Antiviral Chemistry & Chemotherapy (1990) 1(1), 53-60

Modulation of the antiviral activity of poly (A-U) by ethidium bromide and propidium iodide J. M. Jamison,1,2 P. J. Bonilla1 and c-e. Tsai1* 1Department of Chemistry, Kent State University, Kent, Ohio 44242, USA. 2Department of Microbiology and Immunology, Northeastern Ohio Universities College of Medicine, Rootstown, Ohio 44272, USA. Summary The role of ethidium bromide (EB) and propidium iodide (PI) in modulating the antiviral and interferoninducing activities of poly(adenylate-uridylate) (poly (A-U» was examined using the human foreskin fibroblast-vesicular stomatitis virus (HSF-VSV) bioassay system in which the concentration of poly (A-U) was fixed at 0.05 mM or 0.2 mM while the EB or PI concentration was varied to produce variable EB (or PI)/ribonucleotide ratios ranging from 1:16 to 2:1. EB, PI and poly (A-U) tested individually were not efficacious antiviral agents. When poly (A-U) was combined with the ethidium bromide or propidium iodide the antiviral activity was potentiated 15- to 22-fold at EB (or Pl)/ribonucleotide ratios in the region of 1/4. The interferon-inducing activity of the EB (or Pl)/poly (A-U) combinations were equal to the sum of the interferoninducing activity of the poly (A-U) and the EB or (PI). These results indicate that the EB and PI potentiate the antiviral activity of the poly (A-U) without superinduction of interferon. The direct viral inactivation study demonstrated that EB, PI, poly (A-U) and the EB (or Pl)/poly (A-U) combinations did not inactivate the VSV at concentrations near the 50% viral inhibitory dose.

1972; 1976). Other polyribonucleotides, such as poly (adenylate-uridylate) (poly (A-U»,have also been tested as interferon inducers (De Clercq and Torrence, 1977). Poly (A-U) thermally activated in the presence of divalent cations (Ca2 +, Mg 2 +) has been shown to possess a potentiated antiviral activity and an enhanced resistance to endonuclease degradation (De Clercq et aI., 1971; De Clercq, 1977). Diederich and co-workers (1972; 1973) observed potentiated antiviral activity when cationic dyes were added to the dsRNA instead of the divalent cations. The fact that the dyes employed by these investigators exhibit a structural homology with intercalative acridines suggests that a dye-nucleic acid interaction may be responsible for the enhanced antiviral activity. To test this hypothesis, we have designed experiments to systematically examine and quantify the modulation of the antiviral and interferon-inducing activities of poly (A-U)and poly (I) • poly (C) by ethidium bromide and propidium iodide. Ethidium bromide and propidium iodide (Fig. 1) have been employed because the pharmacological actions of these phenanthridines stem from their ability to intercalate into double-stranded nucleic acids (Waring, 1965) and because the precise nature of the intercalation has been elucidated by a variety of physical techniques including single crystal X-ray analysis (Tsai, 1978; Tsai et al., 1975; Patel and Canuel, 1977).A parallel set of experiments was performed with ethidium bromide and propidium iodide in conjunction with poly (A-U) in an effort to elucidate the effect of a bulky, alkyl substituent and an additional positive charge of the propidium iodide to modulate the antiviral and interferon-inducing activities of the polyribonucleotides.

Introduction

Polyriboinosinic-polyribocytidylic acid (poly (I) • poly (C) has been shown to be an effective antiviral (Hill et al., 1971; 1972) and antitumour agent (Levy, 1980) in animals. However, poly (I) • poly (C) therapy is only marginally effective in humans because of its toxicity and lability in human serum (Greene et al., 1984). Investigators have employed chemically modified forms of poly (I) • poly (C), such as ampligen and poly (lCLC), in an effort to improve the therapeutic index of poly (I) • poly (C) (Carter et aI.,

Received 8 August, 1989; revised 13 October, 1989. 'For correspondence.

Results

The purpose of this study is to elicit the role of ethidium bromide and propidium iodide in modulating the antiviral and interferon-inducing activities of poly (A-U). The first set of experiments examines the antiviral activity of ethidium, propidium and poly (A-U) for concentrations ranging from 0.0008 mM to 0.4 mM. The antiviral activities expressed in terms of equivalent interferon titres and 50% viral inhibitory doses are shown in Table 1. The equivalent interferon titres for the ethidium bromide (EB), propidium iodide (PI) and poly (A-U) increase as a function of

54

J. M. Jamison, P. J. Bonilla and

c.-« Tsai

increasing ethidium, propidium or poly (A-U) concentration, until they plateau at a concentration of approximately 0.2mM. Poly (A-U) attains a maximum equivalent interferon titre of approximately 1300 Units mr' at a concentration of 0.4 mM, while ethidium and propidium produce optimum equivalent interferon titres of 400 Units ml- 1 and 670 Units rnl", respectively. Ethidium bromide and propidium iodide possess average 50% viral inhibitory doses of 15.4 ± 7.7 J.LM and 20.9 ± 7.8J.LM, while poly (A-U) exhibits

