LTR Inhibitors, a New Kind of Potential AIDS Treatment: Interview with Arthur B. Pardee, Ph.D.

In a paper published in March 1993, Chiang J. Li and others
at Dana Farber Cancer Institute and Harvard Medical School
reported results of laboratory tests which identified three
substances which should be investigated as potential
treatments for AIDS or HIV.(1) The three are curcumin (an
antioxidant and anti-inflammatory agent found in turmeric,
the mild spice used in curry), topotecan (an experimental
treatment now in phase II/III trials for cancer, but never
tested in patients as a treatment for HIV), and beta-
lapachone, a plant derivative which has not been tested in
humans. The researchers found these three by screening
selected substances for activity against the LTR (long
terminal repeat), which is the Ron-off switchS of the HIV
virus; when the RswitchS is off, the virus is not being
stimulated to reproduce. After these tests, all three
substances were tested against live HIV, and all showed
activity in both acutely and chronically infected cells. In
short, this new approach to finding AIDS treatments has
quickly found three attractive potential candidates --
substances which are available, already in human use in two
cases, active against both acute and chronic HIV infection in
cell cultures, and which work by a completely different
mechanism than standard treatments.

To help the AIDS community understand LTR inhibitors as an
approach to drug development, AIDS TREATMENT NEWS interviewed
Arthur B. Pardee, Ph.D., who is head of the group which
initiated the research cited above. Dr. Pardee is Professor
of Biological Chemistry and Molecular Pharmacology at Dana-
Farber Cancer Institute and Harvard Medical School, and Chief
of the Division of Cell Growth and Regulation at Dana-Farber.
He is a member of the National Academy of Sciences, and has
co-authored 400 technical papers and other publications. He
is a molecular biologist; two other researchers on the
project, Clyde S. Crumpacker, M.D., and Bruce J. Dezube,
M.D., have run AIDS clinical trials for years.

The laboratory test for LTR inhibitors is well established in
basic research. (The particular version of the test used by
PardeeUs group was donated by the HerzenbergUs laboratory at
Stanford University -- which has studied NAC, an experimental
HIV treatment which may reduce LTR activation indirectly.)
Dr. PardeeUs team has used this test to study anti-HIV
properties of available, known, nontoxic compounds -- to look
for a new class of potential drugs which target the on-off
switch of the virus. And they have used the same test to show
that these kinds of HIV inhibitors may work far better
together, in certain combinations which block different paths
involved in LTR activation.(2) (This recent paper, on
combining a Tat inhibitor with pentoxifylline in laboratory
tests, was described in AIDS TREATMENT NEWS #188, December 3,
1993).

Clearly this line of work should be pursued rapidly --
especially since other approaches have had only limited
success. But much of this research is not funded; only the
fact that the laboratory tests are inexpensive enabled it to
be done at all. The problem seems to be that searching for
AIDS treatments by screening for LTR inhibitors has not yet
become an influential part of the mainstream thinking, which
has focused on other kinds of drugs -- mostly AZT, etc., or
high-tech approaches which may not be ready for years.

One great barrier stopping new AIDS drugs today is the lack
of funding and other support for preclinical work needed to
move promising ideas from the laboratory into early human
tests. This barrier is now impeding the practical development
of LTR inhibitors, slowing or preventing the fairly obvious
research steps that are needed next.

Background for Non-Technical Readers

After HIV infects a cell, its genetic information becomes
part of the DNA (the genetic inheritance) of the cell. Each
cell contains a complete copy of all the personUs DNA -- the
information needed to construct all the organs of the body.
Of course not all of this DNA is active in any particular
cell. Instead, complex regulatory mechanisms determine which
genes are expressed, and when. (Each gene is a part of the
DNA which codes the information for creating one protein.)

When a cell has been infected and contains the HIV DNA, that
DNA is often inactive, not making new virus or doing any
other harm. Only when the HIV is activated is it harmful.
Various different substances can increase or decrease this
activation, and many of them have their effect through the
LTR (the long terminal repeat of the AIDS virus).

An LTR inhibitor is a compound which prevents or reduces this
activation, and thereby blocks viral replication and other
viral functions.

There is a rapid, inexpensive test which can be used to
screen substances to identify the ones which show promise as
LTR inhibitors. This test, created by genetic engineering,
uses living cells which have additional, artificially added
DNA. This additional DNA includes the LTR of HIV (or of
another virus being studied), together with a RreporterS
gene. A reporter gene is a gene which produces a protein
(usually an enzyme) that can easily be detected in the
laboratory (for example by causing a color change in a test
tube). In the test cells, the reporter gene is controlled by
the HIV LTR, so that it produces its protein only when the
LTR is activated.

