely transplanted NSPCs as a stress response [16]. Consistent with this finding, almost all genes were downregulated in NSPCs transplanted into acutely injured spinal cords compared with those transplanted into naïve spinal cords (Fig. 3F). However, this stress response was not significant in chronically or subacutely transplanted NSPCs (Fig. 3F). The number of downregulated genes was similar to that of upregulated genes when they were transplanted into subacutely injured spinal cords. Interestingly, contrary to the finding in acutely transplanted NSPCs, many genes were upregulated in chronically transplanted NSPCs (Fig. 3F). qPCR analysis confirmed the decreased transcriptional activities in acutely transplanted NSPCs and the increased transcriptional activities in chronically transplanted NSPCs (Fig. 3G). These results indicate that transcriptional activities increase when NSPCs are transplanted into chronically injured spinal cords.

To further assess the environmental effects of different phases of SCI on transplanted NSPCs, a principal component analysis (PCA) was performed using 13,480 genes. The value of chronically transplanted NSPCs was higher than that of the control on principal component (PC) 1 and PC2 axes (Fig. 3H). In addition, NSPCs transplanted into acutely and chronically injured cords were positioned on sides opposite to the control (Fig. 3H), which suggested opposite effects in acutely and chronically injured spinal cords with transplanted NSPCs. In the analysis for the genes with significant factor loadings (p > .05), the neural differentiation-associated genes, Mbp, Gfap, and neuronatin (Nnat) positively correlated with both PCs, whereas the oligodendrocyte differentiation-associated gene Cnp (also known as Cnpase) positively correlated only with PC1 (Fig. 3H). Because subacutely transplanted NSPCs demonstrated the highest scores on PC1, subacutely injured spinal cords may promote the oligodendrocyte differentiation of transplanted NSPCs. Furthermore, neuronal differentiation-associated gene B3-tubulin (Tubb3) and thymosin beta 4 (Tmsb4x) positively correlated only with PC2, thus suggesting that the neuronal differentiation of transplanted NSPCs was promoted in chronically injured spinal cords.

Differentiation Phenotype of NSPCs Transplanted at Different Phases of SCI

While transplanted NSPCs reported to have a potency to generate neurons or oligodendrocytes, their differentiation phenotypes are largely dependent on the graft microenvironment [8, 15, 25]. Specifically, engrafted NSPCs differentiated into oligodendrocytes and neurons in subacute SCI microenvironments but rarely differentiated into these lineages in acute SCI microenvironments [16, 22]. To assess the environmental influence of chronically injured spinal cords on NSPC differentiation, we analyzed the differentiation-associated gene expression in engrafted NSPCs determined by RNA-Seq. A set of differentiation-associated gene expression in engrafted NSPCs tended to show upregulation in chronically injured spinal cords but downregulation in acutely injured spinal cords compared with naïve spinal cords (Fig. 4A). This result indicates that chronically injured spinal cords are permissive for the differentiation of engrafted NSPCs. We further compared the expression of lineage-specific genes among NSPCs transplanted at different phases of SCI. Consistent with the results of PCA (Fig. 3H), gene expression of the transcription factor Ascl1, which induces neuronal commitment and differentiation, was higher in NSPCs transplanted into chronically injured spinal cords than in those transplanted into other conditions, including naïve spinal cords (Fig. 4A). In addition, NSPCs transplanted in chronically injured spinal cords showed highest gene expression of the immature neuron markers doublecortin (Dcx), Dpysl3 (also known as Tuc4), and Tubb3 among four conditions (Fig. 4B, 4C). These results were confirmed by qPCR (Fig. 4D). Gene ontology analysis [26] also showed that “neurogenesis,” “generation of neurons,” and “neuron differentiation” were significantly over-represented in NSPCs transplanted into chronically injured spinal cords compared with those transplanted into naïve spinal cords (adjacent p value < .001). Furthermore, at 42 DAT, the proportion of Hu/GFP double-positive immature neurons to GFP-positive transplanted cells was highest in chronic SCI among the four groups (Fig. 4E) and graft-derived NeuN-positive mature neurons were observed after the chronic phase of transplantation (Fig. 4F). These results demonstrate that neuronal differentiation of engrafted NSPCs occurs most prominently in chronically injured spinal cords.