50% viral inhibitory doses of 7.4 ± 4.7 J.LM. These results indicate that poly (A-U), EB and PI are not efficacious antiviral agents. The second set of experiments has been formulated to elicit the role of the ethidium and propidium in modulating the antiviral activity of poly (A-U). The concentration of poly (A-U) is fixed at 0.05 mM and 0.2 mM, while the EB or PI concentration is varied to produce EB (or PI)/poly (A-U) ratios of 1/16, 1/8, 1/6, 1/4, 1/2, 1/1 and 2/1. Both EB and PI enhance the antiviral activity of poly (A-U) (Table 2) with the magnitude of enhancement depending on the drug/ ribonucleotide ratio as well as the concentration cif the poly (A-U). The equivalent interferon titres increase as the EB (or PI)/ribonucleotide ratios increase, reach a maximum at an EB (or PI)/ribonucleotide ratio in the region of 1/4; and remain constant or decrease slightly as the EB (or PI)/ribonucleotide ratios approach 2/1. Conversely, the 50% viral inhibitory doses (1050) decrease as the EB (or PI)/ribonucleotide ratios increase, reach a minimum at an EB (or PI)/ribonucleotide ratio in the region of 1/4, and increase as the EB (or PI)/ribonucleotide ratios approach 2/1. When the equivalent interferon tit res forthe 1/4 ratio of

Table 1. Antiviral activity of ethidium, propidium and poly (A-U).

Ethidium

NHz

Ethidium

R = C2H5

Propidium R

+

= (CH2)3NCH3(C2H5)2

Fig. 1. Chemical structure of ethidium and propidium.

Concentration (mM)

Equivalent interferon titre" (Units rnr")

0.40 0.20 0.10 0.050 0.025 0.Q13 0.0063 0.0031 0.0016 0.0008

400 330 270 180 130 ND ND ND ND ND

± ± ± ± ±

140' 100 100 120 80

Propidium

50% viral inhibitory dose b (J.LM)

Equivalent interferon titre" (Units ml- 1)

25.0 22.4 10.2 11.5 8.0 NI NI NI NI NI

670 400 270 170 120 ND ND ND NO ND

± ± ± ± ±

220 140' 100 50 40

50% viral inhibitory

Poly (A-U)

(J.LM)

Equivalent interferon titre" (Units rnr ')

31.5 25.0 20.1 16.2 11.5 NI NI NI NI NI

1300 1100 730 500 330 150 ND ND ND ND

dose"

± ± ± ± ± ±

410 460 160 250 100 60

50% viral inhibitory

dose" (J.LM) 16.0 9.0 6.6 5.6 3.9 3.1 NI NI NI NI

ND, not detected; NI, not inhibitory; " estimated error. Mean equivalent interferon titres are the arithmetic averages of six readings and mean 50% viral inhibitory doses are the geometric averages of six readings. The antiviral activities of ethidium bromide (EB), propidium iodide (PI) and poly (A-U) were tested for concentrations ranging from 0.0008 mM to 0.4 mM. Each of these EB, PI and poly (A-U) solutions was serially diluted with PBS in twelve two-fold dilutions to form test solutions. These test solutions were co-incubated with HSF cells for 3h at 37°C. Subsequently, the test solutions were removed and the cells were washed twice with 100J.L1 PBS, overlain with 50J.L1 EMEM (5% FCS), incubated for 17h at 37°C and challenged with VSV (m.o.i. = 0.1). Antiviral activity was evaluated by 50% CPE 17h post-infection. a. In the present assay, antiviral activity is evaluated on the same cells that were exposed to the drug (poly (A-U)).Therefore, the antiviral activity observed may be a composite of interferon induction, direct viral inactivation or the inhibition of cellular or viral processes. However, since the assay was calibrated with [3-interferon, the antiviral activity of the test agent may be directly compared to the amount of [3-interferon required to produce the same 50% CPE endpoint. b. The 50% viral inhibitory dose (ID5D) represents the concentration of the test agent found in the 50% CPE endpoint well. The average ID5D for EB, PI, and poly (A-Uj is 15.4J.LM ± 7.7 J.LM, 20.9J.LM ± 7.8J.LM and 7.4 J.LM ± 4.7 J.LM, respectively. Since this antiviral assay was performed using serial two-fold dilutions of the test agent, ID5D values that lie within a concentration range from two-fold greater than or two-fold less than the average ID5D are within the error of the technique and are considered to be equivalent to the average ID5D•

Phenanthridines and polyribonucleotide

55

Table 2. Antiviral activity of ethidium bromide-poly (A-U) and propidium iodide-poly (A-U) cotnbinations. EthidiumlO.05 mM Poly (A-U)

DIP

2/1 1/1

1/2 1/4 1/6 1/8

1/16

Equivalent interferon titre (Units mr ') 4300 5900 5900 8500 4300 3700 1300

± 1700 ± 1300 ± 1300 ± 3700 ± 1700 ± 2200 ± 410

Propidium/0.05 mM Poly (A-U)

EthidiumlO.2 mM Poly (A-U)

PropidiumlO.2 mM Poly (A-U)

DIP in 50% viral inhibitory doses (fLM)'

Equivalent interferon titre (Units ml-')

DIP in 50% viral inhibitory doses (fLM)'

Equivalent interferon titre (Units ml-')

