Gallo on HHV-6

[James Scutero]

Immunology Today Vol 16 #2   February 1995

Human herpesvirus 6 in AIDS
by Paolo Lusso and Robert C. Gallo

Multiple lines of clinical and experimental evidence suggest that
human herpesvirus 6 (HHV-6) may act as an accelerating factor in
the natural history of human immunodeficiency virus (HIV) infection.
Although, in common with HIV, HHV-6 has a primary tropism for CD4+
T cells, its potential effects on the immune system are broader.
For instance, HHV-6 can also infect and kill CD8+ T cells, natural
killer cells and mononuclear phagocytes. Here, Paolo Lusso and Robert
Gallo suggest that understanding the immunopathogenic role of 
HHV-6 in the course of HIV infection may shed new light on the complex
mechanisms of disease progression in AIDS.

During the past decade, an unprecedented scientific effort worldwide
has generated an extraordinary volume of information concerning AIDS
and its causative agent, human immunodeficiency virus (HIV) (Ref.
1). Nevertheless, our understanding of the pathogenetic mechanisms
of the severe immunodeficiency that underlies most of the clinical
manifestations of AIDS is still limited. Particularly puzzling is
the observation of wide temporal disparities in the natural history
of HIV infection among different patients, some of whom remain
asymptomatic for several years, while others rapidly progress to
full-blown AIDS. One of the most credited hypotheses to explain such
discrepancies is the selective intervention of putative accelerating
factors in patients with rapid disease progression. These factors
may act either indirectly, by potentiating the virulence and the
pathogenic effects of HIV, or directly, by attacking the immune system
in synergy with HIV itself. As suggested as early as 1988, human
herpesvirus 6 (HHV-6) is one of the most suitable candidates to play
such a catalytic role in AIDS (Refs 2,3).

A new herpesvirus discovered during the AIDS era
    Following the discovery of the Epstein-Barr virus (EBV) in 1964,
22 years elapsed before a novel human herpesvirus was identified.
The new agent was isolated from the peripheral blood of six patients
with lymphoproliferative disorders, two of whom also had AIDS (Ref.
4). Initially, the virus was propagated in vitro with difficulty
and appeared to infect B cells preferentially; hence, the original
designation 'human B-lymphotropic virus'. However, subsequent studies
established that the virus is primarily T-cell tropic(2), and only
in rare instances are B cells infected. Thus, the virus was renamed
in accordance with the taxonomy of human herpesviruses(5).
    Soon after the discovery of HHV-6, seroepidemiological studies
revealed that infection with the virus is highly endemic in the human
population and is commonly acquired in early childhood(6). However,
in recent years, the epidemiology of HHV-6 has gained new complexity
with the identification of two major viral subgroups (HHV-6A,
prototyped by strain GS; and HHV-6B, prototyped by strain Z29), which
exhibit a significant degree of biological, immunological and genetic
divergence(7-9). Although the relative prevalence of the two subgroups
is still unknown, anecdotal evidence indicates that infection by
subgroup B is far more frequent than A in the general population
and may be linked to different disease entities. However, conclusive
studies are lacking because no specific serological methods have
been developed yet, and subgroup identification relies either on
restriction endonuclease analysis or DNA sequencing.
     The first etiological association of HHV-6 with a human disease
was established in 1988, when Yamanishi and colleagues demonstrated
that primary infection with HHV-6 is linked to exanthem subitum(10)
(or roseola infantum), a benign febrile illness of early childhood.
After primary infection, similar to other herpesviruses, HHV-6 can
persist in the host in a latent form that has been detected in
circulating monocytes and in the epithelia of the bronchial and
salivary glands. Nevertheless, the main in vivo reservoir of the
virus has yet to be defined(6). Equally unclear are the events leading
to reactivation of HHV-6 from latency. However, it is conceivable
that any transient or sustained suppression of the host's immune
surveillance may lift the tight immunological restraints that control
the expression of HHV-6 in vivo. Consistent with this concept,
HHV-6 is emerging as  a potentially life-threatening opportunistic
agent in immunocompromised hosts. The spectrum of ailments putatively
associated with HHV-6 in such patients includes fever, hepatitis,
failure of bone marrow engraftment(11), encephalitis(12) and
interstitial pneumonitis(13,14). Conversely, in immunocompetent
adults, sporadic reports have linked HHV-6 to a variety of disorders,
including EBV- infectious mononucleosis, autoimmune disorders, chronic
fatigue syndrome, fulminant hepatitis, non-Hodgkin's lymphomas and
Hodgkin's disease(6); however, the pathogenic implications of these
findings remain unclear at present. It is noteworthy that most of
the viral strains identified in early childhood and in adult life
belong to subgroup B (Refs 7-9), whereas HHV-6A strains have been
recovered only in rare instances, and almost exclusively in
immunocompromised patients.