thumbnail image

Figure 4. Neuronal differentiation potential of neural stem/progenitor cells (NSPCs) transplanted at different phases of spinal cord injury (SCI). (A): Heatmap showing differences in gene expression of differentiation-associated genes (258 genes) in NSPCs transplanted at different phases of SCI. This heatmap represents normalized expression profiles (red, upregulated; blue, downregulated) in NSPCs transplanted into injured spinal cords compared with those transplanted into naïve spinal cords. (B): Gene expression levels of neurogenesis-associated genes in NSPCs transplanted into different conditions as determined by RNA sequencing (RNA-Seq). (C): Visualization of expression data for Ascl1 and Tubb3. Wiggle plots directly represent the number of each individual read. Their expression was prominent in chronically transplanted NSPCs. (D): Comparison of gene expression of neurogenesis-associated genes among NSPCs transplanted at different time points by quantitative polymerase chain reaction (n = 6 per group). (E): Comparison of the rate of neuronal differentiation of NSPCs in different conditions at 42 days after transplantation (DAT). The proportion of Hu/GFP-double-positive cells to total GFP-positive cells was analyzed. *, p < .05 versus control by ANOVA with Dunnett's post hoc test. (F): Immunohistochemical analysis of injured spinal cords after chronic transplantation at 42 DAT using anti-GFP (green) and anti-NeuN (red) antibodies. Arrowhead indicates NeuN positive engrafted cells. Scale bar = 100 μm. Data are represented as the mean ± SEM. Abbreviations: FPKM, fragments per kilobase of exonic sequences per million aligned reads; GFP, green fluorescent protein.

Download figure to PowerPoint

Regarding oligodendrocyte differentiation, many GSTπ-positive oligodendrocytes derived from NSPCs were observed at 42 DAT after chronic phase transplantation (Fig. 5A). Although the gene expression of the obligate oligogenic transcription factors Olig1 and Olig2 was highest in subacutely transplanted NSPCs, their expression levels in chronically transplanted NSPCs were higher than those in NSPCs transplanted into naïve spinal cord (Fig. 5B, 5C). Moreover, gene expression of the oligodendrocyte markers Cspg4 (also known as neuronal/glial 2 (Ng2)), platelet-derived growth factor receptor (Pdgfr)-α, Sox10, and Cnpase was prominent in NSPCs transplanted into subacutely injured spinal cords (Fig. 5B, 5C). Expression levels were similar between NSPCs transplanted into chronically injured spinal cords and those transplanted into naïve spinal cords. These results were confirmed by qPCR (Fig. 5D). Furthermore, quantitative analysis confirmed the highest percentage of oligodendrocyte differentiation in NSPCs transplanted into subacutely injured spinal cords (Fig. 5E), demonstrating that oligogenesis by engrafted NSPCs was most prominent in subacute SCI. The proportion of graft-derived oligodendrocytes in chronically injured spinal cords was similar to that in naïve spinal cords (Fig. 5E). These results reveal that oligodendrocyte differentiation of transplanted NSPCs is facilitated in subacute SCI microenvironments and that oligodendrocyte generation by engrafted NSPCs is not inhibited in chronic SCI microenvironments.

