Inhibition of murine leukemia virus with poly-2'-O-(2,4-dinitrophenyl)poly[A]

Abstract

Poly-2'-O-(2,4-dinitrophenyl)poly[A] (DNP-poly[A] is a potent inhibitor of reverse transcriptases from a variety of sources (I. Kang and J. H. Wang, J. Biol. Chem. 269:12024-12031, 1994). In the present study, its inhibitory effect on the reverse transcriptase (RT) from Moloney murine leukemia virus (MuLV) was investigated. DNP-poly[A] was found to enter the virus spontaneously and to completely inhibit the RT within 30 min at 0 degree C. The inhibitor was also spontaneously transported into isolated human lymphocytes and leukocytes at 37 degrees C. Animal studies have demonstrated the effectiveness of DNP-poly[A] as an antiviral drug when administered intraperitoneally at various doses from 1 to 100 mg/kg of body weight. MuLV-infected mice show the presence of RT in their blood as well as increased numbers of leukocytes. After the administration of DNP-poly[A] at a dosage of 100 mg/kg of body weight three times a week over a 3-week period, RT could no longer be detected by an ultrasensitive RT-PCR assay. Autopsy showed that the spleens of infected but untreated mice were enlarged 2- to 10-fold, with fused nodules and the proliferation of large abnormal lymphocytes, whereas the spleens of infected but treated mice resembled the normal spleens of uninfected control mice. These observations indicate that further study of DNP-poly[A] as a general antiretroviral agent is desirable.

Full text

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Oct. 1996, p. 2311–2317 Vol. 40, No. 10
0066-4804/96/$04.0010
Copyright q1996, American Society for Microbiology
Inhibition of Murine Leukemia Virus with
Poly-29-O-(2,4-Dinitrophenyl)Poly[A]
MARY APEA ASHUN, YIN HU, INSUG KANG, CHIH C. LI, AND JUI H. WANG*
Bioenergetics Laboratory, Natural Sciences Center, State University
of New York, Buffalo, New York 14260-3000
Received 4 April 1996/Returned for modification 21 June 1996/Accepted 22 July 1996
Poly-2*-O-(2,4-dinitrophenyl)poly[A] (DNP-poly[A]) is a potent inhibitor of reverse transcriptases from a
variety of sources (I. Kang and J. H. Wang, J. Biol. Chem. 269:12024–12031, 1994). In the present study, its
inhibitory effect on the reverse transcriptase (RT) from Moloney murine leukemia virus (MuLV) was inves-
tigated. DNP-poly[A] was found to enter the virus spontaneously and to completely inhibit the RT within 30
min at 0&C. The inhibitor was also spontaneously transported into isolated human lymphocytes and leukocytes
at 37&C. Animal studies have demonstrated the effectiveness of DNP-poly[A] as an antiviral drug when
administered intraperitoneally at various doses from 1 to 100 mg/kg of body weight. MuLV-infected mice show
the presence of RT in their blood as well as increased numbers of leukocytes. After the administration of
DNP-poly[A] at a dosage of 100 mg/kg of body weight three times a week over a 3-week period, RT could no
longer be detected by an ultrasensitive RT-PCR assay. Autopsy showed that the spleens of infected but
untreated mice were enlarged 2- to 10-fold, with fused nodules and the proliferation of large abnormal
lymphocytes, whereas the spleens of infected but treated mice resembled the normal spleens of uninfected
control mice. These observations indicate that further study of DNP-poly[A] as a general antiretroviral agent
is desirable.
Since human immunodeficiency virus (HIV) is believed to
be the causative agent of AIDS, arresting its replication has
been the focus of much research. The currently approved anti-
HIV drugs zidovudine, didanosine, and zalcitabine serve as
inhibitors or chain terminators of the reverse transcriptase
(RT) reaction (5, 6, 13). However, the virus is able to mutate
in such a way as to elude these drugs. In a novel attempt to
overcome the development of drug resistance because of the
rapid mutation of HIV, the macromolecular inhibitor poly-29-
O-(2,4-dinitrophenyl)-poly[A] (DNP-poly[A]) was designed
and synthesized (7) on the basis of crystallographic data on the
active site of HIV type 1 (HIV-1) RT (1, 8). This inhibitor was
designed to fit the gross structure of the entire, ;7.0-nm bind-
ing cleft of HIV-1 RT instead of fitting precisely the detailed
structure of a small region in the enzyme. The specific binding
of a precisely fitted small inhibitor molecule can be abolished
by mutation of one or two amino acid residues at the binding
site. However, since the entire binding cleft in RT cannot be
abolished by mutation if the virus is to remain viable, DNP-
poly[A] is expected to be not only a mutation-insensitive in-
hibitor of HIV but also an effective inhibitor of other retrovi-
ruses with similar RT structures. DNP-poly[A] was indeed
found to be a highly potent inhibitor of RTs isolated from
HIV-1, HIV-2, and zidovudine- and nevirapine-resistant strains
of HIV, as well as from avian myeloblastosis virus (14) and
Moloney murine leukemia virus (Mo-MuLV) for which the
50% effective concentration (EC
50
) is in the subnanomolar
range (7). Mo-MuLV originated from sarcoma 37, a trans-
plantable connective tissue neoplasm of mice (9). It has been
reported to produce a generalized lymphocytic neoplasm in
mice within a short period (3, 4) and is specific with regard to
age, strain, and species. Both Mo-MuLV and HIV belong to the
group of negative-stranded RNA viruses which replicate
through a DNA intermediate. In the present work the effect of
DNP-poly[A] on Mo-MuLV was studied both in vitro and in
Mo-MuLV-infected mice as a convenient animal model rele-
vant to AIDS research.