The first T-lymphotropic human herpesvirus
     The initial indication that HHV-6 could contribute to the
dramatic loss of CD4+ T cells in AIDS came from the demonstration
that HHV-6 shares with HIV a predominant tropism for CD4+ T cells,
both in vitro(2) and in vivo(15), and exerts a marked cytopathic
effect on them. This property differentiates HHV-6 from other DNA
viruses (e.g. herpes simplex virus, EBV, cytomegalovirus, hepatitis
B virus) that have been proposed as potential cofactors in AIDS.
By preferentially attacking CD4+ T cells, HHV-6 may not only play
a direct role in their progressive disappearance in the course of
HIV infection, but may also acquire the opportunity of synergizing
directly with HIV in vivo.
     Despite the propensity of HHV-6 to target CD4+ T cells, it is
now clear that its cellular host range is broader, suggesting that
a putative HHV-6 receptor is expressed on cells of different lineages.
Nevertheless, the full, lytic, virus replication cycle is efficiently
completed almost exclusively in lymphoid cells(6). Two types of
lymphoid cells that play a critical role in the defense against viral
infections in vivo-CD8+ T cells and natural (NK) cells- have recently
been found to be permissive for HHV-6 (Refs 16,17), although they
are remarkably more sensitive to HHV-6A than to HHV-6B isolates (P.
Lusso, unpublished). Interestingly, all primary CD3-CD56+ NK-cell
clones that are susceptible to HHV-6 belong to a subset of circulating
NK cells that lack the ability to kill autologous HHV-6-infected
cells, despite high cytolytic activity against K562(Refs 17,18).
By contrast, NK-cell clones with high cytolytic activity against
infected cells cannot be productively infected(17), suggesting that
active target recognition may induce resistance to HHV-6 infection
in the effector cells.
      A complex relationship exists between HHV-6 and cells of the
mononuclear phagocytic system which, along with CD4- T cells, are
another important target of HIV infection in vivo. In vitro, 
HHV-6 replicates very inefficiently in cultures of monocyte-derived
adherent macrophages (19,20), yet it exerts a cytopathic effect on
these cells and inhibits macrophage maturation in
growth-factor-stimulated bone marrow cultures(20). In vivo, HHV-6
has been detected in circulating monocytes of convalescent
children(19) and in tissue macrophages derived from patients with
HHV-6-associated pneumonitis(13). Thus mononuclear phagocytes are
certainly a target for HHV-6, and the virus may exert, directly or
indirectly, detrimental effects on their correct maturation, function
or viability.
      In summary, HHV-6 is emerging as an 'immunotropic' herpesvirus
that can directly infect, or interfere with, the function of several
elements of the immune system, including CD4+ and CD8+ T cells, NK
cells, some B cells and mononuclear phagocytes. Moreover, it has
recently been shown that HHV-6 can also target [gamma-delta] T cells
(P. Lusso et al., unpublished). The ability of HHV-6 to attack
antiviral effector cells may represent a strategy that has evolved
to counteract the host's protective mechanisms against viral
infections.