To run the test, the cells are treated with a chemical (such
as tumor necrosis factor) which is known to activate the LTR,
and the resulting activity of the reporter gene is observed.
A substance being tested can be added; if it inhibits the
LTR, it will turn off or reduce the activity of the reporter
Gene. This test is easy to do, and it does not require live
HIV, so it does not require the elaborate safety precautions
needed to work with HIV itself. It can be done in most
laboratories, and could quickly screen many substances to
look for potential LTR inhibitors. Positive results must be
confirmed using laboratory tests with the live virus, but
this only needs to be done for the much smaller number of
substances which show good results from the initial
screening. (These screening programs often test hundreds, or
even thousands, of random chemicals to find one which may
have value. PardeeUs group had better success looking for LTR
inhibitors -- three found out of a dozen chemicals tested --
because they were not funded for mass screening, and instead
used intelligent guesses about what was likely to work.)

Of various factors which increase HIV activity by activating
the LTR, the most powerful is the protein produced by the Tat
gene of HIV. Another important cause of viral activation is
excessive levels of a different protein, NF-kB (NF-kappa B)
-- which (unlike Tat) is normally present in the body. Each
of these binds to the LTR (at different sites), activating
HIV. NF-kB can be induced by a variety of stimulants,
including T-cell activation, inflammation, bacterial
products, and oxidative stress.

Other research groups had found that HIV is much more active
when both the Tat protein, and also high levels of NF-kB, are
present.(3) It appears that the two together act in concert
in a way that is far more than additive; in one experiment,
NF-kB alone upregulated the virus by 16 times, Tat alone by
80 times, and both together 388 to 760 times, depending on
the concentration of Tat used.

Dr. Pardee and his colleagues have built on basic research to
find potential LTR inhibitors, and especially synergistic
combinations of these compounds. Recently Debajit K. Biswas,
Ph.D., and others showed that a combination of drugs that
inhibit both NF-kB and Tat worked much better than either one
alone -- so much better that the concentrations of both drugs
(pentoxifylline, and the Roche Tat inhibitor Ro 24-7429)
could be reduced ten times, and still give the same
inhibition as the larger concentration of either one
alone.(2)

This paper gives some insights as to why the Hoffmann-La
Roche trials of its Tat inhibitor failed to reduce viral
activity in humans. Since NF-kB activity is likely to be
excessive in people with HIV, due to oxidative stress caused
by infections, it may be necessary to combine a Tat inhibitor
with another treatment (such as pentoxifylline) to reduce NF-
kB activity. Further improvements may be made by adding other
LTR inhibitors with different mechanisms of action.

The scientific search for LTR inhibitors also led Dr.
PardeeUs lab to investigate topotecan, which inhibits a DNA-
unwinding enzyme which upregulates many genes (in other
words, makes them more active), including those governed by
the HIV LTR. This inhibitor, now in use in cancer trials but
never tested in persons with HIV, can be tolerated in
patients at a concentration that is 200 times that which
inhibits 80 percent of viral production in laboratory
tests.(1)

Dr. PardeeUs lab is also considering a number of substances
to modify levels of pro-inflammatory cytokines, secreted by
certain immune cells, which upregulate the LTR. (In addition
to pentoxifylline to decrease TNF, another example is
tenidap, to inhibit IL-6 production). Also, as certain enzyme
systems become upregulated in people with AIDS, inhibitors of
these pathways might also be targets for inhibiting the LTR.

The synergistic inhibition of viral production, using novel
combinations of drugs at tolerable doses, is a here-and-now
approach. In view of the urgent need for better treatments
for AIDS, we must not overlook a potentially effective
antiviral strategy that is possibly within reach today.

Interview

Dr. Pardee was interviewed by John S. James of AIDS TREATMENT
NEWS, and by Charles Davidson, a graduate student with a
strong interest in Tat and LTR inhibitors, who brought much
of this information to our attention.

James: Could you summarize the rationale for your work?

Pardee: Our approach is a simple one. We look at the LTR of
HIV-1, and see what substances respond. ItUs a quick, easy
test, and you donUt need special safety precautions [which
would be required for working with the live virus]. WeUre
pretty much limited to that test [because of lack of funding
to do the live-virus tests]. But weUve come up with
pentoxifylline, and the other three compounds we reported a
year ago. WeUve done other work on tenidap, which looks
interesting, and we have more potential drugs in the
pipeline. This approach has been productive. We have shown an
effect on the LTR system; now itUs pretty much up to someone
else to carry it forward. Bruce [Dr. Bruce Dezube] can carry
it into the clinic if it looks good, but you need somebody in
between [to do the preclinical development], and we do not
have funds for that.