MATERIALS AND METHODS
Compounds and enzymes. [
3
H]poly[A] (22.8 Ci/mmol) was from Amersham
(Arlington Heights, Ill.), and [
3
H]dTTP (83.8 Ci/mmol) was from NEN DuPont
(Boston, Mass.). GF/C filters were from Whatman (Hillsboro, Oreg.). Oxirane
acrylic beads (250 mm), epoxy-activated Sepharose 6B, poly[A]-[dT]
12
, ultrapure
reagents for PCR, and human plasma were all from Sigma Chemical Co. (St.
Louis, Mo.). RNasin was from Promega (Madison, Wis.), MS2 RNA was from
Boehringer Mannheim (Indianapolis, Ind.), Taq DNA polymerase and gel elec-
trophoresis reagents were purchased from Life Technologies (Grand Island,
N.Y.), and HIV RT was purchased from Worthington Biochemical Corp. (Free-
hold, N.J.) and was stored in 10 mM potassium phosphate (pH 7.1) with 1 mM
dithiothreitol (DTT) and 50% (vol/vol) glycerol at 2208C.
Virus and cells. Ecotropic Mo-MuLV particles, counted by electron micros-
copy and suspended in Dulbecco’s modified Eagle medium (high concentration
of glucose with 10% fetal bovine serum and 50 mg of gentamicin per ml) and
elutriated cell preparations of human lymphocytes and leukocytes were pur-
chased from Advanced Biotechnologies (Columbia, Md.).
Assay of RT activity in solution. The RNA-dependent DNA polymerase ac-
tivity of RT was assayed by injecting 3 ml of enzyme solution into 23 ml of an assay
mixture at 378C containing 125 mM Tris-HCl (pH 8.2), 1 mM MgCl
2
, 125 mM
KCl, and 250 mM[
3
H]dTTP as the substrate and 23 nM poly[A]-[dT]
12
as the
template. After 10 min, the amount of [
3
H]poly[dT] that had formed was deter-
mined by precipitation with 10% trichloroacetic acid, removal of the ligand with
glass microfiber filters, and washing of the precipitate with cold 10% trichloro-
acetic acid and 50 mM sodium pyrophosphate and then ethanol. The radioac-
tivity in the washed precipitate was then determined by liquid scintillation count-
ing (7).
Assay of RNA-dependent DNA polymerase activity of Mo-MuLV. The RNA-
dependent DNA polymerase activity of Mo-MuLV was assayed as described
previously (12). Briefly, 5 ml of medium containing 6 310
8
virus particles was
pelleted by centrifugation at 100,000 3gfor 15 min in a Beckman Airfuge
through a cushion of 20% glycerol containing 0.05 M Tris HCl (pH 7.8) and 0.1
M KCl. The pellet was resuspended in 100 ml of a solution containing 0.05 M Tris
HCl (pH 7.8), 0.1 M KCl, 20 mM DTT, and 0.1% Triton X-100. A 10-ml aliquot
was added to a 60-ml total reaction mixture containing 0.05 M Tris HCl (pH 7.8),
0.06 M KCl, 2 mM DTT, 0.6 mM manganese acetate, 0.02 A
260
units of poly[A]-
[dT]
12
, and 1.5 mM [
3
H]dTTP. The reaction mixture was then incubated for 60
min at 378C, and the reaction was terminated by the addition of 3 ml of 10%
trichloroacetic acid (TCA). The acid-precipitated product was collected on GF/C
* Corresponding author. Fax: (716) 645-6949.
2311

filters and was washed extensively with TCA, water, and ethanol and then dried,
mixed with counting cocktail, and assayed in a Wallac scintillation counter.
Testing the stability of DNP-poly[A] in 0.01 M HCl at 37&C. DNP-poly[A] was
incubated in 0.01 M HCl at 378C for 4 and 24 h. At the end of the incubation
periods, each tube was immersed in liquid nitrogen to stop any further inactiva-
tion. The 50% inhibitory concentrations (IC
50
s) of HIV RT (at 25 nM) were
determined by the procedure reported elsewhere (7). Control experiments were
also run with diethyl pyrocarbonate-treated water.
Preparation of Sepharose 6B with covalently attached DNP-poly[A]. About
200 mg of epoxy-activated Sepharose 6B beads was suspended in 4 ml of DNP-
poly[A] solution (2.5 mg/ml). The suspension was mixed with 1 ml of 1.0 M
K
2
CO
3
solution, adjusted to pH 10 to 11 with KHCO
3
, and gently shaken for 2
days at 258C. The derivatized resin suspension was transferred to a 3-ml syringe
fitted with a filter. From the decreased A
259
of the filtrate, the yield was estimated
to be 2 mg of DNP-poly[A] per g of resin. The resin was sequentially washed with
water–1.0 M KCl solution–buffered medium before use.