HHV-6 and HIV:a killer alliance
     The simultaneous infection of a single cell undoubtedly
represents a sine qua non requisite for the direct interplay between
different viruses. In 1989, it was demonstrated that HHV-6 and HIV-
1 can coinfect and simultaneously replicate within the same CD4-
T cell(21). The main biological consequence of such coinfection was
a dramatic acceleration in the kinetics of HIV expression and cell
death (21,22). Thus, at least in vitro, HHV-6 can synergize with
HIV in determining the destruction of CD4- T cells. Although some
investigators have subsequently reported a suppressive, rather than
enhancing, effect of HHV-6 on the replication of HIV-1(Refs 23,24),
there is agreement that coinfection of CD4- T cells by HHV-6 and
HIV results in an accelerated rate of cell death (21,23). Consistent
with this effect, HHV-6 was found to be a potent trans-activator
of the long terminal repeat (LTR) of HIV-1, HIV-2 and simian
immunodeficiency virus (SIV) (Refs 21,22,25). Using deletion mutants
of HIV-1 LTR, the HHV-6-responsive elements were mapped to the
enhancer motifs (e.g. NF-kB and Spl-binding sites) and demonstrated
to function independently from the TAR element (a sequence that is
responsive to the HIV trans-activator Tat). Indeed, an additive effect
was demonstrated in the simultaneous presence of HHV-6 and Tat(Ref.
22). Indeed, an additive effect was demonstrated in the simultaneous
presence of HHV-6 and Tat(Ref. 22).

A unique type of viral interaction
     A novel interaction between HHV-6 and HIV, which may have
important implications for the role of HHV-6 in AIDS, was recently
identified. It is well documented that HIV, unlike HHV-6 (Ref. 26),
but similar to HHV-7 (Ref. 27) (the latest human herpesvirus to be
discovered), uses the CD4 glycoprotein as a cellular membrane receptor
(28,29). HHV-6 has the unique ability to upregulate the CD4 receptor
in CD4 cells, and even to induce de novo expression of CD4 mRNA and
protein in CD4- lymphoid cells, such as NK cells(17) and mature 
CD8- T cells(30) (CD8+ T cells thereby reacquire an early, cortical
CD4+CD8+ thymic phenotype). By contrast, CD3 is transcriptionally
downregulated by HHV-6 infection(16).
     The HHV-6-induced transcriptional activation of CD4 must be
extremely powerful because the expression of the CD4 gene is tightly
repressed in post-thymic CD8+ T cells, presumably by DNA methylation,
whereas in NK cells, which follow an extrathymic differentiation
pathway, CD4 is never expressed even during [sorry, bad xerox]...
recently obtained with CD4- [gamma-delta] T cells (P. Lusso et al.,
unpublished). Intriguingly, the HHV-6-induced upregulation of CD4
renders CD8+ T cells, NK cells and [gamma-delta] T cells susceptible
to productive infection by HIV-1 (Refs 17, 30). Thus, HHV-6 may
broaden the range of cells susceptible to HIV in vivo and, thereby,
enhance the spread of HIV in coinfected hosts.

A powerful inducer of cytokine release
     Besides the direct infection of immune cells, indirect mechanisms
may enable HHV-6 to modulate the function of the immune system and
the replication of HIV. One such mechanism is the induction of the
release of cytokines, or other bioactive molecules, either by the
infected cells or by neighboring uninfected cells. HHV-6 is a powerful
inducer of inflammatory cytokines such as interferon [alpha] 
(IFN-alpha)(Ref. 31), tumor necrosis factor [alpha](TNF-alpha) and
interleukin 1 [Beta](IL-1Beta)(Ref. 32), which can influence the
expression of HIV in vitro and may play a role in the pathogenesis
of AIDS or some AIDS-associated disorders(33).
     
Ubiquitous but pathogenic:an apparent paradox
     A common argument against the concept that HHV-6 may play a
role in AIDS is its high prevalence in the human population. How
can a microbe that virtually everybody harbors from early childhood,
possibly throughout the entire lifetime, be an important cofactor
in HIV infection? The key to understanding this apparent paradox
is the fundamental two-facedness of herpesviruses, which may persist
in the host in a latent form and yet be reactivated concomitantly
with immunosuppressive events, leading to potential clinical
consequences. In particular, the reactivation of HHV-6, an
'immunotropic' herpesvirus, may dramatically amplify the immunological
damage present in HIV-infected individuals. In turn, the increasing
immunological imbalance in the host, combined with the synergistic
effects of HHV-6 and HIV, may accelerate the replication and the
spread of HIV, leading to the final destruction of the immune system.
      The main drawback related to the ubiquitous distribution of
HHV-6 is that conventional serological studies have a limited value
for establishing an association with disease. Indeed, previous
exposure to the virus, as revealed by a seropositive status, does
not imply active infection. Therefore, the use of IgG antibody testing
for the diagnosis of HHV-6 infection may be misleading, particularly
in HIV-infected patients whose antigen-specific immune responses
are progressively impaired with the development of the disease. With
this perspective, the report of an inverse correlation between
anti-HHV-6 IgG titers and progression of HIV infection (34) was not
surprising. Thus, to elucidate the pathogenic role of HHV-6, efforts
should focus on the identification and use of markers of active virus
replication in vivo.

