Tip:
Highlight text to annotate it
X
Alright, so, we've considered in the last section the contributions that the latency program might make
to at least some aspects of the tumor.
And that is to say the pro-inflammatory and pro-angiogenic components of the tumor.
But we're now going to turn our attention to the question of,
is latency the whole story?
And why are we interested in this question?
We're interested in it because I've already told you that the latency program of KSHV is not powerfully immortalizing.
In fact, we can barely show that it's immortalizing at all.
And we don't observe prolongation of cell survival in culture,
that could be a laboratory shortcoming, that we haven't found the right conditions under which to assay it,
but, you know, it's clearly a very different situation from EBV,
where the latency program is very powerfully immortalizing.
And that led us to wonder whether the latency program was really sufficient
to engender all the pathogenesis in KS.
And has led us to consider some other lines of investigation.
Now, I'm going to take a somewhat of a detour on my way to that question
through one other aspect of latent cell biology.
And this concerns our old friend LANA, the latent protein that I told you was the major antigen in our initial serologic work.
But that others have shown plays a very important role in maintaining the nuclear episome
during latency. So, you'll recall that I said that in latency, there's no integration.
The viral genome, which is linear, in the virus particle, undergoes circularization in the nucleus
and is then stably maintained in the nucleus at a low copy number,
about twenty plasmids per cell.
It turns out that KSHV, as might be predicted, encodes machinery that is involved in that replication.
And this is the work largely of Ken Kaye, Rolf Renne, Erle Robertson,
Adam Grundhoff in my lab contributed some to this as well.
But these groups showed that LANA protein can specifically bind to sequences in the terminal repeats of the genome,
and that those repeats are very important in promoting episomal replication,
so that they function as a latent origin of replication.
But also, these groups have shown that KSHV can bind several components of host chromatin,
including histones 2A and B, Brd4. And that's led to a model
that suggests that LANA protein is not only involved in episomal replication
but also in the segregation of KSHV genomes to daughter cells,
the so-called tethering model, in which one domain binds to the KSHV episome
and another domain at the other extremity of the molecule
binds to chromatin and as a result during metaphase and during mitosis,
these KSHV episomes can piggyback onto the host chromosomes,
whose spindle apparatus is ensuring their proper segregation to both daughter cells.
So, in this fashion, any latently infected cell that undergoes cell division can assure that statistically,
with a pool of twenty, it's very likely that both daughter cells will be infected.
So, this is the received wisdom about KSHV replication as derived by studying primary effusion lymphoma cells in culture
and other cells that have been selected to maintain the episome.
But when we began to look at authentic KSHV infection, not transfection and selection of isolated plasmids,
we discovered that this mechanism must function very inefficiently in authentic infected cells
such that KSHV latency, at least as studied in cell culture, is extremely unstable.
This came to light when Adam Grundhoff infected a whole series of different cells in culture with KSHV
and then looked at, over time, how well they retained the episome
by looking periodically at what percent of the cells were still positive for LANA.
And what he found is that most cell lines behave like this,
that within a couple of weeks, if the cells are growing in culture, they actively lose the episome.
So, this was true for immortalized endothelial cells, this was true for HeLa cells
and a whole bunch of other cells, would lose the episome altogether.
There were a couple of cell lines, like SLK and 293, that lost their plasmids very rapidly in the beginning
and then out a few weeks later, a sub-population of about 10 percent of the cells
would stably maintain the episome, ten to fifteen percent of the cells.
And when Adam went to these SLK cells, so SLK is an endothelial cell line,
and he made single cell clones out at 3 or 4 weeks and found that 90 percent of these clones
were negative for the genome, had no LANA staining and no KSHV DNA.
So, they hadn't simply extinguished LANA protein expression,
they'd actually lost the episome.
While about ten percent of the clones actually retained KSHV DNA and continued to stain for LANA.
In fact, the brightness of LANA was correlated with the copy number of episomes.
So, faintly LANA positives had lower copy numbers than brightly LANA positives.