thumbnail image

Figure 5. Oligodendrocyte differentiation potential of neural stem/progenitor cells (NSPCs) transplanted at different phases of spinal cord injury (SCI). (A): Immunohistochemical analysis of injured spinal cord 42 days after chronic transplantation. Oligodendrocytes derived from engrafted NSPCs (green) were identified by their GST-π expression (red). Scale bar = 100 μm. (B): Visualization of expression data for Olig1 and Cnp. Their expression levels were prominent in subacutely transplanted NSPCs. (C): Gene expression levels of oligogenesis-associated genes in NSPCs transplanted in different conditions as determined by RNA sequencing (RNA-Seq). (D): Validation by quantitative polymerase chain reaction of gene expression levels of oligogenesis-associated genes in NSPCs transplanted at different time points (n = 6 per group). (E): Comparison of the rate of oligodendrocyte differentiation among NSPCs transplanted at different time points after SCI. The proportion of GST-π/GFP double-positive cells to total GFP-positive cells was analyzed at 42 DAT. *, p < .05 versus control by ANOVA with Dunnett's post hoc test. Data are represented as the mean ± SEM. Abbreviations: FPKM, fragments per kilobase of exonic sequences per million aligned reads; GFP, green fluorescent protein.

Download figure to PowerPoint

Secretory Activity of NSPCs Transplanted at Different Phases of SCI

The important role of engrafted NSPCs for CNS disorders lies not only in their capacity for cell replacement but also in their ability to become trophic mediators [4, 5]. Although it is evident that NSPC secretion intermediates neural inflammatory and regenerative responses [5], their secretory activities after transplantation remain unclear. To elucidate these activities, we compared gene expression of secretory molecules among NSPCs transplanted at different phases of SCI. RNA-Seq identified many secreted molecules that were expressed in transplanted NSPCs (Fig. 6A) and revealed that NSPC secretion decreased markedly in acutely injured spinal cords but increased in chronically injured spinal cords, compared with naïve spinal cords (Fig. 6A, 6B). qPCR confirmed the activated production of regenerative molecules in NSPCs transplanted into chronically injured spinal cords (Fig. 6C). We further quantified the gene expression of these growth factors in chronically transplanted settings at 7 DAT by qPCR and found that they were significantly increased compared to controls (Fig. 6D). These results demonstrate that the secretory activities of NSPCs for several neurohumoral factors are very high when transplanted into chronically injured spinal cords.

thumbnail image

Figure 6. Secretory activities of neural stem/progenitor cells (NSPCs) transplanted at different phases of spinal cord injury (SCI). (A): Heatmap showing differences in gene expression of secretory molecules among NSPCs transplanted at different phases of SCI. This heatmap represents expression profiles (red, upregulated; blue, downregulated) of NSPCs transplanted into injured spinal cords normalized by those transplanted into naïve spinal cords. (B): Visualization of expression data for Igf2 and Hdgf. They were expressed at high levels in chronically transplanted NSPCs. (C): Gene expression levels of secreted molecules in NSPCs transplanted at different phases of SCI as determined by quantitative polymerase chain reaction (qPCR). Data were normalized by the expression of NSPCs transplanted into naïve spinal cord. (D): Gene expression of secretory molecules increased in the injured spinal cord after chronic transplantation. Gene expression of secretory molecules in chronically transplanted injured spinal cords at 7 days after transplantation were compared with controls by qPCR (n = 6 per group). Data are represented as the mean ± SEM.

Download figure to PowerPoint

Functional Improvement After NSPC Transplantation

After confirming the cellular therapeutic activities of NSPCs transplanted into chronically injured spinal cords by comparing them with activities in other microenvironments, we assessed functional improvement after NSPC transplantation using an open-field motor score system. In the acutely and subacutely transplanted groups, mice showed significantly improved functional recovery compared to the medium-injected control groups after SCI (Fig. 7A, 7B). However, chronically NSPC-transplanted mice did not exhibit significantly improved locomotor recovery compared with pretransplanted mice (repeated measured ANOVA, p > .05) or medium control mice (one-way ANOVA, p > .05) at 42 DAT (Fig. 7C). Furthermore, grip walk test (Fig. 7D) and footprint analyses (Fig. 7E) also did not show improved locomotor function after chronic NSPC transplantation.