Infection with HHV-6 is active and widespread in patients with AIDS
     Since the early days of HHV-6 research, the virus has frequently
been isolated from patients with AIDS (Ref. 6). These data were
confirmed by studies using the polymerase chain reaction (PCR) on
peripheral blood leukocytes, which demonstrated a higher prevalence
of HHV-6 in patients with AIDS than in HIV-seronegative controls(35),
as well as a significant correlation between the frequency and extent
of HHV-6 infection and the absolute number of circulating CD4+ T
cells(36), suggesting an association between HHV-6 and the
disappearance of CD4+ T cells. However, it has to be remembered that
circulating leukocytes provide only a partial, and often inaccurate,
mirror of the immune system as whole.
     Important new information has recently been derived from the
analysis of autopsy specimens, which has documented an active and
disseminated HHV-6 infection in patients with terminal AIDS. Using
PCR, the presence of HHV-6 was demonstrated in the vast majority
of the anatomical sites, including lymphoid tissue, parenchymatous
organs and the central nervous system (37). By contrast, a restricted
tissue distribution and a significantly lower virus load were observed
in HIV-seronegative controls. The wide distribution and a
significantly lower virus load were observed in HIV-seronegative
controls. The wide distribution of HHV-6 in terminal AIDS patients
was confirmed using immunohistochemical techniques, which also showed
conclusively that the infection is active(38). In one patient who
died of 'idopathic' pneumonitis, the level of viral replication in
the lungs was so extensive as to suggest an etiological role of HHV-
6 in this disorder.(38)
     Although the results obtained on autopsy specimens have certainly
strengthened the concept that HHV-6 may play a role in AIDS, their
major limitation is they do not define the time of HHV-6 reactivation
(or exogenous reinfection) during the course of HIV infection. Indeed,
several important questions remain unanswered. Is the widespread
replication of HHV-6 a primary pathogenetic event in AIDS? Conversely,
even if it is a mere epiphenomenon of the immunodeficiency, does
HHV-6 behave as a bona fide opportunistic agent, which can be
responsible for some of the clinical manifestations of AIDS? To begin
to address these questions, it is essential to identify reproducible
methods for the detection and quantitation of active HHV-6 infection
in vivo and to use them in longitudinal patient studies. In this
respect, a novel, practical test that measures the level of HHV-6
replication in vivo(i.e. the detection of cell-free viral DNA in
serum or plasma by quantitative PCR) has permitted the documentation
of active HHV-6 infection, mostly by HHV-6A strains, in early
symptomatic HIV-infected patients but not in immunocompetent
individuals.(39)

