Circulating T cell subsets are altered in individuals with chronic spinal cord injury

The local institutional review board approved this study, and informed consent was obtained from all participants prior to study enrollment. Inclusion criteria for individuals with SCI were: ≥18 years old, a history of SCI at any level, an initial injury that occurred ≥1 year prior and injury classification with an American Spinal Injury Association Impairment Scale (AIS) grade of A–D. Potential SCI participants were excluded or study visits rescheduled if they had a concurrent infection as indicated by laboratory or clinical evidence, pressure ulcers, cancer, chemotherapy, neutropenia or autoimmune disease. Uninjured participants were ≥18 years old, without history of SCI, and selected to be within an age range similar to the chronic SCI participants. Peripheral blood samples were collected into sodium heparin-coated tubes via standard venipuncture. Data are derived from male individuals with SCI (N = 22 T cell, N = 19 T_reg) or uninjured controls (N = 11, T cells and T_reg; Table 1). Due to technical issues with sample processing or staining, one participant included in the T_reg analysis was not included in the T cell analysis and four participants included in the T cell analysis were not included in the T_reg analysis. Complete omission of these participants (N = 5) does not eliminate the significance (P values <0.05) of the findings in either the T cell or T_reg analysis. Additional peripheral immune cell types and immune system data analyzed from these and other participants recruited in this study will be described elsewhere.

Peripheral blood leukocytes were isolated from blood using Ficoll Paque Plus Gradient (GE Healthcare), according to manufacturer’s instructions. Cells (2 × 10⁶ cells/100 μl) were incubated for 25 min on ice in the dark with antibodies conjugated to fluorophores, washed in FBS buffer (BD Biosciences Cat# 554656) and fixed in 4 % paraformaldehyde. Antibodies were purchased from BD Biosciences (BD), unless otherwise indicated. T cell panel contained the antibodies: CD3-Alexa 700 (Cat# 557943), CD4-PerCP-Cy5.5 (Cat# 560650), CD8-APC-Cy7 (Cat# 557760), CD38-APC (Cat# 555462) and HLA-DR-FITC (Miltenyi Cat# 130-095-295). (CD69 was also included in the panel, but failed to stain sufficient numbers of cells and so was not used for analysis.) The T_reg panel contained the following antibodies: CD3-Alexa 700 (Cat# 557943), CD4-PerCP-Cy5.5 (Cat# 560650), CD25-PE (Cat# 557138), CD127-APC (Biolegend Cat# 351316), CCR4-PE-Cy7 (Biolegend Cat# 359410) and HLA-DR-FITC (Miltenyi Cat# 130-095-295). (CD45RO was also included in the panel, but was not used in the analysis.) At least 100,000 and at most 480,000 total events were collected using a BD LSRII Flow Cytometer. Spectral compensation was performed with compensation beads (BD) on each day of staining. Analysis was performed using FlowJo software (Treestar, Inc). Cells were first gated as leukocytes using forward (FSC) versus side scatter (SSC), and then, the singlet population was selected for T cell subset analyses. Frequencies reported are of the parent populations.

We analyzed circulating T and regulatory T cells isolated from chronic SCI (N = 22, 19, respectively) and uninjured (N = 11) participants (Table 1). Clinical and demographic features of participants are shown in Table 1. The age of uninjured participants was 50 ± 3 (mean years ± SEM) and ranged 28–66 years. The age of chronic SCI participants was 56 ± 3 (mean years ± SEM) and ranged 28–80 years. Among chronic SCI participants, the most common mechanism of injury was sports (35 %); other mechanisms included falls (26 %), motor vehicle accidents (MVAs, 26 %) or violence other (13 %). Among SCI participants, the time from initial injury ranged from 1 to 44 years; the average time from initial injury was 17 ± 2.7 (mean years ± SEM). The ASIA Impairment Scale (AIS) grades among SCI participants were: A (56.5 %), B (8.7 %), C (8.7 %) and D (26.1 %). The injury level most common among SCI participants was cervical (48 %), followed by thoracic (43 %).

To better understand the potential roles of T cell subsets in immune dysfunction in human chronic SCI, T cell subsets were distinguished using cell surface markers currently considered to be characteristic of human T cell susbets [23]. PBMCs were isolated from blood and cell subsets labeled with multiantibody cocktails, as described in “Methods.” T lymphocyte subsets were characterized by the expression of cell surface markers: all T cells (CD3+), CD4+ or CD8+ T cells (CD3+ CD4+ or CD3+ CD8+), alone and in combination with HLA-DR (MHCII), which is present on activated human T cells [23, 24] (Fig. 1).

Individuals with chronic SCI (N = 22) had a significantly lower frequency of CD3 + T cells, as compared to uninjured individuals (N = 11; Fig. 2a, P < 0.03). The percentage of CD3+ CD4+ T cells was also significantly lower in chronic SCI as compared to uninjured individuals (Fig. 2b, P < 0.04). However, the frequency of activated CD4+ T cells was elevated in individuals with SCI, as indicated by HLA-DR expression (Fig. 2c, P < 0.003), and particularly in individuals with neurologically complete injuries (N = 13; Fig. 2d, P < 0.01) or in individuals with injury levels at T5 and above (N = 14; Fig. 2e, P < 0.01). This is of interest because sympathetic nervous system innervation of immune organs occurs at T6, as described above [4, 5]. The mean fluorescence intensity of HLA-DR on CD4+ T cells was not significantly different between SCI and uninjured individuals (P > 0.57, mean ± SEM 937 ± 98 vs. 862 ± 76, respectively). These data indicate a disruption in homeostasis of the CD4+ T cell subset in individuals with chronic SCI.

The percentage of CD8+ T cells did not differ significantly between chronic SCI (N = 22) and uninjured individuals (N = 11; P < 0.19, median ± SEM median 19 ± 2 vs. 23 ± 2 %, respectively). The mean fluorescence intensity of HLA-DR on CD8+ T cells was not significantly different between SCI and uninjured individuals (P > 0.8, mean ± SEM 929 ± 106 vs. 1063 ± 157, respectively).

Regulatory T cells are a heterogeneous population of T cells which can suppress activation of most other innate and adaptive immune cell types and are dysregulated in neurological and non-neurological chronic diseases, including multiple sclerosis, type I diabetes and rheumatoid arthritis [7, 25, 26]. Due to their importance in modulating immune responses generally and reports of immune depression in SCI [4, 5, 22, 27, 28], we also measured the frequency of regulatory T cells (T_Regs) here. While the transcription factor FoxP3 is the most widely used marker for murine CD4+ T_Regs, it requires intracellular staining (which precludes functional assays) and its expression patterns and stability in humans are less well understood [7]. CD4+ T_Regs express CD3+ CD4+ and high levels of CD25, the IL-2 receptor. The chemokine receptor CCR4+, which facilitates migration, was also shown to be present on most human peripheral blood T_Regs, where it correlates with the expression of Foxp3+ [23, 29, 30]. CD127, the alpha chain of the IL-7 receptor, is inversely correlated with FoxP3+ in human T_Regs [31, 32] and in combination with CD25+ high CD4+ T cells, identifies T_Regs in human peripheral blood [33, 34]. Expression of HLA-DR on T_Regs is currently accepted to define a population of terminally differentiated, effector T_Regs with high suppressive activity [25, 35]. This population is functionally defective in cells isolated from individuals with multiple sclerosis [25]. Therefore, to characterize T_Regs here, we used a gating strategy that included the expression of CD3+, CD4+, high expression of CD25, low or no expression of CD127, and expression of CCR4, HLA-DR (MHCII) or both CCR4 and HLA-DR (Fig. 3).

Using the T_Regs antibody panel, we again observed a lower percentage of CD3+ T cells (P < 0.03, Fig. 4a) and a trend toward a lower percentage of CD4+ T cells in individuals with chronic SCI (N = 19) as compared to uninjured individuals (N = 11; P < 0.07, Fig. 4b). The percentage of CD3+ CD4+ T cells that were CD25+ CD127lo was not significantly different between chronic SCI and uninjured individuals (P < 0.15, Fig. 4c). However, we observed a significant increase in the proportion of CD25+ CD127lo CD4+ T cells that expressed CCR4+ (P < 0.0002, Fig. 4d). Among individuals with SCI, this difference was significant in individuals with neurologically incomplete (N = 9) or complete injuries (N = 10) (P < 0.01, P < 0.05, respectively) and in individuals with injuries at level T5 and above (N = 13) or T6 and below (N = 6; P < 0.01 and P < 0.05, respectively). We also observed a significant increase in the proportion of CD25+ CD127lo CD4+ T cells that expressed HLA-DR (P < 0.02, Fig. 4e) or both CCR4+ and HLA-DR+ (P < 0.003, Fig. 4f).