So, from these SLK negative segregants, or so-called SLKn,
he established some cell lines and we similarly established lines from SLK positive lines, which he called SLKp.
Now here's the interesting thing about SLKp cells.
So, these cells have maintained the stable episome,
so you would have thought if they'd undergone some genetic change
that allowed them to stabilize the episome, that if you superinfected them
with another plasmid or episome from KSHV, that they would maintain such a plasmid.
But that is not true. When you look at SLK positive, p-cells,
they lose the episome, the incoming episome, very rapidly, just like the parental SLK cells,
even though they continue to harbor their own stable episome.
So, what that suggests is that episome stabilization is a property that acts in cis,
it's not the result of some trans-acting factor that's been turned on or modified.
And in fact, we know from other studies that the rate at which this happens
is very strongly suggestive that that change is epigenetic.
In fact, we've now been able to prove that by taking genomes from a PEL cell line,
like BCBL-1, which is very stably latently infected and therefore has, whatever its genome has,
all the appropriate modifications that are necessary.
If we induce that virus and infect SLK cells, it is rapidly lost.
So, that says that it's not a genetic change, it must be an epigenetic change.
So, stabilization is epigenetic and acts in cis.
Ok, so here's the, let me just re-state those things.
Newly established latent episomes are unstable, and if cells are proliferating,
episome loss is going to happen. Now, this is phenomenon that we observed in culture,
but we have very strong reasons to believe that it's also true in human beings.
And this has to do with early experiments of Bob Gallo's.
You'll recall that I mentioned in the beginning that Gallo showed that you could grow spindle cells from a KS biopsy
if you put them in circumstances where they had conditioned medium from activated T-cells.
So, under those conditions, spindle cells will actually divide.
What was subsequently shown is that under those very same conditions,
episomes of KSHV are lost, such that routinely all the cells that are derived by that protocol
after a month or two in culture, are KSHV negative,
even though they were KSHV positive when they went in.
So, in fact, KS spindle cells in a human host are not stably immortalized.
And that differentiates them from PEL cells, that are stably, latently infected.
In the spindle cells, KSHV hasn't undergone whatever those epigenetic changes are.
Or can't propagate them.
Now, rare cells in culture can undergo episome stabilization
and the in vivo counterpart of that is the PEL cell.
These tumors that are growing in the body cavities of patients.
Those cells have been very strongly selected in the patient
and that selection must have included these epigenetic modifications that are required for stabilization.
Alright, so I want you to keep that in the back of your mind
as we ponder this question.
Can latency alone account for tumorigenesis?
Now, here we inherited a certain amount of dogma in tumor virology
that was derived largely from EBV. And this dogma says that in *** viral oncogenesis,
the latency program does all the heavy lifting.
It does the work of tumorigenesis by stimulating proliferation and extending cell survival
and the latency program plays no role in cancer, because after all,
if a cell enters the, and the lytic program plays no role in cancer,
because after all, if a lytically infected cell, if a cell flips into the lytic cycle, that cell is going to die, ok.
So, how could a lytic infection contribute to a proliferative disease.
Those cells can't contribute to the mass.
Now, everyone has always acknowledged that early in the natural history of infection,
the lytic cycle must play an important role because it's involved in disseminating the infection throughout the host.
And in the case of EBV, that results in infecting a lot of B-cells,
the default pathway in the B-cell, as it is for KSHV, is latency.
So, it's always been acknowledged that the lytic program is necessary early in infection
to spread throughout the body and put a lot of B-cells into the latency program.
But then it is imagined that once in the latency program,
they begin their slow but inexorable mutational march to cancer,
rearranging Myc genes and other genes that are necessary for the final development of the tumor.
So, in that dogma, the lytic infection plays this adjunctive role early,
but no ongoing role in the process of tumorigenesis.
And that model was the model that we inherited as tumor virologists
and tried to adapt to this paradigm, but very clearly, things were amiss with the model.
First of all, we could never show that the latency program was strongly immortalizing,
second, we had this problem that the latency program, at least in KS cells,
appears to be very unstable.
