Tip:
Highlight text to annotate it
X
Alright, so let's move on now to the second part of the talk,
which is going to concern the discovery of the virus of KS itself.
Where did the notion that KS might have an infectious precipitate come from?
Interestingly, it didn't come from any microbiologist or virologist,
it came, rather, from clinicians and epidemiologists who were studying the natural history of KS.
And that natural history revolved around several striking facts.
Of course, when the AIDS epidemic happened and KS went from being an extraordinarily rare tumor
to a tumor that affected perhaps 20 to 25 percent of young gay men with AIDS.
Everyone initially presumed KS itself might be the cause of AIDS.
But this idea soon collapsed,
of course we know that KS occurred in Africa and the Mediterranean prior to the onset and spread of *** through the human population
and it could be shown that in classical KS, there was no evidence of *** in the lesion.
Also, it was soon shown that in the AIDS related form, there was no *** DNA or RNA in the spindle cells themselves.
Most KS biopsies didn't have any *** positive cells in them.
So, right away it became clear that although *** was a very important cofactor,
that put people at risk for KS, it probably wasn't itself the cause, the proximate cause of the tumor.
Another particularly striking feature that supported that idea
was that if you look among *** positive patients themselves,
KS was not uniformly distributed among AIDS patients. So, even among people who had contracted ***,
and had gone all the way to a severe immunodeficiency,
they were not equally at risk for developing this tumor.
The KS was largely the province of gay men with AIDS, 20 to 25 percent of whom, in the early days of the epidemic,
would develop KS. Whereas, if you looked at hemophiliacs,
who had acquired *** from blood transfusions, KS was relatively uncommon in the, 2 percent.
KS was also relatively uncommon in women compared to men,
most women acquired *** either through drug use, needle sharing, or ***, heterosexual contact
with *** positive individuals, and women had a very low incidence of KS,
even when they went on to overt immunodeficiency.
And most impressively, children who acquired *** by vertical transmission,
from the infected mom to the newborn baby, by transplacental or perinatal spread of virus,
almost never got KS, despite the fact that they had a very accelerated course and rapidly succumbed to the immunodeficiency.
So, these things all put the lie to the simple notion that *** itself was the proximate cause of KS.
And when epidemiologists looked at this data, one thing struck them.
Which is of course, the striking prevalence of infection among gay men
suggested that *** transmission of some other agent in the context of an immunodeficiency
might account for this result, particularly if the agent was more likely to be transmitted homosexually,
than by heterosexual transmission.
So, those facts swirled around the epidemiology community for a long time
before they caught the attention of microbiologists and infectious disease specialists
and virologists. Most people weren't sure what to make of these facts
and because of the strong bias that many molecular biologists have
that information from epidemiology is simply correlative and not powerful in any direct way,
many people in AIDS research discounted these results, or simply ignored them.
But two people who did not are shown here.
These are my friends Pat Moore and Yuan Chang who were then at Columbia,
now at the University of Pittsburgh. And they, as we did, took these results very seriously
and began a search for viruses in KS.
And, now, of course, simple attempts to cultivate viruses from KS didn't work
and had been tried before, so their approach was to use a very advanced genomic method
that had been recently described by Michael Wigler of Cold Spring Harbor,
a method called representational difference analysis, or RDA.
RDA is essentially a form of subtractive hybridization, a souped up form of subtractive hybridization
married to PCR amplification and it can be used to explore two different populations of cells,
the DNA of two different populations of cells, looking for segments of DNA that are present in one, but not in the other.
That's the fundamental notion behind subtractive hybridization.
And Pat and Yuan did this on KS tumors and normal cells, normal skin or other tissues from unaffected patients.
And after much to-do, they discovered from this procedure,
from a single KS tumor, two small segments of DNA of about 100 to 200 base pairs in length
which, when sequenced, which were unique to KS,
not present in the surrounding normal skin, and when sequenced, had distant evolutionary relationships
to genomes of known human herpesviruses like EBV, Epstein-Barr virus.
They were also weakly homologous to a simian virus called herpesvirus saimiri.
So, these were not EBV genomes, but they were related to EBV genomes
and they correctly surmised that must indicate the presence of a new herpesvirus
in these KS tumors and they rapidly went on to show that these small sequences
could be used as markers of KS and they were found not only in the AIDS-related form of the disease,
from which they were originally isolated, but all forms, including the classical forms, endemic forms,
whether KS was *** associated or *** unassociated from any part of the world,
KS tumors always had these DNA sequences in them.