Future perspectives
      Box 1 summarizes the experimental observations that suggest
HHV-6 acts as an accerlerating factor in HIV infection. However,
although intriguing, the data accumulated thus far are insufficient
to prove definitively the role of HHV-6 in AIDS. This consideration
underscores the need to devise novel experimental and clinical
approaches. As stated above, it will be critical to perform systematic
longitudinal studies of the prevalence and replication levels of
HHV-6 in individual patients progressing through the different stages
of HIV infection. Such studies will permit the elucidation of whether
any correlation exists between active HHV-6 infection and progression
towards full-blown AIDS. Moreover, the characterization of HHV-6
strains detected in vivo will also clarify whether the two major
HHV-6 subgroups play a differential role in the course of HIV
infection, as suggested by the frequent recovery of HHV-6A strains
from immunocompromised patients(6-9).
     The use of suitable animal models has provided another promising
approach to studying the role of HHV-6 in AIDS. Unfortunately, HHV-
6, particularly subgroup A, has a restricted species host range that
includes, besides humans, only chimpanzees (Pan troglodytes) and
pig-tailed macaques (Macaca nemestrina)(6). Both species are
susceptible to infection by HIV-1, but without clinical consequences.
However, infection of pig-tailed macaques with SIV induces a clinical
immunodeficiency that resembles human AIDS. Thus, these animals
represent a suitable model system to investigate the role of 
HHV-6 in vivo.
      Finally, important insights into the role of HHV-6 in AIDS
may derive from therapeutic trials combinining anti-HHV-6 drugs with
conventional anti-retroviral therapy. For example, foscarnet, one
of the most effective drugs against HHV-6 (Ref.40), in combination
with 3'-azido-3'-deoxythymidine (AZT), was reported to prolong the
survival of HIV-infected patients significantly(41). Although
foscarnet is also an effective anti-retroviral agent, its favorable
effects on the progression of AIDS could be ascribed, at least in
part, to its anti-HHV-6 activity. The identification and use of drugs
with selective anti-HHV-6 activity will be crucial to define the
relative role played by this virus in the natural history of HIV
infection.
----------------------------------------------------------------
Box 1. Biological effects of HHV-6 on the life cycle of HIV

Biological property of HHV-6     Effect on the life cycle of HIV

Tropism for CD4+ T cells(2,15)   Direct interaction with HIV in
                                   individual cells(21), accelerated
                                   kinetics of cell death (21,23)

Cytopathic infection of CD8+     Abatement of anti-HIV cellular 
  T cells and NK cells(16,17)      immunity

Trans-activation of the HIV-1    Induction of HIV expression,
  LTR (Refs 21,22,25)              enhancement of HIV           
                                   replication(21,22)

Release of inflammatory          Immunomodulation, modulation of
  cytokines(31,32)                 HIV replication(33)

Upregulation and de novo         Broadening of the cellular
  induction of CD4                 host range of HIV (Refs 17,30)
   (Refs 17,30)
-----------------------------------------------------------------
Abbreviations: NK, natural killer; LTR, long terminal repeat.
-----------------------------------------------------------------
We thank P. Farci for critically reviewing the manuscript.

Paolo Lusso and Robert C. Gallo are at the Laboratory of Tumor Cell
Biology, National Cancer Institute, National Institutes of Health,
Bethesda, MA 20892, USA.

