Retraction of an interpretation.

WE WRITE TO RETRACT AN INTERPRETATION IN our Report, “Contribution of human αdefensin 1, 2, and 3 to the anti–HIV-1 activity of CD8 antiviral factor” (1), wherein we demonstrated that human α-defensin 1, 2, and 3 account for much of the anti–HIV-1 activity of the CD8 antiviral factor (CAF) that is not attributable to β-chemokines. Although the antiviral activity of human αdefensin has not been called into question, the cellular source of these α-defensins has been reinterpretated in light of more recent experiments. Our experiments were done using purified CD8 T cells from long-term nonprogressors or normal persons that were stimulated with anti-CD3 and anti-CD28 antibodies, recombinant interleukin-2, phytohemagglutinin, and irradiated allogeneic peripheral blood mononuclear cells (PBMC). This method of stimulating CD8 T cells had been commonly used by many groups working on CAF (2–7). However, in a follow-up attempt to define the specific subpopulation of CD8 T cells that produce αdefensins, we have found that in the absence of allogeneic irradiated PBMC (feeders), stimulated CD8 T cell supernatants do not contain α-defensins. Although it could be argued that an allogeneic stimulus is a prerequisite for α-defensin production by CD8 T cells, it is more likely that they are derived from a cell population residing within allogeneic feeders. To pursue the exact source of αdefensins in allogeneic feeders, we positively selected individual cell populations from irradiated PBMC and subjected them to the aforementioned stimulation conditions in the absence of allogeneic exposure. α-Defensins were detected in the supernatants of CD4 and CD8 T cells and CD19 B cells. However, if the allogeneic PBMC feeders were first treated with an antiCD15 monoclonal antibody to eliminate residual neutrophils before being subjected to irradiation, then α-defensins were no longer detectable in the supernatants of stimulated T or B cells. These findings suggest that under our experimental conditions, even minor degrees of neutrophil contamination could result in the detection of α-defensins in the culture supernatant of other cell populations. In a different set of experiments using antibody staining in flow cytometry or immunofluorescence, we also detected αdefensins within several freshly isolated mononuclear cell populations, including CD8 T cells, as we have reported (1). But, in stark contrast, no α-defensin mRNA could be detected in CD8 T cells using a sensitive RT-PCR assay. This discrepancy prompted a series of in vitro cell-mixing experiments, which led us to the conclusion that during the steps of cell processing, fixation, and permeabilization, α-defensins readily leaked from neutrophils into other cells that are not natively producing these proteins. In preliminary experiments, we also find that the release of α-defensins into the supernatant becomes more striking when activated CD8 T cells are present, perhaps because they release cytokines that facilitate degranulation of α-defensins from contaminating neutrophils. Taken together, these new findings convinced us that αdefensins cannot account for the CAF activity in experimental systems that do not use allogeneic feeders (8–11). We wish to emphasize that the experimental findings in our Report (1) are repeatable. In particular, we have solidified the results shown in figs. 3 and 4 of our paper (1). The removal or neutralization of α-defensins by specific antibodies again resulted in the loss of anti–HIV-1 activity in the supernatant of CD8 T cells stimulated by allogeneic feeders. It should be pointed out that there is, in fact, very little residual anti–HIV-1 activity remaining once α-defensins and β-chemokines are eliminated. More importantly, we have also tested human neutrophil-derived α-defensins from independent sources and found their antiviral potency to be equivalent to those previously described (1), irrespective of the viral strain or target cell used in the experiment. Other investigators have confirmed this observation (12, 13). Thus, the anti–HIV-1 activity of αdefensins we have described (1) is not in doubt, and the mechanism of their antiviral effect should be pursued. LINQI ZHANG, PETER LOPEZ, TIAN HE, WENJIE YU, DAVID D. HO Aaron Diamond AIDS Research Center, The Rockefeller University, 455 First Avenue, New York, NY 10016, USA. References 1. L. Zhang et al., Science 298, 995 (2002). 2. T. W. Chun et al., Proc. Natl. Acad. Sci. U.S.A. 98, 253 (2001). 3. R. Geiben-Lynn et al., J. Virol. 75, 8306 (2001). 4. S. Le Borgne, M. Fevrier, C. Callebaut, S. P. Lee, Y. Riviere, J. Virol. 74, 4456 (2000). 5. J. F. Toso et al., J. Infect. Dis. 172, 964 (1995). 6. O. O. Yang et al., J. Virol. 71, 3120 (1997). 7. Y. Cao, L. Qin, L. Zhang, J. Safrit, D. D. Ho, N. Engl. J. Med. 332, 201 (1995). 8. H. Moriuchi, M. Moriuchi, C. Combadiere, P. M. Murphy, A. S. Fauci, Proc. Natl. Acad. Sci. U.S.A. 93, 15341 (1996). 9. C. M.Walker, D. J. Moody, D. P. Stites, J.A. Levy, Science 234, 1563 (1986). 10. A. Mosoian, A. Teixeira, E. Caron, J. Piwoz, M. E. Klotman, Viral Immunol. 13, 481 (2000). 11. G. D. Tomaras et al., Proc. Natl. Acad. Sci. U.S.A. 97, 3503 (2000). 12. T. L. Chang, F. Francois, A. Mosoian, M. E. Klotman, J. Virol. 77, 6777 (2003). 13. C. E. Mackewicz et al., AIDS 17, F23 (2003).