Davidson: You found the topoisomerase inhibitor was quite
effective?

Pardee: Topotecan, the one we used in that work of a year
ago, was by far the most potent drug. We think it has some
promise. It was tested with the virus system very
extensively, as we reported in our paper.(1) For reasons that
arenUt too clear to me, Smith-Kline Beecham does not seem
very interested in pursuing it further. ItUs the companyUs
decision at this point; itUs pretty much out of our hands.

We have largely shifted to some other camptothecin analogs;
they look pretty effective, too. We are working mostly with
camptothecin itself; there are analogs that are clinically
preferable, but there are no patents on camptothecin; you run
into problems when you work with companies.

There are several ways a drug can inhibit the LTR, which is
quite complicated. There are NF-kB motifs, and Tat-TAR
motifs, and some others, SP-1 motifs, etc. The questions are
what is the target for a particular compound, and how can we
set up better assays for finding new compounds, by taking
advantage of what weUve learned with the first generation.

And of course curcumin [first discovered to have anti-HIV
activity by PardeeUs group, using the test for LTR
inhibitors] is another one which needs to be investigated. I
understand there will be a clinical trial in the Boston area,
also [to be conducted by the Community Research Initiative of
New England -- in addition to the trial by SEARCH Alliance,
in Los Angeles]. I have no further information on it. We are
starting further work on curcumin. We hear stories about
India, where they eat a lot of turmeric; some people say
thereUs lots of AIDS in India. I understand there is a lot of
HIV infection, but not very much AIDS, so it may slow down
the disease. Dr. Biswas, my Indian colleague, is currently in
India; he is in close contact with top leaders in their
infectious-disease work, and they are interested. There are
also some other Indian plant products which may be quite
interesting; weUre hoping to look at them. Concerning
curcumin, I have been told by another Indian scientist that
there are other compounds in that plant which are more active
than curcumin. There is much we can do if we get the time and
resources.

James: Two callers told me that they didnUt know where to get
the more concentrated curcumin, so they used turmeric from a
grocery store. Both said they had substantial improvement in
blood work. This should be studied.

Davidson: Curcumin is classified as an antioxidant?

Pardee: Yes, but we donUt know what its real mechanism of
action is against HIV. Somebody published an abstract at a
recent meeting saying it may inhibit the HIV protease. They
did a computer study of what the structure should be to block
the specific HIV protease, and came up with a curcumin
structure.

Davidson: If you have several different mechanisms, and there
is synergism, one should be able to use smaller doses of the
drugs?

Pardee: Right, you should be able to get more effectiveness
with less side effects, we hope. But people are very
complicated organisms; just because it works on some simple
system in the laboratory does not mean there will be no
toxicity or other problem.

Pardee: I wish that what might be called the AIDS
establishment was a little more open to innovative
approaches. I have not been able to get a penny from the NIH
to do any of our work, and we have come up with several new
compounds [with potential use for AIDS], and we have several
more in the pipeline. We will soon report on another class of
compounds which is effective against the HIV LTR. But we
canUt go on working without funds. We have to pay researchers
to eat and pay rent.

James: What study groups at NIH review your proposals?

Pardee: It goes into some AIDS study groups. I must say that
one of the reviews, on attempts to follow up on the
topoisomerase inhibitors, was erroneous. It claimed that we
did the work before we submitted the grant, which is
absolutely untrue. We did all this work in the two years of
the grant; we published two good papers, and they said we
didnUt accomplish anything. And they also said we should be
working on HIV-2 as well as HIV-1. Well, we are not a
factory. If we discover something active against the HIV-1
LTR, it could be pursued with HIV-2 by someone else. ItUs
pretty discouraging.

James: Over the years I have seen that this is not a new
problem.

Pardee: The AIDS establishment has to produce something with
all the money they are spending. They give us no
encouragement or help at all [for innovative approaches]; and
we get the same story from colleagues. The whole AIDS
granting process has been dominated by the AZT types, and a
few other interest areas. ItUs like molasses to get funded.
This new panel [the National Task Force on Drug Development]
is likely to have people who donUt rock the boat.

A carefully worded article about how we need innovative
approaches against AIDS, and need to support these innovative
approaches, would hopefully do some good if it gets to the
right people.