Preparation of oxirane beads with covalently attached DNP-poly[A]. DNP-
poly[A]-activated oxirane beads were prepared by derivatization as described
previously (11).
Binding of Mo-MuLV to DNP-poly[A] covalently attached to oxirane acrylic
or Sepharose beads. Epoxy-activated Sepharose or oxirane acrylic beads with
covalently attached DNP-poly[A] were preequilibrated with 2 ml of buffer (20
mM Tris HCl [pH 7.5]) or 1 ml of reconstituted human plasma for 72 h at 258C.
Each mixture was kept in a capped sterilized container on a shaker. At each
titration step, a mixture of 10
8
virus particles in 100 ml of buffer or plasma was
added to the resin suspension, and the mixture was shaken at 08C for 60 min. An
aliquot (50 ml) of the supernatant was then removed, centrifuged at 100,000 3g
in an Airfuge through a glycerol layer (20% glycerol, 0.05 M Tris HCl [pH 7.8],
0.1 M KCl) to pellet the virus, and assayed for RT activity (12).
Inhibition of RT in intact virus by DNP-poly[A]. A sample containing 6 310
8
virus particles in 5 ml of medium was incubated at 08C with 10 mlofa2mM
DNP-poly[A] stock solution. At each stated time, the virus particles were cen-
trifuged at 100,000 3gfor 10 min through a glycerol layer to separate the virus
particles from DNP-poly[A] in the external liquid. The supernatant was carefully
removed, and the pellet was washed with 200 ml of Tris HCl buffer (50 mM; pH
7.8) and again centrifuged. The pellet was then homogenized in 12 ml of 0.5%
Triton X-100 at 08C for 3 min and incubated with the Mo-MuLV RT assay
mixture (12) at 378C for 60 min. The assay reaction was terminated by the
addition of 3 ml of 10% TCA. The incorporated radioactive product was col-
lected on a GF/C filter, washed, and assayed by liquid scintillation counting.
Transport of DNP-poly[A] into human lymphocytes and leukocytes at 37&C. A
400-ml suspension containing 8 310
6
cells in 100 ml of fetal serum albumin, 3.2
ml of RPMI 1640, and 300 mlof[
14
C]DNP-poly[A] (0.05 mg/ml; 3024 cpm/mg)
was divided into several aliquots and incubated at 378C for various lengths of
time. At specified times each incubation mixture was centrifuged at 3,000 3gfor
5 min in an Eppendorf centrifuge, and the supernatant was discarded. The pellet
was carefully washed with 3,000 ml of phosphate-buffered saline (PBS) and cen-
trifuged, and the supernatant was discarded. This washing procedure was per-
formed three times. The homogenized pellet was then immersed in 5 ml of counting
cocktail, and the radioactivity was counted in a Wallac liquid scintillation counter.
Animal experiments. Fifty-four male BALB/c mice (age ;3 weeks) were
purchased from Harlan Sprague-Dawley (Indianapolis, Ind.); they weighed 12 to
15 g prior to inoculation and/or treatment. Thirty mice were inoculated intrave-
nously (via the tail vein) with 10
5
to 10
8
virus particles in saline. DNP-poly[A] was
administered intraperitoneally in saline at various doses of 1 to 100 mg/kg of
body weight. Blood was periodically drawn from the tail and was used in the
product-enhanced RT (PERT) assay described below.
Sample pretreatment for the PERT assay. The PERT assay method of Pyra et
al. (10) was used, with some modifications. Briefly, 5 ml of whole blood was
collected into 5 ml of EDTA (13.8% [wt/vol]) plus 45 ml of saline and frozen at
2208C overnight. The hemolyzed blood was thawed and centrifuged at 4,000 3
gfor 30 min at 48C. The supernatant was removed and again centrifuged at
100,000 3gfor 30 min. The pellet was homogenized in 20 ml of buffer A
containing 50 mM KCl, 50% glycerol, 25 mM Tris HCl (pH 7.5), 0.25 mM
EDTA, 0.5% Triton X-100, and 5 mM DTT. This suspension was kept at 08C for
15 min, and a 3-ml aliquot was removed and used in the PERT assay.
PERT assay and product amplification by PCR. cDNA was synthesized and
amplified by the PERT assay and PCR, respectively, by using genomic MS2 RNA
as described previously (10).
RESULTS
Synthesis of DNP-poly[A]. The inhibitor was synthesized
from poly[A] supplied by Sigma (average molecular weight by
light scattering, 10
5
). After derivatization and purification the
product mixture was dialyzed against water via a membrane
with a cutoff molecular weight of 12,000. From the spectro-
photometrically determined DNP-to-adenine ratio, we esti-
mate the average molecular weight of DNP-poly[A] to be 1.1 3
10
5
, with a lower limit of 1.2 310
4
.
