EDITORIALS Endogenous Retroviruses and Multiple Sclerosis Human endogenous retroviruses (HERVs) are DNA sequences (retroelements) present throughout human chromosomes that by some estimates comprise up to 7% percent of the genome.1 Given their characteristic retroviral organization, with long terminal repeats (LTRs) flanking gag, pol, and env genes, it has been proposed that they are derived from ancestral infectious agents long integrated into human DNA. All identified HERVs—and there are hundreds—are defective; that is, they do not encode infectious particles, are not known to be transmissible, and frequently do not even encode functional proteins. Nevertheless, many are transcriptionally active. Additionally, some HERVs are associated with particle formation, and there is a theoretical possibility that HERV genomes can be transcomplemented, resulting in full-fledged virion formation. Their persistence in the primate and human genome—some authorities place the introduction of primate ERVs back millions of years2—argues for a strong selective pressure against a generally pathogenic phenotype. HERVs have been categorized in several ways, and one frequently used scheme divides HERVs according to the transfer RNA (tRNA) that is potentially used to prime the synthesis of a DNA copy of the (presumed) ancestral RNA genome.1 For example, HERV-W contains a primer binding site that binds the tRNA for tryptophan (single-letter amino acid code W); HERV-K is potentially primed by lysine tRNA; and so forth. Their persistence in the human genome has led some investigators to search for functional roles for HERVs. Perhaps the most compelling evidence for their involvement in a normal cellular function is the expression of HERV-W envelope (termed syncytin) in normal placental syncytiotrophoblasts.3 Mi and colleagues3 demonstrated that high levels of syncytin in placental cytotrophoblasts mediate fusion of the cytotrophoblasts, suggesting a role for this protein in normal human placental development. Moreover, syncytin was not found in any other tissues, with the exception of testis (from which it was originally cloned). To our knowledge, there is no other evidence for positive selection for HERVs. Similarly, there is no conclusive proof that any HERVs play a role in human disease, although investigators have posited several possible pathogenetic mechanisms. For example, HERVs could enhance the transcription of cellular genes placed downstream of the HERV long terminal repeats (LTRs). The LTRs contain the promoter elements and associated regulatory sequences that interact with transcriptional factors; these are themselves often regulated by various inflammatory cytokines and other cellular proteins.4 Alternatively, HERVs may encode superantigens that result in enhanced inflammatory responses, or they could mimic self-antigens and subsequently lead to the development of autoimmune phenomena such as those associated with systemic lupus erythematosus or multiple sclerosis (MS). Retroviruses have by no means escaped the long list of potential pathogens associated with MS by different investigative groups. For example, a decade ago, human T lymphotropic virus type 1 (HTLV-1) was implicated by polymerase chain reaction amplification from patients with MS.5 Perron and colleagues first proposed the importance of an MS-associated retrovirus that formed morphologically appropriate particles.6 The MS-associated retrovirus was eventually found to be related to an endogenous retroelement, ERV-9, which led to the discovery of the family of HERV-W retroelements.7,8 The increased expression of HERV RNA had been suggested to play a role in multiple sclerosis9,10 as well as schizophrenia,11 without clear causality. The article by Johnston and colleagues in this issue of the Annals of Neurology (pp. 434 – 442) presents clear evidence for increased expression of certain HERV sequences, specifically HERV-W and HERV-K, in the brains of individuals with MS or human immunodeficiency virus (HIV)-1 infection. It further concludes that increased HERV expression is not associated with specific disease pathology. Using both an in vitro model system and pathological analyses of autopsied brain, the authors conclude that the observed change in HERV expression is likely to be the result, and not the cause, of inflammatory diseases within the brain, by showing an association between monocyte/ macrophage activation in vitro and induced expression of certain HERV RNAs. MS and AIDS are two conditions associated with macrophage activation, tumor necrosis factor-␣ (TNF-␣) expression, and transendothelial migration of monocytes and lymphocytes into the central nervous system (CNS); these events could trigger the expression of HERVs, as suggested by Johnston and colleagues.12 Because the increased expression occurs in conditions with such varied cause and pathology, it is unlikely that specific pathological changes or disease entities are associated with the enhanced transcription of these retroelements. The authors appropriately indicate that this study does not preclude a more generic role for HERVs in inflammatory CNS disease, but rather argue against a specific cause-effect relationship. Additionally, the failure to detect significant differences in HERV expression between control brains and those from patients with Alz- © 2001 Wiley-Liss, Inc. 429 