The role of CXCR3 and its ligands expression in Brucellar spondylitis

This study is a descriptive observational. Patients diagnosed with BS were recruited in spine surgery at two hospitals in Xinjiang, Urumqi from 2015 to 2019. The study was approved by the ethics committee of Xinjiang Medical University. The patients were categorized as acute phase (with symptoms less than 3 months), subacute phase (3–6 months) and chronic phase (more than 6 months). The clinical symptoms and laboratory results were collected. The paravertebral cartilage tissue (include close and distant tissue) and peripheral blood of 29 BS patients were collected. 15 healthy controls were enrolled. The consent of patients was received at the same time. The data used to support the findings of this study are available from the corresponding author upon request.

The inclusion criteria of BS patients: 1. Patients have a history of contact with Brucella infected livestock or livestock products, consumption of raw meat and unpasteurized dairy products. 2. The patients have a persistent pain in the lower back, sacroiliac, fever, sweating, fatigue, or joint pain for days or even weeks. 3. Brucella was detected after one-week blood culture. 4. Tube agglutination test is positive. 5. Magnetic Resonance Imaging (MRI) reports that the adjacent surface of vertebra is destroyed, and there are hyperplasia and sclerosis. 6. Exclude patients with other immune, neoplastic diseases and HIV infection.

BS tissue specimen (n = 29) and corresponding peripheral blood samples (n = 29) was obtained. All samples were collected at the close and distant normal tissues, by which tissues in vitro were washed in PBS (Boster, CN), and fixed in 4% paraformaldehyde (Beijing Chemical Works, CN) for 24–48 h. Then the tissues were decalcified in 10% EDTA decalcification solution for 21–30 days. The samples were then embedded in paraffin and sectioned transversely at 4 μm continuously to make parallel sections. Peripheral blood (3 ml) was drawn from each subject and treated with Ficoll-Hypaque Solution (Tianjin TBD, CN) to get peripheral blood mononuclear cells (PBMCs).

H&E staining was performed according to routine procedure, including hematoxylin (Solarbio, CN) staining for 3 min, eosin (Solarbio, CN) staining for 2 min, 1% hydrochloric acid differentiation, ethanol gradient dehydration, neutral gum seal, and microscopic observation of histopathological changes.

In short, sections, 4 μm thick, were deparaffinised and hydrated. After antigen retrieval, they were treated with primary antibody (Dilution ratio: Brucella-Ab, 1:200; IFN-γ, 1:100; CXCR3, 1:200; CXCL9, 1:200; CXCL10, 1:200) (Concentration: Brucella-Ab, 5 μg/ml; IFN-γ, 10 μg/ml; CXCR3, 5 μg/ml; CXCL9, 5 μg/ml; CXCL10, 5 μg/ml) (Beijin Bioss, CN) at appropriate dilution at 4 °C overnight. Sections were incubated with secondary antibody (Zhongshan Jinqiao, CN) for 1.5 h. Visualisation was performed with 3, 3′-diaminobenzidine (DAB) as substrate, applied for 1.5 min. Sections were counterstained with Hematoxylin. Antigen expression was analyzed under × 20 medium power lens. Positive signals were required to meet two criteria: irregular patchy dark brown granules; more than five clusters. IHC positive expression area was processed by Image J. The average optical density (AOD) obtained was the final processing results.

Total RNA was extracted from the PBMCs by Trizol reagent (Invitrogen, USA). RNA quantity (> 80 ng/μl) and purity (OD260/OD280 = 1.8–2.0) were determined by nucleic acid quantifier. RNA was treated with DNase I (Takara, Japan) before reverse transcription to eliminate contaminating genomic DNA. cDNA was synthesized by reverse transcription according to the PrimeScript™ RT reagent Kit (Takara, Japan). The total amount of RNA was 3.75 μl. For RT-PCR, we mixed a 2 μl cDNA, 1 μl forward primers and 1 μl reverse primers (the final concentration of the primers was 10 μM.), 12.5 μl TB Green Premix Ex Taq (Takara, Japan) and 8.5 μl DEPC water. Thermal cycle parameters were: 95 °C for 30 s for 1 cycle, then 95 °C for 5 s, and 60 °C for 30 s for 39 cycles. Primer sequences are shown in Table 1. GAPDH was used as the internal control. The relative expression was calculated by the comparative cycle threshold method (2^-△△t).

Serum levels of IFN-γ, CXCL9, CXCL10 (eBioscience, Austria) and autoantibodies against CXCR3 (CellTrend GmbH Luckenwalde, Germany), were detected in accordance with ELISA kit instructions. Standard wells, blank control wells and sample wells were set on 96-well plates. The diluted standard was added to the standard well, the sample dilution was added to the blank control well, and the test serum was added to the sample well. Then they were placed in 37 °C incubator for 60 min. After washing, color development and termination, the absorbance (A) was detected at a wavelength of 450 nm. For anti-CXCR-3-antibodies, the CXCR-3-receptor has been pre-coated onto a microtiter plate. During the first incubation the anti-CXCR-3-antibodies of the samples are immobilised on the plate. The autoantibodies are detected with a POD labeled antihuman IgG antibody. In the following enzymatic substrate reaction the intensity of the colour correlates with the concentration and/ or avidity of anti-CXCR-3-antibodies. The concentration was determined according to the standard curve.