And that led us to wonder whether the lytic cycle might play some role in KS pathogenesis.
But because we had no animal model of KS, we couldn't put that to a direct experimental test.
However, once again, clinical medicine provided a very strong clue,
at just the time that we were considering this.
And this was a clinical experiment that was designed to ask no question about KS.
It was a clinical experiment that was designed about, to answer a question about
the treatment of another herpesviral infection, cytomegalovirus.
So, some of you may know that AIDS patients are very susceptible
to the reactivation of cytomegalovirus, or CMV,
a particularly devastating complication for them is reactivation in the retina.
So, ocular, or retinal CMV infection, was a leading cause of blindness in patients with untreated AIDS
and during those horrible early days of the AIDS epidemic, those of us who practice medicine
saw many patients who after becoming thoroughly debilitated from their, and wasted, from their *** infection,
went blind from reactivation of CMV in the retina.
Now, in the 90's, treatment with ganciclovir was developed
and ganciclovir is a drug that was not very well absorbed orally,
so ultimately the standard of care became the development of implantable reservoirs
of ganciclovir that could be implanted right into the eye.
And the treatment was to implant these right into the affected eye
and the release of ganciclovir there controlled CMV in that eye.
And pretty effectively controlled it for a period of many months.
The problem was that that treatment does not give rise to bloodstream levels of ganciclovir,
and there was the concern, would patients develop retinitis in the opposite eye that wasn't treated?
So, in order to address that, a clinical trial was started
in which AIDS patients with CMV retinitis got the standard treatment, which was the implant into the affected eye,
and then they were randomized to either receive nothing, a placebo,
or oral ganciclovir or intravenous ganciclovir.
And very large numbers of patients were studied in a trial that was orchestrated by Dan Martin,
a very gifted ophthalmologist at Emory, then at Emory.
This was a multi-center trial, it involved many centers and the results of this trial were truly remarkable.
And when I read this paper in the New England Journal of Medicine, I was just shocked,
because when the paper came out,
well, first of all the results showed that systemic ganciclovir did protect the contralateral eye,
but incidentally, the clinicians noticed that there was a striking decrease in new cases Kaposi's sarcoma
in the patients that got randomized to oral or especially intravenous ganciclovir.
A six-fold drop, or more, in incident KS.
Now, when this paper came out, it was greeted with a giant ho-hum,
oh yes, more evidence that herpesviruses have something to do with KS.
But to a virologist reading this paper, and one steeped in the dogma of tumor virology,
this was a profoundly important result, deeply subversive of the conventional wisdom.
Why do I say that?
I say that for two reasons. There are two things you need to know about
in order to correctly interpret this experiment.
The first is, who gets CMV retinitis?
We know that CMV retinitis is a late complication of AIDS.
Most of these patients have had *** and KSHV infection for over a decade
and that they have less than 50 CD4 cells left,
that most of them will die within 6 months.
So, these are people who in the long natural history of ***
are at five minutes to midnight.
So, that's fact number one.
They've been carrying *** for a long time, if they've got KSHV, they've been carrying it for a long time.
The second thing you need to know is that ganciclovir is absolutely specific for the lytic cycle
and does nothing whatsoever to latent infection.
We often grow latently infected cells in ganciclovir to prevent lytic reactivation.
So, it has no impact on latency, it's specific for the lytic cycle.
So, if you put those two facts together, you can see that this is a profoundly subversive result
of the conventional wisdom because what it really says is that the continuous operation of the lytic cycle
day in and day out, over the full ten years of the co-infection
is absolutely necessary for the development of KS. So, what these numbers reflect the six month incidence of KS
at, you know, in the last six months, after carrying these viruses for ten or fifteen years.
So, really what the experiment says is not only is lytic infection important
early in the natural history, to disseminate to lots of endothelium,
it's necessary continuously, day after day, week after week, month after month,
throughout the entire natural history of the infection.
And that was really a shock, but really, began to make us consider
that the lytic cycle must be playing some important role here.
Ok, alright, so how might lytic replication contribute to KS?