DIP in 50% viral inhibitory doses (fLM)'

Equivalent interferon titre (Units rnr ')

0.9010.45 0.32/0.32 0.16/0.32 0.05/0.20 0.0710.42 0.0710.56 0.09/1.44

4800 5900 6400 6400 2700 1600 930

1.13/0.56 0.501.0.50 0.2010.40 0.1010.40 0.16/1.00 0.14/1.12 0.13/2.10

29000 25600 25600 27700 5300 2900 1500

1.4010.70 0.63/0.63 0.28/0.56 0.1010.40 0.33/1.98 0.4413.50

21300 23500 23500 25600 8500 5300 2000

± 1800 ± 1300 ± 2200· ± 2200· ± 830 ± 550· ± 460

± 10500 ± 9000· ± 9000· ± 7700 ± 1850 ± 650 ± 330

0.44/7.10

± 6600 ± 5200 ± 5200 ± 9000· ± 3300 ± 1700 ± 980

DIP in50% viral inhibitory doses (fLM)'

0.7010.35 0.32/0.32

0.16/0.32 0.07/0.28 0.14/0.88 0.18/1.44 0.35/5.60

DIP, drug/ribonucleotide ratio; ., estimated error. Mean equivalent interferon titres are the arithmetic averages of six readings and mean 50% viral inhibitory doses are the geometric averages of six readings. The roles of ethidium bromide (EB) and propidium iodide (PI) in modulating the antiviral activities of poly (A-U) were examined by experiments in which the concentration of the poly (A-U) was fixed at 0.05 mM or 0.2 mM, while EB (or PI)concentrations were varied to produce EB (or Pl)/poly (A-U) ratios ranging from 1:16 to 2:1. Each of these EB (or PI)/poly (A-U) solutions was serially diluted with PBS in twelve two-fold dilutions to form test solutions. These test solutions were co-incubated with HSF cells for 3 h at 37°C. SUbsequently, the test solutions were removed and the cells were washed twice with 100 fLl PBS, overlain with 50 fLlEMEM (5% FCS), incubated for 17 hat 37"C and challenged with VSV (rn.o.i, = 0.1). Antiviral activity was evaluated by 50% ePE 17 h post-infection. a. The drug/ribonucleotide ratio reflects the concentration of test agent and ribonucleotide found in the 50% ePE endpoint well (ID5 0) for each combination tested. The different ID50 ratios demonstrate that the degree of potentiation of antiviral activity is dependent on the DIP ratio, with a DIP ratio of 1/4 producing the most potentiation of antiviral activity.

the EB (or PI)/poly (A-U) combinations with fixed poly (A-U) concentrations of 0.2 mM are compared to the 1/4 ratio of those with fixed poly (A-U) concentrations of 0.05mM, it appears that the equivalent interferon tit res of the EB (or PI)/poly (A-U) combinations increase in an essentially linear fashion as a function of the poly (A-U) concentration, with the tit res of the former (0.2 mM) being approximately four times the equivalent interferon titres of the latter (0.05mM). In addition, the equivalent interferon titres of the EB/polyribonucleotide combinations are not significantly different from the equivalent interferon titres of the PI/polyribonucleotide combinations. These results suggest that the enhancement process is not affected by the alkyl substituent or the additional positive charge of the propidium iodide. Furthermore, the equivalent interferon titres of the EB (or PI)/polyribonucleotide combinations exceed the sum of the equivalent interferon titres of their constituents. For example, the equivalent interferon titre of the 0.2mM poly (A-U) alone is 1100 Units mr', while the equivalent interferon titre of the ethidium bromide alone at the concentration employed to produce an EB/ribonucleotide ratio of 1/4 is 180 Units mr'. When ethidium bromide is combined with poly (A-U) in a ratio of 1/4, the equivalent interferon titre is 27700 Units mr' (a 22-fold enhancement). The enhancement phenomenon is more striking when one examines the 1050 (50% viral inhibitory dose) values. The average 1050 for EB alone is 15Af,LM, while the average 1050 for poly.(A-U) alone is 704f,LM. When EB is combined with 0.2 mM poly (A-U) to produce an EB/poly (A-U) ratio of 1:4, the ratio of 1050 is 0.10/0040. These results demonstrate that when EB is combined with

