We developed a method for selectively propagating disease-related autoreactive T-cell lines (auTCLs) based on their increased resistance to apoptosis. The generated auTCLs homogeneously co-express CD45RO and CD49a, adhere strongly to extracellular matrix proteins and express high interleukin-17 (IL-17) messenger RNA levels, resembling a T-cell subset proposed to transmigrate into tissues and induce systemic and local inflammation in rheumatoid arthritis. The combinations of T-cell oligoclones that comprise probable multiple sclerosis (pMS) disease-related lines use a unique portfolio of T-cell receptor beta-chain variable allele (BV genes) combinations forming 'disease-specific cluster patterns'. The auTCL derived from different patients and from different myelin epitopes display striking similarities in BV gene allele clusters and are derived primarily from a disease-prone hotspot residing in the BV gene locus between Vbeta6 and Vbeta9. Conversely, healthy subject TCLs use different BV gene allele sets, forming 'healthy responder usage formats'. These formats were absent from the pMS patient V-beta gene allele combinations evaluated in this study. Hierarchical clustering of the BV gene combinations, distinguish three pMS auTCL groups, implying existence of up to three disease-related immune response patterns. These subgroup patterns may reflect different disease subclasses or alternatively they may suggest immune reactivity to different aetiological agents. Analyses of clonal-clustering patterns may potentially aid in subclassification of MS or in characterizing aetiological agents of this disease.