And here's where two of the things that I've already mentioned, and emphasized, might come into play.
First of all, if latency isn't immortalizing,
then if latently infected cells are destined to die,
then how can a mass ever expand? The only way a mass can expand is if
newly, cells newly infected and recruited to latency outnumber the number that are dying at any given time.
So, and, that is a function of the lytic cycle, lytic infection releases virus
and allows new cells to become infected. The default pathway is latency.
So, in order to constantly be recruiting new cells to latency,
you need the operation of the lytic cycle at some level to generate enough virus to do that,
to replace dying cells.
Second, I've shown you that latency isn't very stable,
that latently infected cells, if they begin to proliferate, will segregate uninfected cells.
And those uninfected cells, if they lack whatever KSHV confers as a survival advantage
are not going to contribute to the mass.
So, again, if latency isn't stable, lytic infection allows, again, the recruitment of new, latently infected cells
that can replace these uninfected segregates.
Finally, there's another possibility that I haven't discussed yet.
But you'll recall that I emphasized when we discussed the histology of KS
that KS is really a pastiche of three processes:
proliferation, inflammation and angiogenesis.
Now, we've already discussed that the lytic cycle is not likely to contribute to the proliferative component of the tumor
because lytically infected cells die, they lyse.
So, they can't really become a part of the mass.
But inflammation and angiogenesis are process that are driven by paracrine factors.
So, if the lytic cycle could produce paracrine factors,
cytokines, chemokines, growth factors, that would operate in trans,
they could stimulate a) the proliferation of other cells in the culture
and b) the influx of inflammatory cells and new blood vessels
that would contribute to this environment that we've already shown is a big part of the mass.
So, here's a third as yet-hypothetical way. Now, interestingly, when you look at the viral genome,
it turns out that KSHV has captured a number of paracrine factors.
There are three CC chemokines in the genome that could contribute to inflammation and angiogenesis.
There's a viral homologue of IL-6, a very potent cytokine that we've already earlier observed
is important in the natural history of multicentric Castleman's disease.
It's a very strong promoter of B-cell survival, and others have shown, is an up regulator of angiogenesis.
And I want to emphasize here that the identification of these factors is not our work
but is the work of Pat Moore, John Nicholas and many other groups that have contributed to this.
But the point is that these genes are not expressed during latency,
they are only expressed during the lytic cycle.
In addition, Enrique Mesri and Silvia Montaner and others have shown that a virally encoded G-protein coupled receptor
is a signaling molecule that is not itself secreted from the cell,
but that can trigger a signaling pathway that up regulates the release of other pro-angiogenic factors
and especially prominently is VEGF among them.
Now, it's a little controversial about the role of the GPCR in pathogenesis
because the GPCR, as a lytic cycle gene, is not secreted from cells
and we showed a few years ago that lytically infected cells shut off many host genes
so that there's only a rather narrow window in which expression of the GPCR
can up regulate genes and lead to secretion, but this is almost without a doubt a part of the pathogenesis of KS.
So, whether these secreted factors are host derived or virally derived,
the lytic cycle can produce many powerful angiogenic and pro-inflammatory substances
that surely must play some role in the evolution of a KS lesion.
So, let's sum up. KSHV's latency program can provoke inflammation,
it can provoke cell survival, at least in the B-cell environment, still not clear whether it does that in an endothelial environment,
and possibly proliferation, although we've never observed that in culture.
But these effects are modest in comparison with other known tumor viruses.
And from clinical observations that suggest that the continuous operation of the lytic cycle is also required,
we believe that the lytic cycle is playing a special role in KS pathogenesis
by regenerating latently infected cells that are lost either to cell death or plasmid segregation.
And by providing paracrine signals for inflammation and angiogenesis.
Now, obviously, there's a lot we have left to learn.
We still don't know what the function of the cyclin gene is in latency,
we still don't know whether latency can provoke proliferation
under conditions that we simply haven't replicated yet in cell culture, so there's lot of unanswered questions.
But from what we already know, we can make some inferences that may have some clinical significance.