Inhibition of RT in solution by DNP-poly[A]. Using poly[A]-
oligo[dT]
12
as the template, we found that DNP-poly[A] is a
potent inhibitor of RT in solution. The dependence of the
observed ratio of the catalytic activity of inhibited enzyme (A)
to that of the uninhibited enzyme (A
0
) on the total inhibitor
concentration is illustrated in Fig. 1; the apparent IC
50
was ;2
nM. It is noteworthy from the data that 5 nM DNP-poly[A] can
cause 70% inhibition of the activity of 25 nM RT, which im-
plies that each long DNP-poly[A] molecule can bind and in-
hibit several RT molecules.
The actual dissociation constants (K
d
s) for the inhibitor
bound to RT should be much lower than the apparent IC
50
given in Fig. 1, because the abscissa of Fig. 1 gives the total
inhibitor concentration, which is sometimes much higher than
the concentration of the free inhibitor. It is very difficult to
treat the data in Fig. 1 quantitatively, because for [RT] .
[inhibitor], the RTs that bound successively to the same inhib-
itor probably have different K
d
values. For an [RT] of ,1 nM,
catalytic activity is too low to be measured reliably.
In order to overcome this difficulty, we attached DNP-
poly[A] covalently to oxirane acrylic or Sepharose beads and
equilibrated the affinity beads that were obtained with different
concentrations of RT or Mo-MuLV. The K
d
of the first binding
sites on each anchored inhibitor molecule was calculated by
measuring the observed removal of RT activity from the liquid
phase.
Inhibition of RT inside Mo-MuLV by DNP-poly[A]. Previ-
ous experimental studies showed that DNP-poly[A] can pro-
tect susceptible lymphocytes from HIV-1 in cell cultures with
an EC
50
of 0.2 mg/ml or 2 nM (7). The observed protection
could be due to the transport of DNP-poly[A] into HIV and
inhibition of the RT inside, the transport of the inhibitor into
lymphocytes and the prevention of reverse transcription after
infection, or the prevention of attachment of HIV to lympho-
cytes by DNP-poly[A]. To explore these possibilities, the rates
of transport of DNP-poly[A] into Mo-MuLV or lymphocytes
were measured.
FIG. 1. Inhibition of HIV-1 RT in solution by DNP-poly[A] at 258C. Aand
A
0
represent the catalytic activities of the RT in the presence and absence of
DNP-poly[A], respectively. The total concentration of RT is 25 nM. The con-
centration of DNP-poly[A] indicated by the abscissa corresponds to the total
concentration of inhibitor (bound and free). The DNP-poly[A] used in this
experiment has an average molecular weight of 1.1 310
5
, and the DNP/adenine
molar ratio is 1:1.2.
2312 ASHUN ET AL. ANTIMICROB.AGENTS CHEMOTHER.

Intact Mo-MuLV particles in aqueous buffer exhibit no RT
activity toward substrate and primer-template in external me-
dium. However, the activity can be measured by first releasing
the endogenous RT with low concentrations of Triton X-100
(0.1 to 0.5%). In the present work, the transport of DNP-
poly[A] into Mo-MuLV was monitored by mixing the virus in
the absence of detergent, taking aliquots of the mixture at
different time intervals after mixing, centrifuging each aliquot,
and rinsing each sediment to remove external DNP-poly[A].
The washed virus was subsequently treated with Triton X-100
to release the RT and was finally incubated in the assay solu-
tion for determination of RT activity. The rate of transport of
DNP-poly[A] into Mo-MuLV was measured by monitoring the
decrease in RT activity with the time of incubation with inhib-
itor. At 378C we found that this decrease was too fast to follow.
At 08C, however, the decrease in RT activity with time was
slow enough to be monitored in this way. The result of a typical
experiment is illustrated in Fig. 2, in which the observed ratio
of RT activity after incubation (A) and before incubation (A
0
)
is shown as a function of incubation time with the inhibitor.
The first point at zero incubation time in Fig. 2 was obtained
by adding the Mo-MuLV suspension to DNP-poly[A] solution,
within 5 s putting the mixture into an Airfuge centrifuge tube
containing a layer of glycerol at the bottom, and immediately
centrifuging the virus through the glycerol layer to remove the
bulk of adhering liquid. It was found repeatedly that the RT
activity of the washed pellet from the zero-time sample (A) was
essentially the same as that from the virus incubated with
buffer in the absence of the inhibitor (A
0
). These control ex-
periments indicate that the removal of externally adsorbed
inhibitor by this procedure was sufficiently complete. The curve
in Fig. 2 indicates that at 08C the decrease in endogenous RT
activity is slow enough to be monitored by this procedure.
The DNP-poly[A] used in the experiment whose results are
presented in Fig. 2 has an average molecular weight of 1.1 3
10
5
, and the adenine/DNP molar ratio was 1.5. Additional
control experiments with [
3
H]poly[A] with a similar molecular
weight show that this underivatized polymer is not transported
into Mo-MuLV under the same conditions, because the level
of radioactivity of the washed virus was only three times the
background counting rate. Since this low radioactivity level was
reached within 5 min (2), we conclude that it was due to
external adsorption of [
3
H]poly[A] by Mo-MuLV.