poly (A-U), the 1050 of the poly (A-U) decreases from 704f,LM to 0040f,LM (an 18-fold decrease), while the 1050 of the EB decreases from 1504f,LM to 0.10 f,LM (a 154-fold decrease). A similar potentiation of antiviral activity is observed for the PI/polyribonucleotide combinations. This suggests a positive synergism between the poly (A-U) and the EB (or PI)at the concentrations employed in this study. For the poly (I) • poly (C) experiments, the concentration of poly (I) • poly (C) has been fixed at 0.2 mM, while the concentration of EB or PI is fixed at 0.05 mM, to produce an EB (or PI)/poly (I) • poly (C) ratio of 1/4. In contrast to the antiViral activity of the EB (or PI)/poly (A-U) combinations, the antiviral activity of poly (I) • poly (C) is not affected by the addition of EB or PI (Table 3). The equivalent interferon titre of 0.2 mM poly (I) • poly (C) alone is 14900 Units mr' (10 50 = 0.71 f,LM). When the 0.2 mM poly (I) • poly (C) is combined with 0.05 mM EB or 0.05 mM PI, the equivalent interferon (lFN) titres decrease to 8500 Units mr" (ratio of 1050 = 0.32/1.28) and 7500 Units (ratio of 1050 = 0.36/1044), respectively. These results suggest that potentiated antiviral activity occurs only when EB and PI are added to poly (A-U). The results shown in Table 4 demonstrate that 0.05 mM ethidium, 0.05 mM propidium, 0.2 mM poly (A-U) and the EB (or PI)/poly (A-U) combinations with a DIP ratio of 1/4 induce I:HFN. The interferon induced by the ethidium or propidium alone produced a 16% (p < 0.01) reduction in the number of plaques, while the poly (A-U) alone, the EB/poly (A-U) combination, the PI/poly (A-U) combination and the poly (I) • poly (C) produced a 28% (p < 0.001), a 31 % (p < 0.001), a 32% (p < 0.001) and a 100% reduction

mr'

56

J. M. Jamison, P. J. Bonilla and C.-c. Tsai

Table 3. Effect of ethidium and propidium on the antiviral activity of poly (I) • poly (C).

Test agent 0.20mM 0.05 mM 0.05mM 0.05 mM 0.05 mM

poly (I) • poly (C) alone ethidium bromide alone propidium iodide alone EB + 0.20 mM poly (I) • poly (C) PI + 0.20 mM poly (I) • poly (C)

Equivalent interferon titre (Units ml~1)

50% viral inhibitory dose (fLM)

14900 ± 230 ± 170 ± 8500 ± 7500 ±

0.71 11.50 16.20

5200 100 50 3300 2600

D/P in 50% viral inhibitory doses (fLM)

0.32/1.28 0.36/1.44

Mean equivalent interferon titres are the arithmetic averages of siX readings, and mean 50 % viral inhibitory doses are the geometric averages of six readings. The role of ethidium bromide (EB) and propidium iodide (PI) in modulating the antiviral activity of poly (I) • poly (C) was examined by experiments in whioh the concentration of the poly (I) • poly (C) was fixed at 0.2 mM, while EB or PI concentration was fixed at 0.05 mM to produce an EB {or PI)/poly (I) • poly (C) ratio of 1:4. Each of these EB {or PI)/poly (I) • poly (C) solutions was serially diluted with PBS in 12 two-fold dilutions to form test solutions. These test solutions were co-incubated with HSF cells for 3 h at 37°C. Subsequently, the test solutions were removed and the cells were washed twice with 100 fLl PBS, overlaid with 50 fLlEMEM (5% FCS), incubated for 17h at 3rC and challenged with VSV (m.o.i. = 0.1). Antiviral activity was evaluated by 50% CPE 17h post-infection. Table 4. Interferon characterization and neutralization.

Test agent Virus alone Virus + antibody 0.20 mM poly (A-U) alone 0.20mM poly (A-U) + antibody 0.05mM ethidium bromide alone 0.05mM ethidium bromide + antibody 0.05mM propidium iodide alone 0.05mM propidium iOdide + antibody 0.05mM EB + 0.20mM poly (A-U) 0.05mM EB + 0.20mM poly (A-U) + antibody 0.05mM PI + 0.20mM poly (A-U) 0.05mM PI + 0.20mM poly (A-U) + antibody 0.20mM poly (I) • poly (C) 0.20mM poly (I). poly (C) + antibody

Percent survival (% of control)

97 72 95 84 96 84 97 69 92 68 93

o 33

The 0.05 mM ethidium bromide (EB), 0.05 mM propidium iodide (PI), 0.2 mM poly (A-U) and the EB {or PI)/poly (A-U) combinations with a test agent/nucleotide ratio of 1/4 and a fixed poly (A-U) concentration of 0.2 mM were assayed in triplicate for their ability to induce l3-interferon (13-IFN). Each 35-mm-titre dish was seeded with 500000 HSF cells suspended in 2 ml MEM supplemented with 2% FCS and incubated at 37°C and 5% CO 2 for 24 h. EB or PI was combined with poly (A-U) to produce an EB {or PI)/poly (A-U) test solution with a test agent/nucleotide ratio of 1/4 and a fixed poly (A-U) concentration of 0.2 mM. Poly (A-U) alone was tested at a concentration of 0.2 mM. One ml of each test agent was combined with an equal volume of MEM (2% FCS) and was incubated on confluent monolayers of HSF cells. After 3h, each test agent was removed, the monolayer was washed exhaustively with PBS and overlain with 2 ml of MEM. After 24 h, the interferon-containing supernatant was divided into two 1-ml aliquots. The first aliquot was combined with 1 ml of MEM and was added to a second titre dish. To demonstrate that 13-IFN was produced, the second aliquot was combined with an equal volume of a 1:1600 dilution of NIH anti-I3-IFN antibody (G-028-501-568) for 1 h at 37°C and 5% CO 2 before being added to a second titre dish. After 24 h of incubation at 3rC and 5% CO 2 , the HSF cells were challenged with VSV (100 p.f.u./plate). After 1 h. the supernatant was removed and the cells were overlain with 0.5% methylcellulose. After a 48h incubation at 37°C and 5% CO 2 , the methylcellulose was aspirated off and the cells were stained with 0.5% crystal violet (in 70% ethanol) for 30 minutes. The culture plates were then air-dried and scored for plaques. Poly (I) • poly (C) (Po L. Biochemicals) was employed as the positive control while plain medium served as the negative control. The anti-I3-IFN antibody was mixed with VSV to determine the non-specific inactivation of the VSV by the antibody.