References
1 Gallo, R.C. and Montaigner, L.(1989) Sci. Am. 259, 41-48
2 Lusso, P.,Markham, P.D., Tschachler, E. et al. (1988) J. Exp. Med.
167, 1659-1670
3 Gallo, R.C. (1988) J. AIDS 1,521-535
4 Salahuddin, S.Z., Ablashi, D.V., Markham, P.D. et al. (1986) Science
234, 596-601
5 Ablashi, D.V., Salahuddin, S.Z., Josephs, S.F. et al. (1987) Nature
329,207
6 Lusso, P. and Gallo, R.C. in Bailliere's Clinical Hematology (Young,
N., ed.), Bailliere Tindall(in press)
7 Ablashi, D.V., Balachandran, N., Josephs, S.F. et al. (1991)
Virology 184, 545-552
8 Aubin, J.T., Collandre, H., Candotti, D. et al. (1992) J. Clin.
Microbiol. 29, 367-372
9 Schirmer, E.C., Wyatt, L.S., Yamanishi, K., Rodriguez, W.J. and
Frenkel, N. (1991) Proc. Natl Acad. Sci. USA 88, 5922-5926
10 Yamanishi, K., Okuna, T., Shiraki, K. et al.(1988) Lancet i, 1065-
1067
11 Drobyski, W.R.R., Dunne, W.M., Burd, E.M. et al. (1993) J. Infect.
Dis. 167, 735-739
12 Drobyski, W.R.R., Knox, K.K., Majewski, A. and Carrigan, D.R.(1994)
New Engl. J. Med. 330, 1356-1360
13 Carrigan, D.R., Drobyski, W.R., Russler, S.K. et al. (1991) Lancet
338, 147-149
14 Cone, R., Hackman, R.C., Huang, M.L.W. et al (1993) New Engl.
J. Med. 329, 156-161
15 Takahashi, K., Sonoda, S., Higashi, K. et al, (1989) J. Virol.
63, 3161-3165
16 Lusso, P., Malnati, M., De Maria, A., De Rocco, S., Markham, P.D.
and Gallo, R.C.(1991)J. Immunol. 147, 685-691
17 Lusso, P., Malnati, M., Garzino-Demo, A., Crowley, R.W., Long,
F.O. and Gallo, R.C. (1993) Nature 362, 458-462
18 Malnati, M.S., Lusso, P., Ciccone, F., Moretta, L. and Long, E.O.
(1993) J. Exp. Med. 178, 961-969
19 Kondo, K., Kondo, T., Okuno, T., Takahashi, M. and Yamanishi,
K.(1991)J. Gen. Virol. 72, 1401-1408
20 Carrigan, D.R. (1992) in Human Herpesvirus-6, Epidemiology.
Molecular Biology and Clinical Pathology(Ablashi, D.V., Krueger,
G.R.F. and Salahuddin, S.Z., eds), pp. 281-302, Elsevier
21 Lusso, Pl, Ensoli, B., Markham, P.D. et al. (1989) Nature 337,
368-370
22 Ensoli, B., Lusso, P.,Sehachter, F. et al. (1989) EMBO J. 8
3019-3027
23 Carrigan, D.R., Knox, K.K. and Tapper, M.A. (1990) J. Infect.
Dis. 162, 844-851
24 Levy, J.A., Landay, A. and Lennette, E.T. (1990) J. Clin.
Microbiol. 28, 2362-2364
25 Horvat, R.T., Wood, C. and Balachandran, N. (1989) J. Virol. 63,
970-973
26 Lusso, P., Gallo, R.C., De Rocco, S.E. and Markham, P.D. (1989)
Lancet i, 730
27 Lusso, P., Secchiero, P., Crowley, R.W., Garzino-Demo, A.,
Berneman, Z.N. and Gallo, R.C. (1994) Proc. Natl Acad. Sci. USA 91.
3872-3876
28 Dalgleish, A.G., Beverley, P.C.L., Clapham, P.R., Crawford, D.H.,
Greaves, M.F. and Weiss, R.A. (1984) Nature 312, 763-767
29 Klatzmann, D., Champagne, F., Chamaret, S. et al. (1984) Nature
30 Lusso, P., De Maria, Malnati, M. et al. (1991) Nature 349,
533-535
31 Kikuta, H., Nakane, A., lu, H., Taguchi, Y., Minagawa, T. and
Matsumoto,S. (1990) J. Infect Dis. 162,35-38
32 Flamand, L., Gosselin, J., D'Addario, M. et al. (1991) J. Virol
65, 5105-5110
33 Ensoli, B., Barillari, G. and Gallo, R.C.(1992) Immunol. Rev.
127,147-155
34 Spira, T.J., Bozeman, L.H., Sanderlin, K.C. et al. (1990) J.
Infect. Dis. 161,567-570
35 Buchbinder, A., Josephs, S.F., Ablashi, D.V. et al. (1988)J. Virol.
Methods 21, 133-140
36 Fairfax, M.R., Schacker, T., Cone, R.W. et al. (1994) J. Infect.
Dis. 169, 1342-1345
37 Corbellino, M., Lusso, P., Gallo, R.C., Parravicini, C., Galli,
M. And Moroni, M.(1993) Lancet 342, 1242
38 Knox, K.K. and Carrigan, D.R. (1994) Lancet 343, 577-578
39 Secchiero, P., Carrigan, D.R.,Asano, V, et al. J. Infect. Dis.
(in press)
40 Streicher, H.L., Hung, C.L., Ablashi, D.V. et al (1988)J. Virol.
Methods 21,301-304
41 Studies of Ocular Complications of AIDS Research Group, in
Collaboration with the AIDS Clinical Trials Group(1992)New Engl.
J. Med. 326,213-220
********************
-James M. Scutero