heimer’s disease argues against a role for HERVs in directly inducing neurodegeneration. Nonetheless, unforeseen effects of HERV transcriptional activation by inflammatory factors in the CNS cannot be discounted, and further studies in other neurodegenerative diseases may demonstrate other consequences of the presence of these retroelements. These experiments extend previous reports of HERV detection in cells of patients with MS by the direct quantification of HERV transcripts in brain tissue, and by demonstrating a possible mechanism for induction of HERV-W overexpression. Macrophage activation, which occurs in brain tissue in many inflammatory CNS diseases such as multiple sclerosis and AIDS, is linked directly to HERV transcriptional induction. This conclusion is supported by this work as well as by previous studies showing enhanced HERV expression in vascular endothelial cells stimulated with TNF-␣ and other proinflammatory cytokines.4 Johnston and associates further contend that the state of monocyte differentiation also affects HERV expression in cells of the macrophage lineage. While the effect on HERV expression noted after induction with phorbol esters is convincing, the relationship between stage of differentiation and HERV expression is only partially proven. The U937 cells, which are transformed and monocytoid in origin, are not ideal surrogates for differentiating monocytes/macrophages but were used in these studies because of the inherent variability of HERV expression in primary monocytes from different donors. Nonetheless, a comparison of HERV expression patterns in undifferentiated monocytes with the pattern in fully differentiated monocytederived macrophages from the same donor would provide additional evidence of the importance of the differentiation program in retroelement expression. The significance of the presence of HERV transcripts in the human brain will likely remain contentious. HERVs are considered to be essentially nonfunctional components of the human genome, although their mere presence suggests the potential for adaptive or maladaptive influences on host cell functions under conditions of induced expression.11 HERV gene expression may be regulated by mechanisms similar to those affecting other retroviruses, such as HIV. HERVs demonstrate at least partial sequence similarity to HIV, with a characteristic LTR-gag-pol-env-LTR organization. As such, HERVs may be similarly subject to transcriptional regulation, suggesting a transactivation phenomenon resulting in increased HERV transcription under inflammatory conditions. Consistent with this, Johnston and colleagues found elevations in HERV RNAs in brain tissues showing concomitant elevations of TNF-␣, and in cultured macrophages stimulated with phorbol esters and lipopolysachharide. In summary, this interesting study provides strong 430 Annals of Neurology Vol 50 No 4 October 2001 evidence that the HERVs found in MS are a byproduct of the inflammatory component of this disease, putting to rest thoughts that they could be the exogenous pathogen that has been the “holy grail” of studies on the pathogenesis of MS. At the same time, the experiments raise the possibility that the increased expression accompanying the inflammatory infiltrate could end up by itself contributing to the pathology, albeit in a nonspecific way. Dennis L. Kolson, MD, PhD, and Francisco González-Scarano, MD Department of Neurology University of Pennsylvania Medical Center Philadelphia, PA References 1. Bock M, Stoye JP. Endogenous retroviruses and the human germline. Curr Opin Genet Dev 2000;10:651– 655. 2. Barbulescu M, Turner G, Seaman MI, et al. Many human endogenous retrovirus K (HERV-K) proviruses are unique to humans. Curr Biol 1999;9:861– 868. 3. Mi S, Lee X, Li X-p, et al. Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis. Nature 2000;403:785–789. 4. Katsumata K, Ikeda H, Sato M, et al. Cytokine regulation of env gene expression of human endogenous retrovirus-R in human vascular endothelial cells. Clin Immunol 1999;93:75– 80. 5. Reddy EP, Sandberg-Wollheim M, Mettus RV, et al. Amplification and molecular cloning of HTLV-I sequences from DNA of multiple sclerosis patients. Science 1989;243:529 –533. 6. Perron H, Gratacap B, Lalande B, et al. In vitro transmission and antigenicity of a retrovirus isolated from a multiple sclerosis patient. Res Virol 1992;143:337–350. 7. Blond JL, Beseme F, Duret L, et al. Molecular characterization and placental expression of HERV-W, a new human endogenous retrovirus family. J Virol 1999;73:1175–1185. 8. Fujinami RS, Libbey JE. Endogenous retroviruses: are they the cause of multiple sclerosis? Trends Microbiol 1999;7:263–264. 9. Christensen T, Dissing Sorensen P, Riemann H, et al. Molecular characterization of HERV-H variants associated with multiple sclerosis. Acta Neurol Scand 2000;101:229 –238. 10. Rasmussen HB, Lucotte G, Clausen J. Endogenous retroviruses and multiple sclerosis. J Neurovirol 2000;6:S80 –S84. 11. Karlsson H, Bachmann S, Schroder J, et al. Retroviral RNA identified in the cerebropsinal fluids and brains of individuals with schizophrenia. Proc Natl Acad Sci U S A 2001;98: 4634 – 4639. 12. Johnston JB, Silva C, Holden J, et al. Monocyte activation and differentiation augment human endogenous retrovirus expression: implications for inflammatory brain diseases. Ann Neurol 2001; 50:434 – 442.