in the number of plaques, respectively. The addition of 1ml of a 1/1600 dilution of NIH (National Institutes of Health) human f3-IFN antiserum to an equal volume of supernatant from EB- or PI-treated cells, poly (I)• poly (C)-treated, poly (A-U)-treated and EB (or PI)/poly (A-U)-treated cells, destroyed 33% of the antiviral activity of the poly (I) • poly (C) and essentially all of the antiviral activity of the other test agents. In this study, 0.2 mM poly (I) • poly (C) exhibited an equivalent interferon titre of 14900 Units ml- 1 compared to an equivalent interferon titre of 27700 Units rnr' for the EB/poly (A-U) combinations and 1100 Units mr' for the poly (A-U) alone. This suggests that the antiviral activity of the EB/poly (A-U) combination is greater than or equal to the antiviral activity of the poly (I) • poly (C). However, the results from the interferon neutralization assay suggest that the interferon-inducing capability of the EB (or PI)/poly (A-U) combinations closely approximates the sum of the interferon-inducing capabilities of the poly (A-U)and the EB (or PI)employed. These results suggest that EB and PI potentiate the antiviral activity of the poly (A-U) without superinduction of interferon. As an initial step in the process of elucidating the mechanism responsible for the potentiated antiviral activity, EB, PI, poly (A-U) and the EB (or PI)/poly (A-U) combinations have been tested for their ability to directly inactivate vesicular stomatitis virus (ySV)(Table5). Neither EB nor PI alone inactivates VSV, while the 0.2 mM poly (A-U)alone and the EB(or PI)/poly (A-U)combinations with a D/P ratio of 1/4 and a fixed poly (A-U) concentration of 0.2mM completely inactivate the VSV. However, poly (A-U) alone and the EB (or PI)/poly (A-U) combinations do not directly inactivate the VSV when they are assayed at concentrations four times greater than the concentration in the 50% cytopathic effect (CPE) endpoint well. These results illustrate that direct viral inactivation by EB, PI and poly (A-U) is not responsible for the enhanced antiviral

Phenanthridines and polyribonucleotide Table 5. Direct viral inactivation of the test agents and test agenVpoly (A-U) combinations. Test agent Poly (A-U) Poly (A-U) Ethidium Ethidium Ethidiurnlpoly (A-U) Ethidium/poly (A-U) Propidium Propidiurn Propidiurnlpoly (A-U) Propidium/poly (A-U)

DIP ratio

1/4 1/4

1/4 1/4

Concentration (mM) 0.20 0.0016 0.05 0.0004 0.05/0.20 0.0004/0.0016 0.05 0.0004 0.05/0.20 0.0004/0.0016

57

Percent survival (% of control)

o 100 100 100

o 100 100 100

o 100

DIP, drug/ribonucleotide ratio. The 0.05 mM ethidium bromide (EB), 0.05 mM propidium iodide (PI) and 0.2 mM poly (AcU)alone and the EB (or PI)/poly (A-U) combinations with a test agenVnucleotide ratio of 1:4 and a fixed poly (A-U) concentration of 0.2 mM were tested for their ability to directly inactivate the VSV. The test agents were assayed in triplicate by combining them with an equal volume of MEM (2%) containing 100 p.f.u. ofVSV and then incubating the test agent VSV mixture for 2 h at 37"C. SUbsequently, each test agenWSV solution was subdivided into three equal aJiquots and added to three 35-mm-titre dishes. After 1 h, the supernatant was removed and the cells were overlaid with 0.5% methylcellulose. After a 48h incubation at 37"C and 5% CO2 , the methylcellulose was aspirated off and the cells stained with 0.5% crystal violet and were scored for plaques. Inactivation was expressed as percent of control with 100% of control representing no inactivation and 0% of control representing complete inactivation. The direct viral inactivation ofVSV by the EB (or PI)/poly (A-U) combinations was also evaluated at concentrations four times greater than the concentration in the 50% CPE endpoint well.

activity of the EB (or PI)/poly (A-U) combinations at concentrations near the ID50 . Discussion