And this result was rapidly affirmed by many other labs including our own.
So, now it's clear that these sequences define the existence of a new virus,
a new herpesvirus, a human herpesvirus.
Prior to this, there were seven known herpesviruses, now there are eight.
The formal name of the KS virus is human herpesvirus-8,
but most of us use the more informative name, KS associated *** virus
because it indicates the major disease state with which the virus is associated.
Now, when the KS genome, KSHV genome was completely sequenced, as it was by 1996,
it became clear where it stood phylogenically.
So, all of the human herpesviruses cluster into three families:
the alpha herpesviruses, the beta herpesviruses, and the gamma herpesviruses.
And these, this segregation is now defined by DNA sequences, but initially these groups were defined biologically.
Alpha herpesviruses are viruses that reside, or establish a latent infection in neuronal cells;
gamma herpesviruses are viruses that establish latency in lymphoid cells.
So, that was a huge surprise, that this virus that was derived from a predominantly endothelial neoplasm
actually sorted with the lymphotropic herpesviruses.
And for a while there was confusion about this, people wondered this was not the right virus,
just some sort of passenger virus, but I'm going to show you very compelling evidence
that it is linked to KS, but this lymphotropism set off a search that turned out to be productive.
Once it became clear that the normal reservoir of KSHV is in fact the B-cell,
students of AIDS and especially AIDS-related lymphomas
began to look for the presence of KSHV in lymphoproliferative disease.
And Ethel Cesarman and her colleagues and others rapidly established that there are two lymphoproliferative diseases that are linked to KS.
One of them is a rare lymphoma called primary effusion lymphoma.
So, some of you may know that because of their immunodeficiency state,
AIDS patients get many forms of cancer, KS being the most common,
second to that is lymphoma, predominantly B-cell lymphomas.
Many of these B-cell lymphomas turn out to be EBV related, but one of them
is KSHV related.
It's a rare tumor called primary effusion lymphoma.
Now, unlike KS, this is a classical lymphoma. This is a clonal B-cell tumor. It's an unusual B-cell tumor
because the bone marrow and lymph nodes are generally normal and the tumor arises in serosal spaces,
spaces that are outside of solid organs.
The plural space, outside of the lung, the pericardial space, outside of the heart,
the perineal space, outside of the intestinal tract.
These essentially form ascites tumors in these cases
and they occur in patients who have very advanced AIDS,
they carry a very poor prognosis. And these cells that are growing already in liquid culture in these spaces
are adaptable readily into tissue culture and grow readily in culture,
unlike KS cells. The other lesion that was found to be linked to KSHV
is a disorder called Multicentric Castleman's disease.
The initial paper on this was written by a man named Soulier, in France,
who discovered that AIDS related multicentric Castleman's disease is very tightly linked to KSHV.
These lesions have KSHV DNA in them.
Multicentric Castleman's disease, or MCD, is an interesting disease that's right, again,
at the cusp between the benign and the malignant.
It is a polyclonal process in which B-cells proliferate, patients develop very high fevers,
big spleens, lymph node involvement, they have sky-high levels of IL-6 in the peripheral blood,
so in those regards, they look like they have some sort of acute disease.
But these patients have a proliferative syndrome and they require chemotherapy for its control.
So, again, a rare disease, but these two diseases are B-cell diseases linked to KSHV.
I'm not going to say anything more about them today,
but just to point out that they are further support for the idea that the principle reservoir of KSHV is in the B-cell
and there are pathologic consequences to the B-cell infection.
An important aspect of this was cultured primary effusion lymphoma, or PEL cells,
were central to advancing the virology of KSHV.
With our colleagues Mike McGrath and Brian Herndier in San Francisco,
who had isolated several PEL cell lines from patients at San Francisco General Hospital,
we showed that these cells lines, which retained KSHV DNA,
could be coaxed to produce the virus particles themselves.
So, the B-cell lymphoma lines are latently infected, they have cryptic infection of KSHV,
but if treated with phorbol esters or other chemicals, we showed that they could undergo lytic reactivation,
and produce infectious virus particles, electron micrographs of which are shown in this slide.