It is also worth pointing out that the concentration of DNP-
poly[A] in the incubation mixture (1.3 mM) is 1,000 times
higher than the apparent IC
50
of the inhibitor for RT in solu-
tion or the EC
50
for HIV in microculture plates. This high
inhibitor concentration was chosen deliberately to eliminate
slow binding as a possible rate-limiting step in the inhibition
process.
In an attempt to measure the backward diffusion of DNP-
poly[A] out of Mo-MuLV, we found that after the virus has
been incubated with the inhibitor at 1.3 mM for longer than 30
min at 08C, no detectable RT activity can be recovered, even by
reincubation of the inhibited Mo-MuLV in fresh medium for
24hat08C. Under the same conditions the RT activity of
uninhibited virus decreased less than 10%.
When considered together, the observations presented
above indicate that the rate-limiting step in the observed time-
dependent inhibition of endogenous RT in Mo-MuLV is mem-
brane transport, not a slow rate of absorption or binding.
Transport of DNP-poly[A] into mammalian cells. The trans-
port of DNP-[
14
C]poly[A], [
14
C]poly[A], and [
3
H]poly[A] into
human lymphocytes or leukocytes was measured at 378C. The
data are summarized in Fig. 3. These results indicate that the
hydrophobic effect of a large number of DNP groups can
compensate for the hydrophilic effect of the charged phosphate
groups in each long DNP-poly[A] molecule sufficiently to allow
it to permeate the membrane. By contrast, the hydrophilic
poly[A] of equal chain length was not transported at all under
the same conditions. Natural oligonucleotides often have a
poor ability to permeate the membranes. The present obser-
vations suggest a convenient derivatization method for con-
verting them to a membrane-permeable form.
Binding of Mo-MuLV to DNP-poly[A] attached to solid
support. To exploit its ability to permeate the membrane, we
linked each DNP-poly[A] molecule at one end covalently to
oxirane acrylic resin and found that the resulting affinity resin
can bind Mo-MuLV selectively from its suspension in human
plasma. Presumably, the covalently anchored DNP-poly[A]
(approximately 0.06 mm in length) can penetrate the retrovirus
(;0.1 mm) and catch it by binding to the endogenous RT, as
illustrated in the inset of Fig. 4. Equilibration studies indicate
that this DNP-poly[A] affinity resin can selectively remove Mo-
MuLV suspended in PBS or reconstituted human plasma,
whereas the underivatized oxirane acrylic resin exhibited no
binding at all under the same conditions. In order to determine
the free energy of binding, a suspension of the affinity resin was
equilibrated with different concentrations of the retrovirus. At
each equilibrium state, the concentration of free virus (C), the
number of moles of virus bound per milligram of resin (n), and
the virus-binding capacity (N) are related by the Scatchard
equation: n/C 5(N2n) (1/K
d
), where K
d
is the dissociation
constant of a virus bound at a specific binding site. The cor-
responding linear plot of the experimental data in Fig. 4 gives
FIG. 2. Transport of DNP-poly[A] into intact Mo-MuLV and inhibition of
the RT inside the cell at 08C.
FIG. 3. Measurement of the transport of labeled poly[A] derivatives into hu-
man cells at 378C. Symbols: ■,[
14
C]DNP-poly[A] into lymphocytes; Ç,[
14
C]
DNP-poly[A] into leukocytes; E,[
14
C]poly[A] into lymphocytes; F,[
3
H]poly[A]
into lymphocytes.
VOL. 40, 1996 INHIBITION OF MuLV WITH DNP-POLY[A] 2313

aK
d
value of 3.4 310
212
M and an Nvalue of 1.4 310
9
virus
particles per mg. The linearity of the Scatchard plot indicates
that in these very dilute mixtures, Mo-MuLV particles were
bound to only one type of very tightly binding site, with a K
d
in
the subnanomolar range. This K
d
is compatible with the IC
50
observed for the binding of DNP-poly[A] to very dilute Mo-
MuLV RT in solution (7), but many orders of magnitude lower
than the half-saturation concentration expected from any non-
specific binding or adsorption.
Stability of DNP-poly[A]. Unlike the natural polynucleoti-
des, DNP-poly[A] is completely stable in solutions containing
RNases A, B, S, T
1
,T
2
, and H as well as phosphodiesterases 1
and 2 (11). DNP-poly[A] was also found to be quite stable in
acidic medium, in that it was able to retain its inhibition po-
tency when it was incubated for several hours in 0.01 M HCl or
water at 378C (15).
Animal study. The observed antiretroviral activity of DNP-
poly[A] in cell cultures as well as its permeation through the
membrane and stability in the presence of RNase or hydro-
chloric acid suggest that this RT inhibitor could be used as an
effective antiretroviral agent. As a test case, we treated Mo-
MuLV-infected mice with doses of DNP-poly[A] in Dulbecco’s
PBS solution by intraperitoneal injection and followed their
blood chemistry and spleen histology in comparison with those
for the control mice (infected and untreated, uninfected and
untreated, and uninfected and treated).
