This in turn makes AZD9291 cell line it possible to identify the factors associated with the selective expansion of certain clones in vivo. We found that the chief determinants of clonal abundance were the transcriptional orientation of the provirus

and its position (upstream or downstream), relative to the nearest host transcriptional start site. Proviruses integrated within a host gene were significantly more frequent in clones of high abundance in vivo than in those with low abundance, but only when integrated in the same transcriptional sense as the host gene. Because of the known mitogenic properties of Tax, we postulated that Tax-expressing clones would reach a higher mean abundance than non-expressing clones in the circulation. But again the results confounded expectation: the frequency of Tax expression was significantly greater in low-abundance clones (Fig. 2) [80]. Although it was unexpected, this result is consistent with the observations noted above that orientation of the provirus in the same transcriptional sense as the flanking host gene is associated with silencing Crenolanib mw of Tax expression [80] and with high clone abundance [72] and [80]. Since the proviral load is higher in HAM/TSP patients than in asymptomatic HTLV-1 carriers, and oligoclonal proliferation is frequently detected more easily in samples

from patients with HAM/TSP [54], it was natural to infer that oligoclonal proliferation was stronger in HAM/TSP and therefore that it might contribute to the pathogenesis of the inflammatory disease. However, this inference could not be formally tested in the absence of an objective measure of oligoclonality. What is required is a measure of the non-uniformity or entropy of the clone frequency distribution. A widely used entropic index, the Shannon index, the is of very limited usefulness here because this index is correlated with the sample size, which can be very large in high-throughput

sequencing. We therefore defined [72] the oligoclonality index (OCI), an application of the Gini index (Fig. 3). An OCI of 1 indicates perfect monoclonality, whereas an index of 0 indicates that each clone has the same frequency. This index allows a rigorous quantitative comparison of the degree of oligoclonality between disease states. We found that, contrary to expectation, there is no significant difference in oligoclonality (as measured by OCI) between patients with HAM/TSP and asymptomatic carriers [72]; The OCI in patients with malignant disease, ATLL, is significantly higher, as expected. Further, the degree of oligoclonality (OCI) does not correlate with the proviral load in patients with non-malignant infection [72]. Rather, the proviral load correlated with the total number of distinct clones, and this number is significantly greater in patients with HAM/TSP than in asymptomatic carriers.