And that provided a great source of early antigens and virus for future work,
so finally, we had a system for cultivation, in which virus particles could be collected and used to infect other cells,
but also, these PEL cells, because they were infected, could be used as a source of antigen
to develop a serologic test.
I want to tell you about the serologic test,
but I wanted to emphasize why we were interested in making a serologic test for KSHV.
Because remember, at the time, the question was, was KSHV really the agent predicted by the epidemiologists?
Was it really causally linked to KS?
We didn't want to spend a decade working on the molecular biology of a virus that was not the causative agent of KS,
so we and others, many others, including Pat Moore, Thomas Schulz, Chris Boshoff, many other groups, Denise Whitby,
began to try to develop, based on PEL cells in culture, serologic tests.
And the one that we developed is schematized here.
It's an immunofluorescence test in which the antigen from latently affected primary effusion lymphoma cells,
we incubate these with a patient's serum, wash and then come in with a second antibody,
a FITC conjugated, a fluorescinated anti-human IgG.
And this is what it looks like, the nuclei of the cells light up,
with an antigen that we called latency-associated nuclear antigen, or LANA.
And it's a very crude test, those of you who do immunofluorescence know that it's an insensitive method.
The sensitivity of this method was about 75 percent, 65-75 percent. But it is highly specific.
The pattern of staining, nuclear, punctate nuclear staining, was highly specific.
The test is laborious, it all has to be done manually,
but it was our first entree into developing a test that would diagnose KSHV infection,
or exposure to KSHV in a human serum.
Nowadays, this kind of testing has gotten much more advanced, it's typically done by enzyme immunoassay,
often for LANA with one or more lytic antigens thrown into the mix
and the tests have improved to a sensitivity of 90 percent and specificity of 97 percent,
although there are still no commercially available tests for routine diagnosis.
But using our first order, homebrewed immunofluorescence test, Dean Kedes, an infectious disease fellow in my lab,
studied over, in the aggregate, 5,000 patients
from various groups. And his work and the work of others established a striking fact,
which is that KSHV prevalence in the population strikingly mirrored KS risk.
So, if you look, as Dean did, at the blood donor population, a proxy for the whole population,
only 2 or 3 percent were positive in this test.
Similarly low rates were found among *** positive hemophiliacs,
and only slightly higher rates were shown in *** positive women,
groups that we knew were at very low risk for KS.
But, when he looked at *** positive gay men, seropositivity rates rose to anywhere from 30 to 60 percent,
depending on the population and the test.
And this is a group where we know roughly 20 percent of these folks had or would develop Kaposi's sarcoma.
So, there was a striking segregation of KSHV infection, or exposure, with known risk for KS.
And this was a very powerful fact that strongly advanced the argument that the genome discovered by Moore and Chang
really was the agent that had been predicted by the epidemiologists.
There's a lot more in the decade that has followed,
many groups now have examined populations all over the world
and what was true in the United States is true globally. Where KS is uncommon, the seroprevalence of KSHV is low,
so in the US and Western Europe, it's somewhere between 2 and 7 percent,
whereas in endemic zones of the Mediterranean basin, southern Italy, Sicily, Sardinia,
between 20 and 25 percent of the whole population is seropositive.
And in central and southern Africa, where KS has long been known to be a highly prevalent tumor,
the seroprevalence is between 50 and 70 percent.
So, again, not only in our domestic population, but globally, where KSHV infection is frequent, KS is common,
where infection is rare, KS is rare.
The other thing about these low prevalence and high prevalence zones
is they define two different epidemiologic patterns of KSHV spread.
And the mechanisms here are still not fully understood,
but in Western Europe and America, where seroprevalence is low,
infection is principally derived by *** transmission in adulthood,
and especially among gay men,
something about homosexual *** congress appears to be a particularly efficient way of spreading KSHV
and this is unusual in sexually transmitted diseases,
most of them have relatively equal rates of exchange among the different sexes.
But in western Europe and America, that doesn't seem to be the pattern.
In the endemic zones, however, in Sicily and Sardinia,
and especially in Africa, there is very efficient spread by some non-*** route.
We know this because many children, many pre-pubital children in these cultures
seroconvert in the first decade or two of life, before the onset of *** activity.