The hematocrit measurements for 11 mice in the four groups
were obtained (data not shown). The readings gave some in-
dication of the progress of infection during the early stages of
the disease, although the total number of mice (n511) was
too small to be able to provide conclusive results. At about 3
months after infection, the infected and control mice were
sacrificed. Mice that were treated immediately after viral in-
fection with a dosage of 10 mg/kg of body weight three times at
3-day intervals did not manifest other leukemic symptoms such
as large splenic cells or the large spleens which were observed
in infected and untreated mice. Another group of infected
mice that were treated 4 months after infection had lower
spleen weights and fewer abnormal cells than the infected but
untreated mice. The data are summarized in Table 1.
The spleens were preserved in formalin and were subse-
quently stained with hematoxylin and eosin for microscopic
examination. Figure 5 indicates that under low magnification,
the spleen from a healthy mouse (Fig. 5A1) and the spleen
from an uninfected but treated mouse (Fig. 5D1) have similar
spleen architectures, but the greatly enlarged spleen from an
infected but untreated mouse (Fig. 5B1) has lost the normal
spleen architecture completely. The spleen of an infected
mouse which was treated at a later stage of infection (Fig. 5C1)
has some regions of hypernodularity but still maintains some
regions of red and white pulp. Under high magnification, the
spleens from a healthy mouse (Fig. 5A2), an infected and
treated mouse (Fig. 5C2), and an uninfected but treated mouse
(Fig. 5D2) are quite similar. However, large abnormal cells can
be seen to proliferate in the spleen from an infected but un-
treated mouse (Fig. 5B2) to such an extent that normal-size
lymphocytes are very rare in the whole field.
The effect of continued treatment of Mo-MuLV-infected
mice with DNP-poly[A] was also determined by monitoring the
retroviral load in blood by the PERT assay as described in
Materials and Methods. The results are summarized in Fig. 6.
About 4 months after infection, blood samples from infected
but untreated mice all showed strong ethidium bromide-
stained bands in the electrophoresis gel (Fig. 6A). These bands
correspond to about 100 bp according to the marker ladder
which was run alongside it, and the intensities of the bands are
proportional to the concentration of RT in the sample. The
different intensities of the ;100-bp band in lanes 2 to 7 of Fig.
6A reflect different steady-state concentrations of RT in blood.
Our positive controls with pure virus (lane 8 in all panels of
Fig. 6) gave strong bands, and negative controls with normal,
uninfected blood showed no band (not shown in Fig. 6) in the
electrophoresis gel at the 100-bp position. Figure 6B, C, and D
indicates that these bands progressively decreased in intensity
after the treatment with DNP poly[A] was introduced and
eventually disappeared. Indeed, viral RT was still not detect-
able 2 weeks after the treatment was terminated (data not
shown), when all the mice were sacrificed for autopsy.
As toxicity controls, uninfected mice were also administered
1 to 100 mg of DNP-poly[A] per kg by intraperitoneal injec-
tion. No visible ill effect was observed, and the hematocrit,
leukocyte, and erythrocyte levels for these mice remained with-
in normal limits.
DISCUSSION
The use of macromolecular inhibitors as therapeutic agents
is often hampered by their failure to cross biological transport
barriers. We have now found that derivatives of poly[A] with
DNP groups attached to the 29-Opositions via an ether linkage
can be transported freely and rapidly into viruses and cells,
whereas poly[A] itself cannot. This poly[A] derivative was pre-
FIG. 4. Equilibrium binding of Mo-MuLV in human plasma to acrylamide
beads with covalently attached DNP-poly[A]. (Inset) Assumed mode of binding.
TABLE 1. Spleen weight and percentage of abnormal
cells at autopsy
Class of mice No. of
mice Average
wt (g) % Abnormal
cells
a
Normal 4 0.10 60.00 0.0 60.0
Treated immediately after virus
inoculation
b
4 0.12 60.1 7.0 62.0
Treated 4 months after virus
inoculation
c
6 0.20 60.1 37.0 67.0
Inoculated with virus but not
treated for 4 months
d
6 0.42 60.3 59 619
a
The cells were counted from four different areas of each spleen.
b
The mice were inoculated intravenously with 10
4
to 10
5
virus particles and
were then treated intraperitoneally with one 10-mg/kg dose of DNP-poly[A] and
sacrificed 3 months after infection for autopsy to see the effect of inhibitor on
free retrovirus in blood.
c
The mice were inoculated intravenously with 10
9
virus particles. Four months
after infection, they were treated intraperitoneally with 100 mg of DNP-poly[A]
per kg three times per week for 3 weeks. These mice were sacrificed 5 months
after infection for autopsy to see the effect of inhibitor on virus-infected cells.
d
The mice were infected intravenously with 10
9
virus particles, never treated
with any drug, and sacrificed either 3 or 5 months after infection.
2314 ASHUN ET AL. ANTIMICROB.AGENTS CHEMOTHER.

viously shown to protect susceptible lymphocytes in the pres-
ence of HIV (7). The new transport data suggest a protection
mechanism involving the permeation of the inhibitor through
viral and cellular membranes and then inhibition of the RT
inside the cell.
