Expression of semaphorin 3A and its receptors in the human intervertebral disc: potential role in regulating neural ingrowth in the degenerate intervertebral disc

A range of post-mortem and surgical samples were used, in which there were 14 samples (7 AF and 7 NP) in each cohort. The post mortem cohort ranged from aged 30 to 79 years (mean: 57 years) and the surgical cohort ages 29 to 56 years (mean: 39 years). Cells were isolated from each sample as previously reported [49] and RNA extracted using Trizol™ (Invitrogen, Paisley, UK) according to the manufacturer's instructions. RNA was then treated with 4U DNAse I (Invitrogen) to remove any DNA contamination and reverse transcribed using an Applied Biosystems (Warrington, UK) High Capacity cDNA Reverse Transcription kit as per the manufacturer's instructions.

Aliqouts of cDNA were subjected to PCR using primers specific for the house-keeping gene GAPDH, as well as sema3A, sema3F, NRP-1, NRP-2, PLXA1, PLXA2, PLXA3, PLXA4 and PGP9.5 (Table 1). Each 25 μl reaction mixture consisted of: 19.125 μl molecular biology grade water, 2.5 μl 10 × PCR buffer w/o MgCl₂, 0.75 μl 50 mM MgCl₂, 1 μl 5 mM dNTP mix, 0.25 μl of 100 μM respective forward and reverse primers, 0.125 μl Platinum Taq polymerase (Invitrogen) and 1 μl of respective cDNA (5 ng/μl). Articular cartilage cDNA and molecular grade water were used as positive and negative controls respectively. Amplification was performed in a thermal cycler (Veriti, Applied Biosystems, UK) under the following conditions: enzyme activation at 95°C for 90 seconds followed by 38 cycles of 95°C for 20 seconds, 60°C for 20 seconds and 72°C for 60 seconds; and a final extension step of 72°C for 10 minutes. PCR products were then run on a 2% agarose gel containing GelRed (Biotium, Hayward, CA, USA) alongside a 100 bp DNA ladder (Bioline Hyperladder IV, London, UK). Bands were visualised by UV transillumination and images acquired using Genesnap software (Syngene, Cambridge, UK).

Immunohistochemistry for sema3A, PGP 9.5 and CD31 was conducted on cohorts of healthy and degenerate disc tissue. The healthy disc cohort consisted of 10 healthy samples which were histologically graded between 0 and 3 according to previously published criteria [48] and had an age range of 30 to 61 years (mean: 51 years). There were a further 10 samples considered to be mildly degenerate (grades 4 to 6) ranging from ages 30 to 74 years (mean: 53 years), a moderately degenerate cohort of 10 samples (grades 7 to 9) ranging from ages 28 to 79 years (mean: 61 years) and a severely degenerate cohort (grades 10 to 12) of 10 samples ranging from ages 22 to 78 years (mean: 41 years). The latter two cohorts primarily consisted of samples obtained during surgery. The immunohistochemistry protocol followed was as previously published [49, 50]. Antigen retrieval was performed using: type II trypsin for sema3A (1 mg/ml for 20 minutes at 37°C), heated citrate buffer for PGP9.5 (10% v/v citrate buffer pH6.0 for 15 minutes at 95°C) and pepsin for CD31 (1 mg/ml in 0.2 M HCL for an hour at 37°C). Subsequently, the slides were washed with tris-buffered saline (TBS) and non-specific secondary antibody binding was blocked with 25% normal goat serum diluted in 1% bovine serum albumin (BSA)/TBS for sema3A & PGP9.5 and 10% normal goat serum diluted in 1% BSA/TBS for CD31. Each primary antibody was diluted in 1% BSA/TBS. Sema3A was applied at a 1:250 dilution, PGP9.5 at 1:20 and CD31 at 1:10 and incubated at 4°C overnight. Slides were then washed with TBS and the biotinylated secondary antibody was applied: goat anti-rabbit (sema3A) and goat anti-mouse (CD31/PGP9.5). Both antibodies were applied at a 1:300 dilution in TBS. Positive controls included rat brain for sem3A, human tonsil for CD31 and human brain for PGP9.5. Binding of the secondary antibody was disclosed with the streptavidin-biotin complex (Vector Laboratories, Peterborough, UK) technique with 3,3'-diaminobenzidine tetrahydrochloride solution (Sigma, Poole, Dorset, UK). Sections were counterstained with Mayer's haematoxylin (Raymond A. Lamb, East Sussex, UK), dehydrated, and mounted with Pertex (CellPath Plc, Powys, UK).

