When the End Is the Story

by Welkin John­son

Some­times, dis­cov­ery in bi­ol­ogy is about dis­cern­ing rules and some­times it is about pur­su­ing ex­cep­tions. In this spirit, Hu­man Her­pesvirus six (HHV‑6), the eti­o­logic agent of the com­mon child­hood ill­ness rose­ola in­fan­tum, is shap­ing up to be an in­trigu­ing ex­cep­tion. As every vi­rol­o­gist knows, mem­bers of the Her­pesviri­dae main­tain their large dou­ble-stranded DNA genomes (typ­i­cally 100–250kb) as au­tonomous, co­va­lently closed cir­cles (epi­somes) dur­ing la­tent in­fec­tion of host tis­sues. Nev­er­the­less, there is now con­vinc­ing ev­i­dence that the HHV‑6 genome can, at least on oc­ca­sion, be­come in­te­grated into host-cell chro­mo­somes. In­ter­est­ingly, the first hints that this could hap­pen did not come from hy­poth­e­sis-dri­ven lab­o­ra­tory ex­per­i­ments, but from a hand­ful of clin­i­cal case re­ports of in­di­vid­u­als with ex­cep­tion­ally high lev­els of HHV‑6 DNA in pe­riph­eral blood.

It looks like a her­pesvirus, but does it repli­cate like one? Elec­tron­mi­cro­graphs show­ing ma­ture HHV‑6 par­ti­cles emerg­ing from an in­fected cell. Source

HHV‑6 was dis­cov­ered in 1986, and its sib­ling, HHV‑7, in 1989. In 1988, a link be­tween HHV‑6 and rose­ola in­fan­tum was es­tab­lished. To­gether HHV‑6 and HHV‑7 are now cat­e­go­rized as Rose­oloviruses, con­sti­tut­ing their own genus within the fam­ily Her­pesviri­dae. As with the other hu­man her­pesviruses, HHV‑6 and HHV‑7 in­fec­tions are wide­spread among hu­mans, but are gen­er­ally not as­so­ci­ated with se­vere patho­gen­e­sis ex­cept un­der con­di­tions such as ac­quired or in­duced im­mun­od­e­fi­ciency. Of the eight known hu­man her­pesviruses (of­fi­cially re­ferred to as HHV‑1 through HHV‑8), most of us have heard of the Her­pes sim­plex viruses (HHV‑1 and HHV‑2), the agents be­hind cold sores and gen­i­tal sores, re­spec­tively, and we are fa­mil­iar with Vari­cella Zoster virus (HHV‑3), the cause of chick­en­pox, and Ep­stein-Barr virus (HHV‑4), the cause of "kiss­ing-dis­ease" (in­fec­tious mononu­cle­o­sis).

In con­trast to the vo­lu­mi­nous re­search ac­corded to their more no­to­ri­ous rel­a­tives, the sci­en­tific lit­er­a­ture on the Rose­oloviruses is scant. Nonethe­less, a quick search of this lit­er­a­ture turns up some­thing peculiar—a smat­ter­ing of case re­ports de­scrib­ing fam­i­lies in which HHV‑6 DNA ap­pears to be in­her­ited ver­ti­cally. Im­por­tantly, the viremic in­di­vid­u­als within such fam­i­lies in­clude one par­ent and at least one child. In some stud­ies, flu­o­res­cence in situ hy­bridiza­tion (FISH) us­ing HHV‑6 se­quences as probes re­vealed a close phys­i­cal as­so­ci­a­tion be­tween HHV‑6 DNA and hu­man chro­mo­somes in the cells of the af­flicted in­di­vid­u­als. Com­bined, this gives rise to the rather fan­tas­tic no­tion that HHV‑6 can, from time to time, find its way into an individual's germline DNA and be passed on to the next gen­er­a­tion. When this hap­pens, af­flicted in­di­vid­u­als have HHV‑6 vi­ral DNA in every nu­cle­ated cell in the body.