Previous experiments have indicated that DNP-poly[A] can
protect susceptible lymphocytes from HIV-1 in cell cultures,
with an EC
50
of ;2 nM (7). This protection is most likely due
to inhibition of RT, because the titration of 25 nM HIV-1 RT
with nanomolar concentrations of DNP-poly[A] also gave an
IC
50
of about 2 nM (Fig. 1). If the observed protection of
lymphocytes was due to the antifusion activity of the com-
pound, it would be a rare coincidence for it to exhibit nearly
the same EC
50
. Furthermore, inhibition of internal RT re-
quires transport of the inhibitor across the viral envelope or
cell membrane, or both, whereas antifusion activity by oligo-
nucleotides does not. We found that the presumably mem-
brane-soluble amphipathic DNP-poly[A] can be spontaneously
FIG. 5. Spleens from four sets of mice used for the RT-PCR studies whose results are presented in Fig. 6. All the spleens were preserved in formalin and stained
with hematoxylin and eosin. Magnifications, 3160 for panels A1, B1, C1, and D1 and 31,600 for panels A2, B2, C2, and D2. The four spleen samples were from nor-
mal mouse spleen (A); infected (10
9
virus particles), untreated mouse (B); a mouse infected with 10
9
virus particles, after which we waited for leukemia to develop and
treated the mouse intraperitoneally with a 10-mg/kg dose of DNP-poly[A] three times per week for 3 weeks (C); and a mouse infected with 10
8
virus particles but treated
intraperitoneally immediately with a 10-mg/kg dose of DNP-poly[A] three times per week for 1 week (D).
VOL. 40, 1996 INHIBITION OF MuLV WITH DNP-POLY[A] 2315

transported across the viral envelope (Fig. 2) and cell mem-
brane (Fig. 3), and it can also protect lymphocytes against HIV
in cell culture and inhibit RTs in solution. On the other hand,
the presumably membrane-insoluble poly[A] cannot be spon-
taneously transported across cell membranes under the same
conditions (Fig. 3). We found that it neither protects lympho-
cytes from HIV in cell cultures nor inhibits RTs in solution (7).
For all these reasons it seems very unlikely that DNP-poly[A]
functions as an antifusion agent.
In Mo-MuLV-infected but untreated mice, more abnormal
giant cells were observed within the spleen, while the erythro-
cyte population decreased, eventually giving rise to anemia.
The mice that were treated at 4 months after infection had
presumably already developed the disease to such an extent
that although the viral load in blood was markedly decreased,
the spleen was still infiltrated by cells that looked abnormal.
The statistical tests that were performed indicated that DNP-
poly[A] is significantly more effective when it is administered
immediately following infection, although the therapy was still
somewhat beneficial when it was started 4 months after infec-
tion.
Since the mouse has a very low blood volume (5.5 to 8.0% of
its body weight of ;15 g), multiple tests can be performed only
on a microscale. In the present work, RT-PCR was used to
monitor viral load because it requires only 5 ml of whole blood
per assay. When Mo-MuLV replicates, resulting in more virus
particles, the amount of RT increases. Consequently, the cDNA
produced by reverse transcription also increases. Conversely, a
decrease in intensity of the 100-bp band on an ethidium bro-
mide-stained gel demonstrates a decrease in cDNA and there-
fore RT. Our results indicate that multiple treatments with DNP-
poly[A] can decrease the viral load to an undetectable level.
The following additional properties suggest that DNP-poly[A]
should be further studied as a potential antiretroviral agent.
Specificity. Previous studies indicate that DNP-poly[A] is a
potent inhibitor of RTs (7) and a moderate inhibitor of RNases
(11), but it does not inhibit DNP-dependent DNA polymerase,
DNA-dependent RNA polymerase, glucose-6-phosphate dehy-
drogenase, pyruvate kinase, hexokinase, adenylate kinase, etc.
This functional specificity is also demonstrated by the experi-
ment whose results are illustrated in Fig. 4, in which the K
d
for
equilibrium binding of MuLV in human plasma to acrylamide
beads with covalently attached DNP-poly[A] was determined.
When a similar experiment was conducted with Mo-MuLV in
phosphate buffer, a K
d
of the same order of magnitude was
obtained, which implies that none of the protein in human
plasma can compete with Mo-MuLV RT for binding to the
inhibitors.
On the other hand, DNP-poly[A] does not show species
specificity. It was found to inhibit all the wild-type and mutant
RTs that we tested, and hence may be a mutation-insensitive
inhibitor of HIV.
Bioavailability. In addition to the transport data in Fig. 2
and 3, the bioavailability of DNP-poly[A] is also supported by
the observed gradual disappearance of leukemic symptoms
following the intraperitoneal administration of DNP-poly[A].
In order to stop viral propagation, the inhibitor must permeate
many membrane barriers to reach all affected organs of the
leukemic mouse.
Stability. DNP-poly[A] is resistant to all the nucleases that
we tested and is quite stable at 378C in 0.01 M HCl for several
hours. Pharmacokinetic studies will be conducted to map its
biodegradation path.
