Is immunosenescence infectious? Graham Pawelec 1 , Arne Akbar 2 , Calogero Caruso 3 , Rita Effros 4 , Beatrix Grubeck-Loebenstein 5 and Anders Wikby 6 1 University of Tu ¨ bingen Medical School, Center for Medical Research, ZMF, Waldho ¨ rnlestr 22, D-72072 Tu ¨ bingen, Germany 2 Department of Immunology and Molecular Pathology, The Windeyer Institute For Medical Sciences, Royal Free and University College Medical School, 46 Cleveland Street, London W1T 4JF, UK 3 Universita ` di Palermo, Dipartimento di Biopatologia e Metedologie Biomediche, Gruppo di Studio sull’Immunosenescenza, Corso Tukory 211, 90134 Palermo, Italy 4 David Geffen School of Medicine at UCLA, Department of Pathology and Laboratory Medicine, 10833 Le Conte Avenue, Los Angeles, CA 90095-1732, USA 5 Austrian Academy of Sciences, Institute for Biomedical Ageing Research, Immunology Division, Rennweg 10, A-6020 Innsbruck, Austria 6 School of Health Sciences, Jo ¨ nko ¨ ping University, Department of Natural Science and Biomedicine, P.O. Box 2038, Banarpsgatan 39, SE-551 11 Jo ¨ nko ¨ ping, Sweden Herpes viruses are endemic. Once established, the virus is never eliminated but persists throughout life. The fraction of infected individuals gradually increases with age, such that the majority of elderly people are cyto- megalovirus (CMV) 1 , Epstein – Barr virus (EBV) 1 and Varicella 1 . Clinically relevant reactivation of Varicella causes painful shingles; CMV reactivation can cause fatal pneumonia. Overt reactivation, even in the very elderly, occurs only in immunocompromised individ- uals; however, the necessity for maintaining immunity to these viruses is costly. We argue that this cost is not only reflected in the requirement for continuous immu- nosurveillance against these viruses but, more impor- tantly, results in a re-configuration of T-cell immunity due to the accumulation of dysfunctional virus-specific cells, which fail to be eliminated from the system. Thus, we hypothesize that it is the chronic antigenic stimu- lation by CMV (and possibly other persisting antigens) that leads to an increasing prevalence of senescent, dysfunctional T cells, and therefore contributes to more general alterations in the immune system, which are associated with earlier mortality. Maintenance of protective immunity against cytomegalo- virus (CMV) is clearly essential, which is graphically illustrated by earlier experiences in bone marrow trans- plantation (reviewed in Ref. [1]). Nonetheless, early reports of viral reactivation suggest that immunity is a continuous battle even in normal healthy persons [2]. At that time, the composition of the different T-cell subsets was just being identified, and, concurrently, reports were appearing that CMV infection could markedly alter components of those subsets [3]. One of the early studies showed an increased number of CD8 þ cells in normal healthy donors and even identified expansions of CD8 þ CD57 þ (HNK-1 þ ) subsets in CMV seropositive individuals. It was over a decade later that differences between young and old donors were assessed in the context of CMV status [4]. That study reported that CMV seropositivity was associated with an increased number of both CD4 þ and CD8 þ cells, which were CD28 2 . Importantly, the authors pointed out that this phenotype had previously been associated with age; however, they found that it was primarily associated with CMV status and only secondarily with age, given the increasing frequency of CMV-infection with age. However, both age and CMV status influenced the number of CD8 þ cells and their expression of CD45RA and CD28. Role of CMV in determining the ‘immune risk phenotype’ Age-associated changes in the immune system have been extensively documented over the years, without reference to CMV status (Box 1). However, two different approaches have recently begun to shed more light on the unexpected way in which CMV infection shapes the ageing human immune system. The first is the development of tetramer technology, enabling direct identification of T cells carry- ing receptors for single peptide epitopes. The second is the Box 1. Alterations in the T-cell compartment with age " CD45RO þ cells (reviewed in Ref. [41]) " CD95 þ cells (reviewed in Ref. [41]) # CD28 expression (reviewed in Ref. [41]) " CD152 expression (reviewed in Ref. [41]) " killer cell lectin-like receptor G1 (KLRG-1) expression [15] # apoptosis of CD8 cells (reviewed in Ref. [41]) " apoptosis of CD4 cells (reviewed in Ref. [41]) # interferon-g (IFN-g) production; meta-analysis [42] a # ? interleukin-2 (IL 2) production; meta-analysis [42] b # telomere lengths (reviewed in Ref. [41]) # telomerase induction (reviewed in Ref. [41]) " DNA damage (reviewed in Ref. [41]) # DNA repair (reviewed in Ref. [41]) # stress resistance and heat-shock protein (HSP) expression (reviewed in Ref. [41]) a Data from meta-analysis: of 23 studies, 11 reported decreased IFN-g production, 8 no change and 4 an increase. b Data from meta-analysis: of 28 studies, 14 reported decreased IL 2 production, 3 no change and 11 an increase. Corresponding author: Graham Pawelec (graham.pawelec@uni-tuebingen.de). Available online 2 June 2004 Opinion TRENDS in Immunology Vol.25 No.8 August 2004 www.sciencedirect.com 1471-4906/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.it.2004.05.006