thumbnail image

Figure 7. Effect of neural stem/precursor cell (NSPC) transplantation into chronically injured spinal cords on locomotor function. (A–C): Time course of locomotor function as determined by the Basso Mouse Scale (BMS) of acutely (A, immediately after spinal cord injury [SCI]), subacutely (B, 7 days after SCI), or chronically (C, 3 months after SCI) transplanted groups and control group (n = 8–10 per group). Functional recovery after SCI was significantly improved by acute or subacute transplantation of NSPCs but was not improved by chronic transplantation. *, p < .05, two-way repeated measures ANOVA with Bonferroni test. (D): Grip walk test for chronically transplanted mice after chronic transplantation (n = 11). There is no significant difference among any time point (repeated measured ANOVA, p > .05). (E): Footprint analysis for chronically transplanted mice before (Pre-Tp) and after (Post-Tp; at 42 days after transplantation) chronic transplantation (n = 11). There is no significant difference between the two (paired t test, p > .05). Data are represented as the mean ± SEM.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Although the translation of stem cell approaches developed in the laboratory to the clinic is in progress [5, 6], no studies identified their cellular properties after transplantation. Because cells are not drugs with precise structures, but highly complex biological entities [6], detailed understanding of their cellular biology is essential for successful translation of stem cell-based therapies. Although defining a therapeutic stem cell product is challenging, RNA-Seq can provide complete information about cell architecture and activities by determining the absolute quantity of every molecule in a cell population [17]. In this study, we comprehensively assessed the molecular and cellular properties of NSPCs transplanted into chronically injured spinal cords by the combination of flow-cytometric isolation and RNA-Seq technology.

Although many authors have failed to achieve survival and migration of NSPCs in chronic SCI microenvironments [14, 27, 28], we demonstrated here that chronically injured spinal cords allow for the survival and migration of transplanted NSPCs. This discrepancy may be attributed to differences in the timing of transplantation during the chronic phase of SCI. Most previous studies transplanted stem cells in the early chronic phase of SCI (4–8 weeks after SCI) [14]; however, we transplanted NSPCs in the late chronic phase of SCI (3 months after SCI). Recently, CSPG perturbation was reported to dramatically increase the survival and migration of NSPCs transplanted during the early chronic phase of SCI [12]. That report indicated the central role of CSPGs in the failure of early chronic transplantation strategy. During the early chronic phase, a large amount of CSPGs existed around the lesion site of SCI, whereas most CSPGs were limited to the glial scar at 3 months after SCI (supporting information Fig. 1). Although a few CSPGs were observed around the lesion site at 3 months after SCI, CSPGs were reported to become localized to a limited area of the lesion site thereafter [29–32]. Considering our and previous results, the reduction in extracellular CSPGs would permit the survival and migration of NSPCs transplanted 3 months after SCI and perhaps at later stages. Similar spatial and temporal patterns of CSPG expression were observed in human injured spinal cords [33]; transplanted cells will thus survive and migrate even in injured spinal cords in late chronic SCI patients.

In this study, we found that engrafted NSPCs were not transcriptionally dormant but highly active in chronically injured spinal cords. NSPCs will be activated in response to nutritional stimuli from their microenvironment. Until this study, chronically injured spinal cords have been considered to constitute a growth-inhibiting environment [2, 34] that restricts the reparative capacity of engrafted cells [15, 27]. However, our results demonstrate the robust expression of many growth factors in chronically injured spinal cords. Most of these growth factors were reported to stimulate the transcriptional activity of NSPCs, which promotes cellular survival, differentiation, maturation, and secretion. Especially, IGF and FGF families are key modulators of NSPC activation in the developing CNS [35, 36]. Our results indicate that the nutrient-rich environment of chronically injured spinal cords promotes the cellular activity of engrafted cells, which could amplify the therapeutic advantages of engrafted cells.