Low level of toxicity. In cell cultures no toxic effect was
detected in T4 lymphocytes when the DNP-poly[A] concentra-
tion was varied from 0.05 to 200 mg/liter (7). In healthy mice
treated with DNP-poly[A] as a toxicity control, no apparent
toxic symptoms were observed when the intraperitoneal dos-
age was varied from 1 to 100 mg/kg three times per week over
a 5-month period.
ACKNOWLEDGMENTS
We thank Chester Glomski for valuable advice.
This work was supported by the University of Buffalo Foundation.
REFERENCES
1. Arnold, E., A. Jocobo-Molina, R. G. Nami, R. C. Williams, X. Lu, J. Ding,
A. D. Clark, A. Zhang, A. L. Ferris, P. Clark, A. Hizi, and A. H. Hughes.
1992. Structure of HIV-1 reverse transcriptase/DNA complex at 7A resolu-
tion showing active site locations. Nature (London) 357:85–89.
2. Ashun, M. A. 1996. Inhibition of murine leukemia virus by poly-29-O-(2,4-
dinitrophenyl)-poly A. Ph.D. dissertation. State University of New York,
Buffalo.
3. Dmochowski, L., L. Gross, and F. Padgett. 1962. Electron microscopic stud-
ies of rat leukemia induced with mouse leukemia virus. Proc. Soc. Exp. Biol.
Med. 110:504–508.
4. Dunn, T. B., J. B. Moloney, A. W. Green, and B. Arnold. 1961. Pathogens of
virus induced leukemia in mice. J. Nat. Cancer Inst. 26:382–386.
5. Furman, P. A., J. A. Fyfe, M. H. St. Clair, K. Weinhold, J. L. Rideout, G. A.
Freeman, S. N. Lehrman, D. P. Bolognesi, S. Broder, H. Mitsuya, and D. W.
Barry. 1986. Phosphorylation of 39-azido-39-deoxythymidine and selective
interaction of the 59-triphosphate with human immunodeficiency virus re-
verse transcriptase. Proc. Natl. Acad. Sci. USA 83:8333–8337.
6. Huang, P., D. Farquhar, and W. Plunkett. 1990. Selective action of 39-azido-
39-deoxythymidine 59-triphosphate on viral reverse transcriptases and human
DNA polymerases. J. Biol. Chem. 265:11914–11918.
7. Kang, I., and J. H. Wang. 1994. Design of structure-based reverse transcrip-
tase inhibitors. J. Biol. Chem. 269:12024–12031.
8. Kohlstaedt, L. A., J. Wang, J. M. Friedman, P. A. Rice, and T. A. Steitz.
1992. Crystal Structure at 3.5 Å resolution of HIV-2 reverse transcriptase
complexed with an inhibitor. Science 256:1783–1790.
9. Moloney, J. B. 1960. Biological studies on a lymphoid leukemia virus ex-
tracted from sarcoma 37.1. Origin and introductory investigations. J. Natl.
Cancer Inst. 24:933–951.
FIG. 6. Effect of DNP-poly[A] on viremia of Mo-MuLV-infected mice. Each
of six mice was infected intravenously with 10
8
virus particles. Four months later,
when leukemia symptoms were fully developed, the mice were treated intraperi-
toneally with DNP-poly[A] (100 mg/kg at 3-day intervals). In all panels the
ethidium bromide-stained bands in lanes 2 to 7 were from infected and treated
mice. Lanes 1 and 8, PCR marker and pure virus (positive control), respectively.
(A) Prior to treatment; (B) 1 week after treatment; (C) 2 weeks after treatment;
(D) 3 weeks after treatment.
2316 ASHUN ET AL. ANTIMICROB.AGENTS CHEMOTHER.

10. Pyra, H., J. Boni, and J. Schupbach. 1994. Ultrasensitive retrovirus detection
by a reverse transcriptase assay based on product enhancement. Proc. Natl.
Acad. Sci. USA 91:1544–1548.
11. Rahman, M. H., I. Kang, R. Waterbury, U. Narang, F. Bright, and J. H.
Wang. 1996. Selective removal of ribonucleases from solution with covalently
anchored macromolecular inhibitor. Anal. Chem. 68:134–138.
12. Sherr, C. J., and G. J. Todaro. 1979. Murine leukemia virus reverse trans-
criptase assay. In Methods Enzymol. 58:412–417.
13. St. Clair, M. H., C. A. Richards, T. Spector, K. J. Weinhold, W. H. Miller,
A. J. Langlois, and P. A. Furman. 1987. 39-Azido-39-deoxythymidine
triphosphate as an inhibitor and substrate of purified human immunodefi-
ciency virus reverse transcriptase. Antimicrob. Agents Chemother. 31:1972–
1977.
14. Stromberg, K., N. E. Hurley, N. L. Davis, R. R. Ruekert, and E. Fleissner.
1974. Structural studies of avian myeloblastosis virus: comparison of poly-
peptides in virion and core components by dodecyl sulfate-polyacrylamide
gel electrophoresis. J. Virol. 13:513–528.
15. Wang, J. H., I. Kang, and M. H. Rahman. March 1996. Composition and
methods of application of reactive antiviral polyadenylic acid derivatives.
U.S. patent 5496546.
VOL. 40, 1996 INHIBITION OF MuLV WITH DNP-POLY[A] 2317