The observation that ethidium bromide and propidium iodide potentiate the antiviral activity of poly (A-U) is consistent with the work of several investigators (Diederich et aI., 1972; 1973). De Clercq and co-workers (1971) have shown that preincubating poly (A-U) with Ca2 + or Mg2 + ions at temperatures greater than or equal to 3rC results in elevated antiviral activity, but not elevated interferon induction. Other investigators (Diederich et aI., 1972; 1973) have shown that cationic intercalative dyes potentiate the antiviral activity of poly (I) • poly (C). Sehgal and co-workers (1975) have demonstrated that while neutral red or chloroquine alone are not effective interferon inducers, treatment of human diploid fibroblasts with these dyes 2.5 to 3.5 hours after poly (I) • poly (C) treatment potentiated interferon induction 16- to 64-fold and 4- to 16-fold, respectively. The enhanced interferon induction has been attributed to an increase in the rate and duration of the period of interferon production induced by the drugs' inhibition of RNA and protein synthesis. A more recent study by Baglioni and Maroney (1981) has demonstrated that the application of EB/poly (A) • poly (U) or EB/poly (I) • poly (C) combinations to monolayers of HeLa cells inhibits the activation of 2' ,5'-0Iigo(adenylic acid) synthetase (2',5'-0Iigo A synthetase) as well as the protein kinase. The inactivation appears to be a function of the

binding of the EB to the double-stranded RNA (dsRNA) with complete inhibition of the synthetase occurring at an EB/ribonucleotide ratio of 1116 (one molecule of EB per eight base pairs of poly (A) • poly (U)). In contrast to the work of Baglioni and Maroney (1981), the results of our present study demonstrate that EB and PI potentiate the antiviral activity of poly (A-U)with optimal potentiation of the antiviral activity occurring at a DIP ratio of 1/4 (one EB or PI molecule per four nucleotides). In the study of Baglioni and Maroney, the addition of EB and dsRNA to HeLa cells inhibits the development of an antiviral state by 2',5'-0Iigo A synthetase and protein kinase, while in our study, co-administration of EB with poly (A-U) potentiates the antiviral state and co-administration of EB with poly (I) • poly (C) does not potentiate the antiviral state. These results suggest two possibilities: (i)ethidium bromide interacts differently with the poly (A-U) than with the poly (I) • poly (C); (ii) ethidium bromide potentiates the antiviral state by mechanisms other than the 2 /,5' "oligo A synthetase or protein kinase pathways. Baglioni and co-workers (1982) have proposed three potential antiviral mechanisms for interferon-treated virusinfected cells: (i) activation of a latent endonuclease by 2 /,5/-0Iigo A and the degradation of viral RNA; (ii) activation of a protein kinase that phosphorylates the ct subunit initiation factor e1F-2 and inhibits protein synthesis; (iii) activation of an inhibitor of cap methylation that prevents association of the viral mRNA with the ribosome (Baglioni et al., 1982). If the ethidium bromide-poly (A-U) combination inhibits the 2',5'-0Iigo A synthetase and the protein

58

J. M. Jamison, P. J. Bonilla and C.-c. Tsai

kinase pathways, then inhibition of cap methylation may be responsible for the enhanced antiviral activity. Previous studies of interferon-treated HeLa cells infected with VSV (Baglioni et al., 1982; de Ferra and Baglioni, 1981) have demonstrated that while there is no degradation of cellular mRNA and no detectable 2',5'-0Iigo A formed, VSV replication is impaired because of the inhibition of guanosine cap methylation. While the reasons for the inhibition of guanosine cap methylation in interferon-treated cells have not been elucidated, VSV replication has been shown to be sensitive to inhibitors of RNA methylation (Caboche and La Bonnardiere, 1979). Although it has not been demonstrated that ethidium bromide directly inhibits RNA methylation, ethidium bromide has been shown to inhibit nucleolar 45S RNA synthesis, to decrease the stability of ribosomal structure (Survanarayana and Burma, 1975) and to intercalate into the double-stranded regions of tRNA (Waring, 1964; 1965), 5S RNA from ESCherichia coli (Gray and Saunders, 1971) and dsRNA from Penicillium chrysogenum (Oouthart et al., 1973; Gatti et al., 1975; Ballesta et aI., 1976). The results of these studies suggest that ethidium bromide's ability to impair the processing of the viral RNA may be responsible for the enhanced antiviral activity. This study has shown that while poly (A-U), ethidium bromide and propidium iodide alone are not effective antiviral agents, application of the EB (or PI)/poly (A-U) combinations to the human foreskin fibroblast (HSF) cells enhances the antiviral activities of poly (A-U). The degree of enhancement appears to be a function of the poly (A-U) concentration and the EB (or PI)/ribonucleotide ratio. In all cases, equivalent interferon titres of the EB (or PI)/poly (A-U)combinations with fixed poly (A-U) concentrations of 0.2 mM are approximately four-fold greater than those with fixed poly (A-U) concentrations of 0.05mM. Optimal equivalent interferon titres are observed at an EB (or PI)/ribonucleotide ratio in the region of 1/4 corresponding to one EB or PI molecule per four nucleotides. The observation that the equivalent interferon titres of the EB (or PI)/poly (A-U) combinations consistently exceed the sum of the equivalent interferon titres of each constituent suggests a synergism between the poly (A-U) and the ethidium bromide (or propidium iodide) at the concentrations employed in this study. In contrast to these results, the antiviral activity of poly (I) • poly (C) is not affected when it is combined with ethidium bromide or propidium iodide. The results of interferon neutralization and direct viral inactivation experiments indicate that while I)-interferon is induced by the EB (or PI)/poly (A-U) combinations, neither interferon superinduction nor direct viral inactivation appears to be responsible for the enhanced antiviral activity at concentrations near the 1050 , The enhanced antiviral activity may be a function of ethidium bromide's ability to impair the processing of Viral RNA.