A total of 29 BS patients were enrolled in this study. Detailed demographic information was listed in Table 2.. Among them, 19 patients were male (65.5%) and 10 patients were female (34.5%), with an average age of 43.37 ± 18.27 years old (1 to 76 years old). There were 0 case in the acute phase, 1 case in the sub-acute phase, and 28 cases in the chronic phase. Eighteen patients (62.1%) were cattle and sheep herders, and 10 patients (34.5%) ate raw meat or unpasteurized dairy products.

The clinical symptoms and laboratory results of BS patients were showed in Table 3. Some patients showed hyperhidrosis (65.5%), fatigue (62%) and lower back pain (69%). Some patients showed joint pain (51.7%) and fever (55.1%). There were 20 patients (69%) showed anemia, 19 (65.5%) with increased erythrocyte sedimentation rate (ESR), 18 (62.1%) with increased C-reactive protein (CRP), 11 (37.9%) with reduced albumin, 10 (34.5%) with abnormal leukocytes and 2 (6.9%) with of thrombocytopenia. The agglutination test antibody was positive (≥1: 100) in 23 cases (79.3%) with highest antibody titer of 1: 1600. The above data showed that high ESR, CRP and agglutination test antibody positive (≥1: 100) were important laboratory indicators. BS showed a slice-like low-density shadow around the articular surface or the vertebral body, which was invasive bone destruction. The vertebral body signal of MRI lesions was non-specific, showing a slice-like abnormal signal. T1-weighted image (T1WI) showed low-signal, T2WI showed slightly high-signal, and the lipid-pressing sequence showed high-signal. The enhanced scan showed uneven enhancement.

RT-PCR suggested that in PBMCs of BS patients had a higher mRNA expression: IFN-γ (11.24 ± 5.37, p < 0.001), CXCR3 (2.30 ± 0.58, p < 0.001), CXCL9 (2.37 ± 0.93, p < 0.01) and CXCL10 (6.02 ± 4.56, p < 0.001). Compared with IFN-γ (1.05 ± 0.40, p < 0.001), CXCR3 (1.04 ± 0.27, p < 0.001) and CXCL10 (1.18 ± 0.75, p < 0.001) in controls, the difference was statistically significant (Fig. 1a).

We collected peripheral blood from 29 patients with BS and 15 healthy controls. The expression of CXCR3 related cytokines were analyzed from serum levels. IFN-γ are the main related factors of CXCR3. The levels of serum IFN-γ, CXCL9, CXCL10 and the levels of autoantibodies against CXCR3 in the BS patients were higher than control group, and the differences were statistically significant (p < 0.01) (Fig. 1b).

At close lesion tissue of BS patients, there was significant degeneration of cartilage tissue. There was partial necrosis with a number of inflammatory cells infiltration in cartilage, fibrous tissue and small blood vessel hyperplasia. A minute amount of new cartilage tissue can be seen, no obvious granuloma (Fig. 2 a1). At the distant normal tissue, there were few or none inflammatory cells infiltration, and the cartilage tissue was normal (Fig. 2 a2).

In the lesion tissue, rabbit anti-human Brucella-Ab was used to detect the positive cells. It is suggested that the protein expression of inflammatory cells in close lesion tissue was strongly positive (Fig. 2 b1), while it was rarely or negatively expressed in the distant normal tissue (Fig. 2 b2). By comparing the Average Optical Density (AOD) of IHC results, the difference between the close (0.55 ± 0.13) and distant (0.04 ± 0.03) lesion tissue of Brucella-Ab expression was statistically significant (Fig. 1c).

Immunohistochemistry (IHC) is the process of detecting antigens (e. g. proteins) in cells within a tissue section using specific antibodies. Interestingly, at the lesion tissue of BS patients, those markers not only expressed in immune cells but also in cytoplasm. The protein expression of IFN-γ (0.43 ± 0.09), CXCR3 (0.24 ± 0.07), CXCL9 (0.15 ± 0.05) and CXCL10 (0.26 ± 0.04) in the close lesion tissue were increased (Fig. 3 a1, 3b1, 3c1, 3d1). In distant normal tissue, IFN-γ (0.10 ± 0.05), CXCR3 (0.08 ± 0.04), CXCL9 (0.11 ± 0.03) and CXCL10 (0.06 ± 0.03) were slightly expressed (Fig. 3 a2, 3b2, 3c2, 3d2). The difference between IFN-γ, CXCR3 and CXCL10 was statistically significant (Fig. 1c).