Data shown illustrate the cytokine profiles (pg/ml) of individual GA-reactive T lymphocyte lines from healthy adult controls, polarized for a Th1 or Th2 phenotype (pTh1, pTh2) or nonpolarized (NP), with supernatants removed and assayed at 32 days in culture (D-32). To generate GA-reactive T cell lines from peripheral blood of healthy adults, PBMC were plated in medium containing 10ng/ml IL-7 without antigen, or with 20μg/ml GA. After 7d, medium was enriched with rhIL-2 (20U/ml), and wells were evaluated for GA specificity at 11d. GA-reactive wells (stimulation index >3, and proliferation >3 SD vs. control) were expanded as lines with rhIL-2, and restimulated with GA and autologous irradiated PBMCs. To produce Th1 and Th2 polarization, lines were treated with IL-12 (2.5ng/ml) and anti-IL-4 antibody (2.5mg/ml), or IL-4 (50ng/ml) and anti-IL-12 antibody (5mg/ml) concurrent with GA stimulation. Supernatants were collected from individual GA-reactive Th1 or Th2 polarized and unpolarized control lines undergoing at least three cycles of antigen stimulation. IFNγ and IL-5 were measured as Th1 and Th2 cytokines by sandwich ELISA.

Rodent P1 OPCs were plated into differentiation medium (no growth factors) containing 33% supernatant from GA-reactive lines, OKT3-stimulated control lines, or autologous serum-containing medium. Strengths of GA stimulation and OKT3 stimulation were titrated to produce similar levels of T cell proliferation as assessed by CFSE dilution or tritiated thymidine incorporation (data not shown). (a,b) After maturation for 5d, OPC cultures exposed to GA-reactive Th1- or Th2-polarized cell supernatants contained significantly more total cells than their OKT3-stimulated supernatant counterparts (a, Th1, p<0.001; Th2, p<0.01 ANOVA plus Bonferroni post test), and significantly more CNPase⁺ oligodendrocytes (b, Th1, p<0.01; Th2, p<0.001). Effects of OKT3-stimulated T cell media were weak and reached significance over autologous serum-containing control medium only in the case of CNPase⁺ cells (b, Th2, p<0.05), although interestingly, CNPase⁺ oligodendrocytes comprised a similar percentage of total cells in OPC cultures exposed to both GA-reactive and OKT3-stimulated samples (c). (a-c) *p<0.05, **p<0.01, ***p<0.001, ANOVA plus Bonferroni post test. Data shown are representative of at least three experiments using independent cultures.

(a) First-strand cDNA from Th1- or Th2-polarized GA-reactive T cell lines or unpolarized controls was labeled with α-33P and hybridized to a nylon filter microarray containing 1151 cDNAs. The pixel density of each 33P-labeled cDNA element was determined and raw intensity data normalized for each element by Z transformation. Gene expression raw data, log values, and Z scores were averaged using mean ± SD and compared by regression analysis and Bonferroni’s correction for multiple comparisons. Z score data indicated that GA-reactive lines express multiple factors implicated as regulators of the growth and/or differentiation of oligodendrocyte lineage cells, including factors not previously identified as expressed by GA-reactive cells. Transcripts induced in Th2-polarized lines included IGF-2, CXCL1, FGF1, EDN1, LIF and BMP4, while transcripts induced in Th1-polarized lines included FST, CXCL3 and BMP7. Statistical analysis showed that induction of IGF-2 was strongly significant (p value 9.24 E-10) in Th2-polarized lines, while induction of BMP7 reached significance in Th1-polarized lines (p value 0.0002). Other factors did not reach statistical significance. (b) To determine whether induction of IGF-2 was reflected in changes in IGF-2 protein, supernatants from GA-reactive Th1- and Th2-polarized cells and 5% autologous serum-containing and serum-free control media were analyzed by ELISA. IGF-2 was detected in medium containing autologous serum at concentrations of approximately 830 pg/ml, and concentrations were further increased in supernatants from Th1- and Th2-polarized GA-reactive lymphocytes. In the case of Th2-polarized cells this increase reached significance (*p<0.05, ANOVA plus Bonferroni post test, 1150 pg/ml). IGF-2 was not detected in serum-free control medium, nor in OPC differentiation medium or OPC-conditioned medium (data not shown). (c) Supernatants from GA-reactive Th2-polarized lines were preincubated with 50 μg/ml anti-IGF-2 blocking antibody or IgG control at room temperature 1h, then added to purified OPC cultures as above. After 5d, cultures exposed to supernatants preincubated with anti-IGF-2 antibody displayed a 30% reduction in numbers of both total cells (DAPI⁺) and Olig2⁺ oligodendrocyte lineage cells compared with controls, and in both cases these changes were significant (DAPI⁺ **p<0.01, Olig2⁺ ***p<0.001). In the absence of IGF-2-containing supernatants, anti-IGF-2 antibody had no effect on OPC numbers or maturation (data not shown). (b,c) *p<0.05, **p<0.01, ***p<0.001, ANOVA plus Bonferroni post test (b), Student’s t test (c). Data shown are representative of at least three experiments using independent cultures.

Primary OPC cultures were purified from P1 rat cortices and plated in differentiation medium (no growth factors) or proliferation medium (+10 ng/ml PDGF-AA +10 ng/ml FGF-2) containing 33% supernatant from GA-reactive Th1- or Th2-polarized T cell lines, or matched autologous serum-containing medium control. (a) After 5d, cultures contained Olig2⁺ oligodendrocyte lineage cells and GFAP⁺ astrocytes. Olig2⁺ cells could be subdivided into bipolar A2B5⁺ OPCs, O4⁺ oligodendrocytes, and mature arborized CNPase⁺ cells. (b-f) Cultures in basic medium exposed to GA-reactive Th2-polarized cell supernatants for 5d contained more total cells than controls (b,c, p<0.001, ANOVA plus Bonferroni post test). This effect was associated with significant increases in the absolute numbers and percentage of Olig2⁺ cells (b,d, p<0.01),⁺ oligodendrocytes (b,e, p<0.05), and decreased numbers and percentage of GFAP⁺ astrocytes (f, p<0.01). Effects of Th1-polarized lines also reached significance in all cases (c-f, P values shown). (g) In OPC cultures grown in proliferation medium, GA-reactive T cell supernatants had weaker effects, which reached significance only in terms of total cell number, and only for Th2-polarized cells (p<0.05). (c-g) *p<0.05, **p<0.01, ***p<0.001. Scalebars, (a) 20 μm, (b) μm. 50 Data are representative of at least three experiments in independent cultures.

Primary human oligodendrocyte-enriched cultures prepared from spinal cord samples were plated into differentiation medium containing 33% supernatant from Th1- or Th2-polarized GA-reactive T lymphocytes or autologous serum-containing medium control. (a,b) Cultures exposed to GA-reactive Th2-polarized cell supernatants for 5d displayed a significant increase in total cell number (**p<0.01). (a,c) This effect was also associated with an increase in both the number and percentage of Olig2⁺ oligodendrocyte lineage cells (**p<0.01). Supernatants from Th1-polarized lines had weaker effects, which reached significance only in terms of an increase in the percentage of Olig2⁺ cells (c, *p<0.05). (b,c) *p<0.05, **p<0.01, ANOVA plus Bonferroni post test. Scale bar, (a), 30 μm. Data are representative of at least three experiments using independent cultures.