The results of this study also demonstrate that when EB is combined with poly (A-U) in an EB/poly (A-U) ratio of 1:4, the 1050 of the poly (A-U) decreases 18-fold, and the 105 0 of the EB decreases 154-fold. It is interesting to note that the 1050 of the more toxic EB decreases to a much greater extent than the 1050 of the poly (A-U). The synergism demonstrated by this differential decrease in 50% antiviral inhibitory doses may be useful in antiviral chemotherapy since it may allow toxic antiviral agents to be delivered to cells in lower doses with effective antiviral activity and less toxicity.

Materials and Experimental procedures Poly (A-U) (Biogenics Research Corporation) and poly (I) • poly (C) (P. L. Biochemical) solutions were dissolved and diluted with buffer A (0.15M NaCI, 0.005 M NaH2P0 4 and Na2HP04, pH 7.2) at 4°C to produce the desired concentrations. The desired concentrations of ethidium bromide (Sigma Chemical Company) and propidium iodide (Sigma Chemical Company) were prepared by dissolving the ethidium bromide or propidium iodide in buffer A at 4°C. The concentrations of poly (A-U) (E260 (Molar Absorptivity) = 6.1 x 103 (Biogenics Research Corporation technical data sheet)), poly (I) • poly (C) (E260 = 4.9 X 103 (P. L. Biochemicals)), ethidium bromide (E480 = 5.45 X 103 (Le Pecq, 1971)) and propidium iodide (E494 = 5.83 x 103 (Patel and Canuel, 1977)) were determined spectrophotometrically using a Gilford Model 250 spectrophotometer. All test solutions were pipetted into sterile glass vials and were serially diluted with phosphatebuffered saline (PBS) in 12 two-fold dilutions. In the first set of experiments, each poly (A-U) solution, ethidium bromide or propidium iodide solution was assayed for its antiviral activity. In the second set of experiments, each ethidium bromide (EB)or propidium iodide (PI)solution was combined with poly (A-U)solutions to produce EB (or PI)/poly (A-U) ratios of 1/16, 1/8,1/6, 1/4, 1/2, 1/1 and 2/1 with fixed poly (A-U) concentrations of either 0.05 mM or 0.2 mM. These EB (or PI)/poly (A-U) solutions were tested for their antiviral activities. All ethidium bromide and propidium iodide solutions were prepared in a darkened laminar flow hood in which all experiments were performed in order to prevent photoactivation of the dye.

Antiviral activity assay Antiviral activity was tested using the methods that have been described in our previous paper (Jamison et al., 1988). Each well of a Corning 96-well, flat-bottom microtitre plate was seeded with 20000 human foreskin fibroblast (HSF) cells (M. A. Bioproducts, MHRF) suspended in 50 fLl of Eagle's minimum essential medium (MEM-GIBCO) supplemented with 5% fetal calf serum (FCSGIBCO), pH 7.4 and each titre plate was incubated at 3rC in 5% CO2 for 24 h. Once confluency was attained, the overlying growth medium was removed and the cells were washed twice with 100 fLl of PBS per well. Specifically, EB, PI, poly (A-U) or poly (I) • poly (C) alone or EB (or PI)/poly (A-U) combinations were co-incubated with HSF cells for 3 h. Subsequently, the test solutions were decanted and the HSF cells were washed twice with 100 fLl of PBS, overlaid with 50 fLl of minimum essential medium (MEM) for 17h and challenged for 17h with VSV at a

Phenanthridines and polyribonucleotide 59 multiplicity of infection (m.o.i.) of 0.1. After the viral challenge was completed, the 50% cytopathic effect (50% CPE) endpoint was assessed subjectively for each column with NIH human fibroblast interferon reference standard (G-023-902-527) (Research Reference Reagents Note no. 16A) and poly (I) e poly (C) reference standard (G·020-901-090) (Research Reference Reagents Note no. 13) in each titre plate providing corroboration.

Interferon Characterization and neutralization study . The ability of ethidium bromide, propidiuim iodide, poly (A-U) and the EB (or PI)/poly (A-U) combinations to induce interferon was tested in triplicate in 35-mm titre dishes using the method of De Clercq and Torrence (1977) as modified by Jamison and coworkers (1989).

Direct viral inactivation study The EB, PI and poly (A-U) alone and the EB (or PI)/poly (A-U) combinations were tested for their ability to directly inactivate the virus using the method of Bauer and co-workers (1981).The EB (or PI)/poly (A-U) combinations at a ratio of 1/4,0.2 mM poly (A-U) and 0.05mM EB or PI alone were assayed in triplicate by adding an equal volume of MEM (2% FCS) containing 100 plaqueforming units (pJ.u.) of VSV and then incubating the test agent VSV mixture for 2h at 3rC. Subsequently, each test agenWSV solution was subdivided into three equal aliquots and added to three 35-mm titre dishes. After 1h, the supernatant was removed and the cells were overlaid with 0.5% methylcellulose. After a 48 h incubation at 3rC and 5% CO2 , the methylcellulose was aspirated off and the cells stained with 0.5% crystal violet for 30 minutes. The crystal violet was decanted and the titre dishes were washed with distilled water, air-dried, and scored for plaques. Inactivation was expressed as a percentage of the control. The direct viral inactivation of VSV by the EB (or PI)/poly (A-U) combinations was also evaluated at concentrations four times greater than the concentration in the 50% CPE endpoint well.