Engrafted NSPCs are recognized to improve pathological conditions of the CNS through multiple mechanisms such as by replacing damaged neurons and oligodendrocytes, enhancing endogenous repair [37], promoting angiogenesis [38], modulating inflammation [39, 40], and protecting neural cells [5, 9]. These beneficial effects depend on the differentiation and secretion capacity of engrafted cells. Because these capacities are strongly regulated by environmental signals [16], identification of the environmental impact on NSPC differentiation and secretion is required in chronically injured spinal cords for successful clinical translation into chronic SCI patients. In previous reports, their differentiation was generally directed to the glial lineage in injured spinal cords [14, 22, 25]; the microenvironment after SCI has thus been considered to be gliogenic and to inhibit the neuronal differentiation of NSPCs. However, our results reveal that chronically injured spinal cords allow the neuronal differentiation of NSPCs. This is probably due to significant reduction of astrogenic cues at this phase. The most major extrinsic mechanism for promoting astrocyte differentiation is activation of the p130/JAK/STAT pathway [41–44]. We observed that gene expression of IL-6 and LIF, a potent activator of the JAK/STAT pathway [44, 45], returned to its normal range during the chronic phase of SCI. Because NSPCs have the potential to differentiate into mature functional neurons before transplantation [16], these cells could generate neurons without extrinsic astrogenic direction in chronically injured spinal cords. In addition, FGF-2 and IGF-1, essential regulators of neurogenesis [35, 36, 46], were constantly upregulated during the chronic phase of SCI. These neurogenic factors may support the neurogenesis of engrafted cells. Furthermore, oligodendrocyte generation of engrafted NSPCs was not inhibited by chronically injured spinal cords. Consistent with this result, many researchers successfully generated oligodendrocyte in the early chronic phase of transplantation [11, 12, 15]. Thus, chronically injured spinal cords will be permissive for oligodendrocyte differentiation. Our results reveal that engrafted NSPCs retain their multilineage potential in chronically injured spinal cords, suggesting that the cell replacing strategy by generating functional neurons and oligodendrocytes is more feasible in chronically injured spinal cords than considered previously.

Chronically transplanted NSPCs differentiated mainly into neurons or oligodendrocytes, although a small percentage differentiated into astrocytes. Because astrocytes have dual nature in chronic SCI, this small population of transplanted cells may have a role in preventing or promoting tissue repair [47]. In particular, endogenous astrocytes exert a beneficial effect by replacing oligodendrocytes and secreting trophic molecules, while also preventing axonal regeneration by forming a glial scar and express CSPGs in chronic SCI [48]. In this study, chronically transplanted NSPCs did not express CSPGs (supporting information Fig. 1D) and not contribute to glial scar formation (supporting information Fig. 1E, 1F) probably due to the lack of STAT3 activation [47, 49]. Because this result suggests that chronically transplanted NSPCs did not prevent endogenous repair mechanism, a combination of stem cell-based therapies and the promotion of axonal regeneration may be feasible for treating chronic SCI.

Unlike chronic SCI microenvironments, subacute SCI microenvironments facilitated oligodendrocyte differentiation but not neuronal differentiation of engrafted NSPCs. In the subacute phase of transplantation, Hofstetter et al. [25] failed to generate neuronal cells from Ngn2-overexpressing NSPCs. These genetically engineered NSPCs exclusively differentiated into neurons in vitro but mainly underwent oligodendroglial differentiation after transplantation [25]. In contrast, embryonic stem cell-derived oligodendrocyte precursor cells successfully differentiated into oligodendrocyte in subacutely injured spinal cords [15]. Based on these previous studies and our results, subacutely injured spinal cords appear to contain a huge oligogenic cue that can override the intrinsic strong determinant of neuronal differentiation. Considering the environmental aspect, subacutely injured spinal cords will be relatively suitable for remyelination strategy using stem cells.

Although the vital importance of trophic factors released from engrafted NSPCs for pathological improvement in the CNS are recognized [5], it remains unclear what factors engrafted NSPCs secrete. In this regard, RNA-Seq is quite useful for the discovery of novel transcripts by providing unprecedented opportunities to study all expressed genes [17]. In this study, RNA-Seq analysis identified and quantified every molecule secreted from engrafted NSPCs in each graft environment. Most of these factors are reported to promote CNS tissue repair. In particular, the FGF, VEGF, and PDGF families are potent stimulators of angiogenesis [50]. Neurotrophic factors such as NGF and BDNF strongly induce endogenous neurogenesis [51], and IGF-1 and FGF-2 induce the proliferation, specification, and differentiation of endogenous oligodendrocyte precursor cells [52, 53]. Furthermore, VEGF-α, PDGF, IGF-1, NGF, and BDNF have great therapeutic potential in preventing neural apoptosis in several CNS disorders [50, 51, 54]. The TGF-β family is known to prevent the activation of inflammatory cells and oppose pro-inflammatory cytokines [55]. Given the key roles of these substances in tissue repair and neuroprotection, the pleiotropic actions of NSPCs will be attributed to the production of these growth/neurotrophic factors. Notably, chronically transplanted NSPCs showed higher secretory activities than those transplanted at other phases of SCI. Because NSPC secretion are significant for functional improvement at acute and subacute phases of transplantation [5], NSPCs will potentially have adequate capacity to promote tissue repair and protect neural cells even in chronically injured spinal cords.

As described above, NSPCs transplanted into chronically injured cords had a high capacity for providing neurons/oligodendrocytes and producing trophic factors; however, they did not improve locomotor function. These results indicate that the failure of the chronic transplantation is not due to the lack of cellular therapeutic activities but due to the refractory state of chronically injured spinal cords. Therefore, environmental modification, rather than modulation of transplanting cells, will be feasible and important for chronic SCI treatments [12]. Recent mounting evidence suggests that there exists more intrinsic regenerative potential in the adult spinal cord than originally postulated. Specifically, axonal growth responses were observed in adult spinal cords after various interventions in animal models [56]. In clinical practice, modest neurological recovery was seen with rehabilitation or pharmacologic agents in chronic SCI patients [2, 57]. Most of these successful therapies were targeted to promote endogenous plasticity mechanisms. This indicates that human chronically injured spinal cords have a regenerative capacity that responds to therapeutic interventions. We believe that stem cell-based therapies for chronic SCI patients need to be combined with the therapies to improve tissue plasticity in patients.

Conclusion and Summary

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Although most SCI patients have chronic SCI, many researchers have focused on treatments for acute or subacute SCI and only a few chronic studies have been reported [14]. This is a disquieting issue for patients with chronic SCI, who are the most ardent consumers of information regarding cell transplantation technologies. Although the development of treatments for chronic SCI is challenging and there are many significant hurdles to overcome, stem cell-based therapies have potential in the treatment of chronic SCI because the therapeutic activities of these cells are not dormant. This field will continue to evolve with the hope that further refinement and detailed understanding will increase the chances that cell transplantation will someday emerge as a fruitful treatment for patients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

This work was supported in parts by Grant-in-aid for Scientific Research, Scientific Research on Innovative Areas, Challenging Exploratory Research from the Ministry of Education, Science, Sports and Culture of Japan, and research foundations from ZENKYOREN and the general insurance association of Japan.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion and Summary
  8. Acknowledgements
  9. Disclosure of Potential Conflicts of Interest
  10. References
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
STEM_1404_sm_SuppFig1.pdf547KSupporting Information Figure 1.
STEM_1404_sm_SuppData1.TIF929KSupporting Information Data 1.
STEM_1404_sm_SuppData2.TIF883KSupporting Information Data 2.
STEM_1404_sm_SuppTab1.tif1341KSupporting Information Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.