We need to find out more about the mechanism of action, why
these work. We are likely to find some new targets. When we
know what the target is, and we get an assay for that target,
we can test for other substances that work better. We
published on several potential drugs already. Nobody has
clinically tried topoisomerase I inhibitors on AIDS, and yet
these compounds are pretty powerful, as we reported.

We think we have a fairly clear understanding of
pentoxifylline, how it blocks the LTR. ItUs through NF-kB,
but there are additional details. ItUs part of another very
interesting class of inhibitors of the LTR. ItUs not too hard
to come up with possibilities. Hopefully some of them will
work.

Our approach is to use the LTR as a target, just like others
use the protease as a target. I think itUs a very promising
target.

Davidson: How much do you think has been spent so far on LTR
inhibitors?

Pardee: I donUt know how much companies such as Roche spend.
We get a small amount to do our work on pentoxifylline, from
Hoechst-Roussel Pharmaceuticals, Inc. We luckily got an NIH
grant to work on topoisomerase inhibitors, which I used to
work on the AIDS problem; but that is terminating and wonUt
be renewed, so this work will stop rather fast. We have never
been funded to work on this project specifically; we havenUt
gotten one cent. And very few people elsewhere are working on
this approach. Virtually nothing has been spent on it. WeUll
eke out what little we can.

Davidson: Do you have any idea how the Roche Tat drug [Ro 24-
7429] works?

Pardee: There are accessory proteins involved in the Tat-TAR
interaction. WeUre trying to take these apart as fast as we
can.

[EditorUs note: Obstacles to the research on the Ro 24-7429,
the Hoffmann-La Roche Tat inhibitor, have included not only
lack of funding, but difficulty getting the drug to test, and
lack of access to results from other researchers. Even though
this substance failed to reduce viral activity in humans in a
single-drug trial, it may be important because of the
possibility that it might work in combination -- for example,
with pentoxifylline, since that combination works very well
in the laboratory. Also, it is worth investigating why Ro 24-
7429 did not work by itself in people; there are many
possible reasons that a particular drug will not work in the
body, and the failure of one compound does not invalidate the
whole approach. (A recent paper suggests that Ro 24-7429
alone does not inhibit the Tat protein itself, but another
protein which works with Tat.(4) This might help to explain
the disappointing results of human tests.)

While there are a number of ways to inhibit the LTR through
the NF-kB mechanism, Tat inhibitors are still scarce. A very
different Tat drug is being developed in Canada, by Allelix
Biopharmaceuticals, but it is just beginning human tests (see
AIDS TREATMENT NEWS #187, November 19, 1993). A prescription
drug, d-penicillamine, was tested as a Tat inhibitor in
Europe several years ago; it did not help patients in trials,
but perhaps should be re-examined as a possible part of a
combination.

We believe that the most important work at this time is human
trials of topotecan, curcumin, some other LTR inhibitors
being tested by PardeeUs group, and the combination of Ro 24-
7429 with pentoxifylline. Equally important is the laboratory
search for new LTR inhibitors, and especially for better
combinations of LTR inhibitors, and combinations of these
with other classes of AIDS treatments. This area could move
quite rapidly, if it had support instead of roadblocks from
corporations and from government research agencies.]

References

1. Li CJ, Zhang LJ, Dezube BJ, Crumpacker CS, and Pardee AB.
Three inhibitors of type 1 human immunodeficiency virus long
terminal repeat-directed gene expression and virus
replication. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES,
USA. March 1993; volume 90, pages 1839-1842.

2. Biswas DK, Ahlers CM, Dezube BJ, and Pardee AB.
Cooperative inhibition of NF-kB and Tat-induced
superactivation of human immunodeficiency virus type 1 long
terminal repeat. PROCEEDINGS OF THE NATIONAL ACADEMY OF
SCIENCES, USA. December 1, 1993; volume 90, pages 11044-
11048.

3. Liu J, Perkins ND, Schmid RM, and Nabel GJ. Specific NF-kB
subunits act in concert with Tat to stimulate human
Immunodeficiency virus type 1 transcription. JOURNAL OF
VIROLOGY. June 1992; volume 66, number 6, pages 3883-3887.

4. Braddock M, Cannon P, Muckenthaler M, Kingsman AJ, and
Kingsman SM. Inhibition of human immunodeficiency virus type
1 Tat-dependent activation of translation in Xenopus oocytes
by the benzodiazepine Ro 24-7429 requires trans-activation
response element loop sequences. JOURNAL OF VIROLOGY. January
1994; volume 68, number 1, pages 25-33.