The phe­nom­e­non of chro­mo­so­ma­lly in­her­ited HHV‑6 is un­prece­dented in two re­spects. First, no other hu­man her­pesvirus is known to in­te­grate its DNA into the host cell chro­mo­some. (In fact, sta­ble main­te­nance of epi­so­mal DNA is widely viewed as cen­tral to the leg­endary abil­ity of her­pesviruses to main­tain life­long, per­sis­tent in­fec­tions.) Sec­ond, aside from these stud­ies, there are no re­ports in which a virus has been doc­u­mented to en­ter the germline of a mod­ern hu­man and to sub­se­quently be passed on to a child. An­i­mal genomes, in­clud­ing our own, are full of an­cient retro­vi­ral se­quences (see our post), and the smart money would have been on a retro­virus, rather than a her­pesvirus, to be the first to achieve this feat in mod­ern times.

Flu­o­res­cent in situ hy­bridiza­tion (FISH) high-light­ing telom­eric re­gions at the ends of cel­lu­lar chro­mo­somes (bright spots cor­re­spond to flu­o­res­cent, telom­ere-spe­cific DNA probes). Like cel­lu­lar chro­mo­somes, hu­man her­pes-viruses HHV‑6 and HHV‑7 also have telom­ere-like re­peats at the ends of their genomes. The func­tion of telom­eres in the vi­ral genomes re­mains a mys­tery, but re­cent data sug­gest that they may fa­cil­i­tate in­ser­tion of the vi­ral genome into host cell DNA. Source

At first glance, the genomes of HHV‑6 and HHV‑7 re­sem­ble the typ­i­cal her­pesvirus genome. They are >150 kb long, dou­ble-stranded DNA mol­e­cules brack­eted by long, di­rect re­peat re­gions; they con­tain spe­cific cis-act­ing sig­nals re­quired for DNA repli­ca­tion, and, like the typ­i­cal her­pesvirus genome, they are es­ti­mated to con­tain more than one hun­dred open read­ing frames. How­ever, a no­table pe­cu­liar­ity of the Rose­oloviruses is the pres­ence of hexa­nu­cleotide re­peats com­posed of the se­quence TTAGGG at both the 5' and 3' ends of their genomes. Why is this pe­cu­liar? Be­cause TTAGGG also hap­pens to be the se­quence of the mam­malian telom­eric re­peat, strings of which are found at the ends of every cel­lu­lar chro­mo­some. Telom­eric re­peats are added to the 3' ends of chro­mo­somes by the cel­lu­lar DNA poly­merase known as telom­erase, and are es­sen­tial for the cell to dis­tin­guish the ends of lin­ear chro­mo­somes from 3' ends gen­er­ated by DNA breaks. A func­tional role for the TTAGGG re­peats in the vi­ral repli­ca­tion cy­cle has not yet been es­tab­lished, al­though they most cer­tainly play a role, vi­ral genomes not be­ing known for their tol­er­ance of ex­tra­ne­ous, un­nec­es­sary se­quence. Telom­eric re­peats have also been seen in two other her­pesviruses: Marek's dis­ease virus (MDV) of chick­ens and Equine Her­pesvirus 2 (EHV‑2). Given the phy­lo­ge­netic dis­tance be­tween these viruses, it is likely that the pres­ence of telom­eric se­quences has evolved more than once in the his­tory of the Her­pesviri­dae.

FISH analy­sis of in­te­grated HHV‑6 DNA by E.P.Nacheva and col­leagues. One probe (green) binds to the telom­eric re­gion of a spe­cific chro­mo­some, and the other (red) to HHV‑6. The merged im­age (yel­low) shows co-lo­cal­iza­tion of the HHV‑6 DNA with the telom­eric re­gion at one end of the chro­mo­some. Source

Ar­buckle and col­leagues at the Uni­ver­sity of South­ern Florida un­der­took a de­tailed mol­e­c­u­lar study of sev­eral fam­i­lies with in­her­ited HHV‑6 and pub­lished their find­ings ear­lier this year. These in­ves­ti­ga­tors first used FISH to con­firm the pres­ence of HHV‑6 DNA as­so­ci­ated with chro­mo­somes in mul­ti­ple in­di­vid­u­als from each of four fam­i­lies with high lev­els of HHV‑6 DNA in blood. Al­though the par­tic­u­lar chro­mo­some in­volved dif­fered from one fam­ily to the next, HHV‑6 DNA was in­vari­ably found close to the end of one chro­mo­some. Within each fam­ily, HHV‑6 DNA was as­so­ci­ated with the same chro­mo­some, sup­port­ing the no­tion that the HHV‑6 se­quences in each fam­ily were be­ing ver­ti­cally in­her­ited just like any other ge­nomic lo­cus. The in­ves­ti­ga­tors also used PCR am­pli­fi­ca­tion to cap­ture and se­quence the junc­tions be­tween HHV‑6 DNA and cel­lu­lar chro­mo­so­mal DNA by em­ploy­ing primer pairs, one spe­cific for the ends of the chro­mo­some, the other for a se­quence within the vi­ral genome. Se­quenc­ing of the PCR prod­ucts re­vealed virus-host DNA junc­tions within chro­mo­so­mal telom­eric re­peats, con­firm­ing that the HHV‑6 DNA was in­deed co­va­lently in­te­grated into the host genome.

The main­te­nance of an in­te­grated copy of a her­pesvi­ral genome through at least two host gen­er­a­tions is noth­ing short of amaz­ing. The HHV‑6 genome is >150 kb long, and con­tains a wealth of genes, many of which are likely to be in­volved in vi­ral im­mune eva­sion. What are the con­se­quences of car­ry­ing vi­ral se­quences in the germline, hav­ing them present and pos­si­bly ex­pressed in cells through­out the body? Are some or all of these se­quences seen as "self" by the person's im­mune sys­tem? Does suc­cess­ful in­her­i­tance of chro­mo­so­ma­lly in­te­grated HHV‑6 pos­si­bly re­quire ad­di­tional mol­e­c­u­lar events, such as a mu­ta­tion or the in­ac­ti­va­tion of part or all of the vi­ral genome?

Over 90% of hu­mans are seropos­i­tive for HHV‑6, yet re­ports of fam­i­lies with in­her­ited, chro­mo­so­ma­lly in­te­grated HHV‑6 are few. This sug­gests that in­te­gra­tion of the HHV‑6 genome into hu­man germline DNA is not a typ­i­cal out­come of HHV‑6 in­fec­tion. There are sev­eral pos­si­ble rea­sons for this, e.g., HHV‑6 in­fec­tion of germline tis­sues may oc­cur with very low fre­quency, or the ma­jor­ity of HHV‑6 in­te­gra­tions into germline DNA may be dele­te­ri­ous and con­se­quently never passed on to a new gen­er­a­tion. Nev­er­the­less, the mere ex­is­tence of in­te­grated HHV‑6 DNA in these fam­i­lies raises a some­what hereti­cal idea: could this par­tic­u­lar her­pesvirus have evolved to use in­te­gra­tion as an es­sen­tial step in its in­fec­tious cy­cle? In other words, does in­te­gra­tion hap­pen as a mat­ter of course in the so­matic tis­sues where HHV‑6 nor­mally repli­cates?