First, from the fact that the ongoing operation of the latency program,
of the lytic cycle is important in pathogenesis, we might surmise that drugs like Ganciclovir
might play some beneficial role in the therapy of an established KS lesion.
What I've shown you is that the administration of those drugs can prevent the development of KS,
but what hasn't been shown is whether the use of these drugs could be therapeutically valuable in the treatment of KS.
That's turned out to be a difficult question because we already know
that, you know, in the clinic, because Ganciclovir has toxicities,
it's a bone marrow suppressive drug and there's been some reluctance to treat people
for prolonged periods with Ganciclovir.
The other thing is we have other strategies for dealing with KS
including, in the AIDS related form, the treatment of *** itself,
which is a very successful treatment. So, it's not clear to me that we're ever going to see
a clinical trial of drugs like Ganciclovir.
We don't know, furthermore, how long one would have to treat a patient
with a drug like this in order to see the clinical benefit.
That depends on the rate of plasmid segregation, the half-life of a latently infected cell
and many other things for which we don't have an accurate number.
But, I think it's still out there as a theoretical issue and may one day be the subject of clinical investigation.
Second, what is the role of *** in amplifying the risk of KS?
There are many ideas about this, Gallo's group has advanced the argument
and Barbara Ensoli in particular, now independent many years in Italy,
has advanced the argument that *** makes a specific genetic contribution to KS,
that soluble Tat released from infected cells can play a role as a growth factor on spindle cells.
Be that as it may, I think there are easier explanations for the linkage.
We know that immunodeficiency states in general predispose to amplifying KS risk,
cyclosporin and other anti-rejection drugs in the transplant setting are a good example of that.
And I think now that we know the critical role of the lytic cycle in this process,
we can make some simple analogies. We know what happens to herpesviral infections
when people become immunodeficient. Typically, patients lose control of the lytic cycle,
and develop higher levels of viremia, high viral loads and this contributes to exacerbation of disease
in things like CMV and herpesimplex, it's very well established
that immunodeficiency states lead to loss of control of the lytic cycle.
So, all one has to really imagine is going on in KS is the same thing happening in KSHV.
If *** infected the patients with immunodeficiency lose control of the lytic cycle,
we would expect that they would have higher viral loads and be at increased risk
for the development of KS.
So, I think the things that we've already established here suffice to explain most of the relationship with ***.
But it's still possible that *** does make a specific contribution
over and above the immunodeficiency and that's grist for future scientific investigation.
But I want to conclude on a note about the relationship between science and medicine
because it turns out that not only did medicine at every stage provide all the critical clues
for scientists who were listening to decode a lot of what's going on here,
but medicine continues to do so.
Just in the last year or two, clinical trials in post-transplant KS
have shown that rapamycin is an extremely effective treatment for post-transplant KS
and I think people are now beginning to experiment with rapamycin in other settings,
not related to transplant.
If it turns out to be as effective there as it is in the transplant form,
it'll suggest very strongly that mTOR, the target of rapamycin,
which is a known signaling molecule whose signaling pathway is becoming increasingly understood by cell biologists,
is somehow very influential in KS biology
and we're going to have to start to understand what mTOR signaling does
and how viral infection influences mTOR signaling,
but I think this is pointing already to a very potentially fruitful line of new research
for people interested in the molecular and cellular biology of KS.
So, I'll conclude by reminding you that, as the Talmud says,
"the work of the righteous is done by others,"
nothing I showed you was done with my own hands,
and I want to pay particular tribute to two groups in closing.
First are my students and post-docs over the past 15 years
who have advanced all this work, none of this would have been possible without them.
And finally, a nod to my many colleagues and competitors in the field,
they have set the bar extremely high for us and for each other,
it's been a lot of fun to explore these things and I often say to my students
that we are blessed by having excellent competitors who keep us on our best game
and who set the bar high and advance the work.
So, with that, I'll close, I hope you've enjoyed the things I've had to say,
it's really a remarkable story that large numbers of people have contributed to
and we've been blessed to be a part of it. Thank you.