Quantitative analysis of staining was performed and for each section five random fields of view were observed in each morphologically distinct region of the disc (nucleus pulposus (NP), inner annulus fibrosus (IAF), outer annulus fibrosus (OAF)). Two hundred cells were counted per region and the percentage of positive cells was then calculated for each region of the IVD. Statistically, the data were non-parametric and analysed using the Wilcoxon matched pairs test when comparing data within regions of the disc. Additionally, the Mann-Whitney U test was used to compare data across varying grades of degeneration. Data were then plotted as mean ± standard errors to represent the 95% confidence intervals.

Immunoreactivity to the sema3A antibody was clearly identifiable throughout the entire cohort of IVD samples. Staining was primarily intracellular although some peri-cellular staining was also observed. Both the fibroblast-like cells of the AF and the chondrocyte-like cells of the NP stained positively (Figure 2). Immunoreactivity was also observed in vascular endothelial cells. In addition, the smooth muscle cells in the tunica media of small blood vessels, located primarily in the OAF of the more degenerate disc samples (grades 9 to 12) also exhibited positive staining. Of note, smooth muscle and endothelial cells were not counted for statistical analysis with only the chondrocyte-like cells of the NP and AF considered as sema3A immunopositive positive cells. There was no significant staining within the extracellular matrix of any of the disc samples.

Immunopositivity was localised primarily to cells within the AF with approximately 80% of cells demonstrating positivity for sema3A (Figure 3). A gradient of expression was observed more proximally in the disc as the cells of the IAF showed 40% positivity and the NP cells only showed 5% positivity. There were significant differences (P ≤ 0.01) in the number of cells demonstrating positivity for sema3A between all three morphological regions of the IVD (Figure 3).

Expression of sem3A in the OAF decreased significantly in the mildly degenerate cohort (grades 4 to 6) compared to that of healthy OAF (P ≤ 0.01) (Figure 3). Percentage cell immunopositivity for sema3A was 33% in the mildly degenerate OAF compared to 26% in the IAF and 19% in the NP, although these differences were not significant (P ≥ 0.05).

In moderately degenerate disc samples, immunoreactivity decreased to 12% in the OAF, 21% in the IAF and increased to 18% in the NP when compared to mildly degenerate samples. This positivity was found almost exclusively in cell clusters of the NP. No significant differences were found between the regions of the IVD. However, there were significant differences (P ≤ 0.05 and 0.005) between the IAF and OAF of the moderately degenerate and healthy disc respectively (Figure 3). The severely degenerate disc samples showed a reduction in the number of sema3A immunopositive cells compared to the healthy disc with the OAF and IAF having 2.2% and 5.8% immunopositive cells respectively, which were significantly lower than expression in the healthy disc (P ≥ 0.001 and P ≥ 0.005). The number of sema3A immunopositive cells in the NP was significantly increased (P ≥ 0.05) when compared to cells in the healthy disc with approximately 30% of cells (chondrocyte-like cell clusters) staining positively (Figure 2). Significant differences were also seen between regions of the disc (P ≥ 0.05).

The degenerate sample cohort was further analysed to assess any differences in numbers of immunopositive cells between post mortem degenerate and surgical (symptomatic patients) degenerate samples (Figure 4). The average total disc sema3A cell immunopositivity in the PM degenerate cohort was 32%, while in the surgical degenerate cohort it was 12%. This reduction in sema3A expression was significant (P ≤ 0.05). When comparing the surgical degenerate cohort to the healthy samples, which showed an average 50% cell immunopositivity, a markedly significant difference was observed (P ≤ 0.001) (data not shown).

The staining within the disc was sparse and specific to neuronal cells (Figure 5) with no positivity within the extracellular matrix of the disc. Immunopositivity was mainly localised to the areas of the OAF and IAF. Staining was intracellular, but had a granular appearance. Of the healthy cohort, 20% (2/10) of IVD samples demonstrated positivity while 8% (1/12) of PM degenerate samples showed areas of immunopositivity. Of the surgical degenerate samples 39% (7/18) showed positivity for PGP9.5.

Endothelial cells found within the disc were immunopositive for CD31 with staining found exclusively in endothelial cells surrounding developing blood vessels (Figure 5). The chondrocyte-like cells of the AF and NP were negative. No staining was observed within the extracellular matrix of the respective regions of the disc. Of the healthy samples only 20% (2/10) demonstrated positivity for CD31, 0% of the PM degenerate samples were positive, while 56% (10/19) of the surgical samples showed some level of positivity. Of note, five of the 10 CD31 positive samples were also positive for PGP9.5. Interestingly, the surgical degenerate cohort showed the lowest level of sema3A immunopositivity.