And then of course, after the onset of *** activity, rates of infection continue to rise,
so there is *** transmission in those societies, but there's also another form of transmission,
which we believe is due to salivary exchange.
There's lots of KSHV in saliva, saliva is the principle body fluid that has virus in it,
and it's thought that in those cultures, salivary exchange is an important route of transmission.
Why these patterns are what they are, however, we really don't know,
why some regions display one pattern of spread and others another,
is something that remains to be understood.
However, we did show, with Jeff Martin, a very gifted epidemiologist here at UCSF,
that among gay men, *** transmission is very strikingly efficient
and is strongly correlated with the number of sex partners.
So, here you can see the rate of antibody positivity rising as the number of *** partners increased,
and this was the first evidence for *** transmission in the United States.
Alright, so let me sum up the epidemiologic argument that links KSHV infection to KS,
because I think there's a nuance here that's very important to understand.
So, here are the points.
First of all, KSHV DNA is present in all KS lesions,
irrespective of whether they're *** positive or *** negative.
Second, as I showed you, both locally and globally, KSHV seroprevalence tracks strikingly with KS risk.
I haven't shown you this, but it's true, that infection with KSHV precedes KS development
and predicts an increased risk of KS.
So, in prospective studies that we did with Jeff Martin,
we showed that if you were a young gay man living in San Francisco
and you were dually positive for KSHV and ***, and you received no effective treatment for either virus,
over the next ten years, 30 to 50 percent of those men developed Kaposi's sarcoma.
Finally, we note that KS is never observed in the absence of markers for KSHV.
So, if you do not have KSHV infection, you will not get KS.
So, those are the hard epidemiologic facts that link KSHV to the tumor,
but, it's important to realize that what they establish is that KSHV is necessary for the tumor.
There's good reason to believe that KSHV infection is not sufficient for the tumor.
I told you that somewhere between 3 and 7 percent of the US population is infected with KSHV
but these subjects in general have a very low risk.
The actual magnitude of this risk hasn't been measured because KS is a relatively rare tumor outside of the AIDS population in the United States.
But in Sicily and Sardinia, island populations where there's very good epidemiology,
and where seroprevalence is 20 to 25 percent,
there are pretty accurate numbers that allow us to estimate that if you're KSHV seropositive
and don't have ***, your risk of developing, only 1 in 3,000 to 1 in 10,000 of those people
will develop KS.
Therefore, it's very clear that in addition to KSHV, some other cofactor must be required.
Now, we don't know what that cofactor is in the *** negative form of the disease,
but in the AIDS related form, we know that that cofactor is ***.
Here's one illustration of that fact and this is derived from epidemiologic studies of Jeff Martin back in the 90's,
before there was effective treatment for ***.
It was shown that your risk of developing KS, prospectively, was strongly linked to the viral load.
So, if you had greater than 100,000 copies of *** per milliliter of plasma,
people rapidly and efficiently developed KS,
whereas, if you had an intermediate level, 10,000 to 100,000, your risk was slightly lower,
and if you had less than 10,000 copies of *** per milliliter of blood,
spontaneously, now, without treatment, you had a lower still risk of KS.
So, that was one thing, there was a clear relationship, a dose relationship,
between the activity of the *** infection and the risk for developing KS over ten years.
And a particularly striking feature has been what's happened to the epidemiology of KS
since the introduction of highly active anti-retroviral therapy, or HAART.
Highly active therapy became widely available in mid-90's, around '95, '96,
and since then has been progressively adopted. And as HAART has become more prevalent among *** patients in the United States,
the incidence of new cases of Kaposi's has plummeted, has decreased about 90 percent.
So, the data I'm showing on this slide are from Washington State, but the same thing has been seen all over the country,
and in fact everywhere where HAART has been widely adopted.
So, not only is there a correlation between spontaneous *** replication and KS risk,
but the interdiction of *** replication has led to a striking reduction in KS risk.
So, all of this goes to say that KSHV is the essential driver of the process,
but it requires an amplifier like *** in order to produce clinical disease.
Alright. So, how does KSHV infection predispose to KS risk?
That's the question now. Nobody doubts that KSHV infection is the central element of pathogenesis here,
but how does it predispose to KS?
Of course, to a virologist, this question boils down to one thing:
what viral genes are expressed in the tumor and what do they do?
And this is where, and we'll explore this question in the next part of the lecture.