Acknowledgements This work was supported, in part, by a grant R01-GM31257 from the National Institute of General Medical Sciences, National Institutes of Health.

References Baglioni, C., and Maroney, PA (1981) Inhibition of doublestranded ribonucleic acid activated protein kinase and 2' ,5'oligo(adenylic acid) polymerase by ethidium bromide. Biochemistry20: 758-762. Baglioni, C., Nilsen, T.W., Maroney, PA, and de Ferra, F. (1982) Molecular mechanisms of action of interferon. Tex Rep Bioi Med41: 471-478. Ballesta, J.P.G., Waring, M.J., and Vazquez, D. (1976) Specific release of ribosomal proteins by nucleic acid-intercalating agents. Nucl Acids Res 3: 1307-1322. Bauer, D.J., Selway, J.W.T., Batchelor, J.F., Tisdale, M., Caldwell, I.C., and Young, DAB. (1981) 4',6-dichloroflavin (BW683C), a new anti-rhinovirus compound. Nature 292: 369-370. Biogenics research corporation technical data sheet (1979).

Caboche, M., and La Bonnardiere, C. (1979)Vesicular stomatitis virus mRNA methylation in vivo: effect of cycloleucine, an inhibitor of S-adenosylmethionine biosynthesis, on viral transcription and translation. Virology93: 547-557. Carter, WA, Pitha, P.M., Marshall, L.W., Tazawa, I., Tazawa, S., and Ts'o, P.O.P. (1972) Structural requirements of the rln·rCn complex for induction of human interferon. J Mol Bioi 70: 567-587. Carter, WA, O'Malley, JA, Beeson, M., Cunningham, P., Kelvin, A, Vere-Hodge, A, Alderfer, J.L., and Ts'o, P.O.P. (1976) An integrated and comparative study of the antiviral effects and other biological properties of the polyinosinicepolycytidylic acid duplex and its mismatched analogues. III. Chronic effects and immunological features. Mol Pharmaco/12: 440-453. De Clercq, E. (1977) Interferon induction by polynucleotides: structure-function relationships. Tex Rep Bioi Med35: 29-38. De Clercq, E., and Torrence, P.F. (1977) Comparative study of various double-stranded RNAs as inducers of human interferon. J Gen Viro/37: 619-623. DeClercq, E.,Wells, R., Grant, R., and Merigan, T. (1971)Thermal activation of the antiviral activity of synthetic double-stranded polyribonucleotides. J Mol Bioi 56: 83-100. de Ferra, F., and Baglioni, C. (1981) Viral messenger RNA unmethylated in interferon-treated HeLa cells infected with vesicular stomatitis virus. Virology 112: 426-435. Diederich, J., Lodemann, E, and Wacker, A. (1972) New type of interferon inducer. Naturwissenschaften 59: 172-173. Diederich, J., Lodemann, E.,and Wacker, A (1973)Basic dyes are inducers of interferon-like activity in mice. Arch ges Virusforsch 40: 82-85. Douthart, R.J., Burnett, J.P., Beasley, FW., and Frank, B.H. (1973) Binding to double-stranded ribonucleic acids. Biochemistry 12: 214-219. Gatti, C., Houssier, C., and Fredericq, E. (1975) Binding of ethidium bromide to ribosomal RNA: absorption, fluorescence, circular and electric dichroism study. Biochim Biophys Acta 407: 308-319. Gray, P.N., and Saunders, G.F. (1971) Binding of ethidium bromide to 5-S ribosomal RNA. Biochim Biophys Acta 254: 60-77. Greene, J.J., Ts'o, P.O.P., Strayer, D.R., and Carter, WA (1984) Therapeutic applications of double-stranded RNAs. In Interferons and Their Applications. Chapter 26. Came, P.E., and Carter, WA (eds). New York: Springer Verlag, pp. 535-555. Hill, DA, Baron, S., Levy, H.B., Bellanti, J., Buckler, C.E, Cannellos, G., Carbone, P., Chanock, R.M., DeVita, V., Guggenheim, MA, Homan, E., Kapikian, AZ., Kirschstein, R.L., Mills, J., Perkins, J.C., Van Kirk, J.E, and Worthington, M. (1971)Clinical studies of induction of interferon by polyinosinicpolycytidylic acid. In Perspectives in Virology. Pollard, M. (ed.). New York: Academic Press, pp. 198-222. Hill, DA, Baron, S., Perkins, J.C., Worthington, M., Van Kirk, J.E., Mills, J., Kapikian, AZ., and Chanock, R.H. (1972)Evaluation of an interferon inducer in viral respiratory disease. JAMA 219: 1179-1184. Jamison, J.M., Flowers, D.G., Jamison, E.,Kitareewan, S., Krabill, K., Rosenthal, KS., and Tsai, C.-c. (1988) Enhancement of the antiviral and interferon-inducing activities of poly r(A-U) by carminic acid. Life Sci 42: 1477-1483. Jamison, J.M., