VIEWPOINT IMMUNOLOGY TODAY 0167-5699/99/$ – see front matter © 1999 Elsevier Science. All rights reserved. 258 Vol.20 No.6 JUNE 1999 PII: S0167-5699(99)01471-1 cells from human immuno- deficiency virus type 1 (HIV-1)- infected individuals are particu- larly susceptible to cell death. This occurs when cells are cultured without mitogenic stimuli, but is markedly increased after stimulation through the T-cell receptor (TCR). A number of explanations have been proposed to account for these phenomena, including interactions between virus/viral proteins and TCR/CD4 molecules, lack of appropriate costimulary signals and dysregu- lation of death regulatory proteins (such as Bcl2 and CD95) 1–7 . Apoptotic and necrotic mechanisms of cell death coexist in HIV infection It is commonly believed that cell death occurs from apoptosis, but in reality there are multiple pathways leading to cell death in T cells from HIV-1-infected individuals, and a substantial proportion of T-cell mortality does not appear to occur via classical apoptosis. Apoptotic cells are certainly observed after culture of T cells in the absence of interleukin-2 (IL-2), due to growth factor withdrawal or CD95–CD95 ligand (CD95L) interactions 3–5 . Nevertheless, mitogeni- cally stimulated cells do not appear to die by this mechanism 8–11 . In this article we describe a novel mechanism of activation-induced cell death, that has many features of necrosis 8–11 , a phenomenon referred to as activation-associated necrosis (AAN). Activation-associated necrosis When CD4 + and CD8 + T cells from HIV-1-infected patients are stimu- lated in vitro with phytohaemagglutinin (PHA) or anti-CD3 mono- clonal antibody (mAb), there is an early wave of apoptotic death that occurs within the first 24 h. Later, between 48 and 72 h of culture, further cells die 8–11 , but the cell loss can be attributed to an impaired ability of T cells from HIV-1 infected patients to synthesize nu- cleotides through the de novo pathway 10,11 . It appears that the resul- tant ATP depletion and associated inability to synthesize new protein or complete the cell cycle, leads eventually to necrotic cell death 10,11 Un- like activation-induced cell death (AICD), which is mediated through the CD95–CD95L pathway, cells destined to die via AAN show an initial upregulation of CD25 and major histocompatibility complex (MHC) class II, and commence synthesizing DNA (as measured by bromodeoxyuridine uptake) but die before completing the cell cycle 8,9 . When studied by light- and electron mi- croscopy, apoptotic features such as chromatin condensation or membrane flocculation are absent. Rather, the cells display clear features of necrosis, including membrane leakage and diffused necrotic nuclei (Fig. 1). The percent- age of cells with fragmented DNA (5%, measured by the TUNEL assay) is the same as seen in cultures from HIV-1 - controls. How- ever, all T cells from HIV-1 + individuals, retain the capacity to undergo apoptosis within 24 h of treatment with apoptosis-inducing agents such as okadaic acid 11 (Fig. 1d). The process of AAN is intrinsic to T cells from infected individuals and cannot be trans- ferred from cell to cell in short-term cultures. Thus, T cells from HIV - individuals prolifer- ate normally when cocultured with dysfunctional cells from an HIV-1 + patient 8 . The degree of cell death by AAN is variable, from 10 to 95%, and, although the percentage of cells dying by AAN correlates weakly with disease stage, can be detected in individuals with well-preserved CD4 counts 9 . Furthermore, we found that in a cohort of 435 HIV-1- infected patients high levels of AAN could predict AIDS and long-term risk of death independently of the CD4 counts (M. Bofill, unpublished). Although the CD8 + CD45R0 + population is the most sensitive to AAN, all T-cell subsets, including CD4 + CD45RA + naive cells, are sus- ceptible 9 . We have found that entry into AAN cannot be reversed or prevented by supplementing cultures with lymphokines or fibroblast- derived factors that prevent spontaneous apoptosis, or by addition of neutralizing antibodies to tumor necrosis factors  or  (TNF- or -) or interferon  (IFN-). Furthermore, neither blocking of CD95–CD95L interactions nor the use of the caspase inhibitor ZVAD could prevent cell death, confirming that AAN is not an apoptotic process 11 . Impaired nucleotide metabolism and defective proliferation in HIV-1 infection The ability to synthesize purine and pyrimidine ribonucleotides de novo is essential to all dividing cells, particularly those of the immune system (Fig. 2). Just how essential intact de novo synthetic routes of ribonucleotide formation are for proliferation in PHA-stimulated T cells from healthy individuals has been demonstrated using inhibitors of both purine and pyrimidine biosynthesis 12 . Thus, T cells cultured in the presence of azaserine fail to proliferate in response to 72 h stimu- lation with PHA, compared with uninhibited control cultures. The inhibition is associated with the failure of ATP, GTP, UTP, CTP and Novel mechanism for the impairment of cell proliferation in HIV-1 infection Margarita Bofill, Nicola J. Borthwick and H. Anne Simmonds The synthesis of ribonucleotides is essential to cell proliferation. Defects in the relevant metabolic pathways have been demonstrated in stimulated T cells from AIDS patients and are associated with lymphocyte necrotic death. Here, Margarita Bofill and colleagues discuss the possibility that an impaired ribonucleotide metabolism might be common to all rapidly dividing cells and thus contribute to other recognized symptoms of HIV-1 infection. T