The study by Ar­buckle and col­leagues pro­vides some clues. Us­ing pe­riph­eral blood lym­pho­cytes from fam­i­lies with in­te­grated HHV‑6 DNA, the in­ves­ti­ga­tors were able to in­duce lytic vi­ral repli­ca­tion in cul­ture. While not di­rectly prov­ing that in­te­gra­tion is es­sen­tial, this ex­per­i­ment is im­por­tant be­cause it demon­strates that in­te­grated HHV‑6 DNA is func­tion­ally ca­pa­ble of ex­press­ing prog­eny virus. In a sep­a­rate ex­per­i­ment, stan­dard cell lines (JJHAN and HEK293T) were ex­per­i­men­tally in­fected with a lab­o­ra­tory strain of HHV‑6 in cul­ture. Af­ter al­low­ing the virus to repli­cate and spread in these cells, the in­ves­ti­ga­tors were able to de­tect newly in­te­grated HHV‑6 DNA, prov­ing that in­te­gra­tion can also oc­cur here as a con­se­quence of vi­ral repli­ca­tion. Does this mean that in­te­gra­tion is the rule for HHV‑6? Her­pesviruses repli­cate their DNA in the nu­cleus, so in­te­gra­tion by ho­mol­o­gous re­com­bi­na­tion may hap­pen from time to time sim­ply as a mat­ter of chance, with­out hav­ing any bear­ing on the bi­ol­ogy of the virus. In the case of HHV‑6, ho­mol­ogy pro­vided by the telom­eric re­peats may sim­ply in­crease the prob­a­bil­ity of this tak­ing place. How­ever, the same in­ves­ti­ga­tors were un­able to de­tect epi­so­mal forms of HHV‑6 DNA in ex­per­i­men­tally in­fected cells. Al­though this con­sti­tutes a neg­a­tive re­sult, the ob­ser­va­tion is con­sis­tent with the pos­si­bil­ity that HHV‑6's strat­egy for repli­ca­tion and la­tency is dis­tinct from that of other her­pesviruses.

The pre­cise mol­e­c­u­lar events that give rise to in­her­ited HHV‑6 re­main to be de­ci­phered. Based on the pres­ence of telom­eric re­peats at the ends of the HHV‑6 genome, Ar­buckle and col­leagues fa­vor the view that ho­mol­o­gous re­com­bi­na­tion is in­volved. While plau­si­ble, one can also imag­ine al­ter­na­tive mech­a­nisms, per­haps ones em­ploy­ing vi­ral gene prod­ucts that tar­get the vi­ral genome specif­i­cally to the telom­eres. The telom­eric re­peats as­so­ciate with a com­plex of cel­lu­lar pro­teins in­volved in pro­tec­tion and repli­ca­tion of chro­mo­some ends. Is it pos­si­ble that the vi­ral re­peats also in­ter­act with the cel­lu­lar telom­eric ma­chin­ery, pos­si­bly hi­jack­ing cel­lu­lar com­plexes to meet the virus's own ends (pun in­tended)? If in­te­gra­tion turns out to lie at the core of the HHV‑6 repli­ca­tion cy­cle, how then does an in­te­grated genome em­bed­ded in a large chro­mo­some serve as the tem­plate for pro­duc­ing unit-length prog­eny vi­ral genomes? Could vi­ral genome repli­ca­tion have dele­te­ri­ous ef­fects on the phys­i­cal in­tegrity of the host cell chro­mo­some, or con­versely, might the virus reg­u­late in­te­gra­tion and repli­ca­tion in a man­ner con­sis­tent with host-cell vi­a­bil­ity?

The list of ques­tions raised by these ob­ser­va­tions is long in­deed, and should prove in­trigu­ing to vi­rol­o­gists. Such ques­tions also mean that the Rose­oloviruses are primed for a share of the lime­light hereto­fore ac­corded to their more fa­mous HHV cousins.

Ref­er­ence

Ar­buckle JH, Med­veczky MM, Luka J, Hadley SH, et al. (2010). The la­tent hu­man her­pesvirus-6A genome specif­i­cally in­te­grates in telom­eres of hu­man chro­mo­somes in vivo and in vitro. Proc Natl Acad Sci USA, 107 (12), 5563–5568. PMID 20212114

Welkin is As­sis­tant Pro­fes­sor of Mi­cro­bi­ol­ogy and Mol­e­c­u­lar Ge­net­ics at Har­vard Med­ical School, and an As­so­ciate Blog­ger for Small Things Con­sid­ered.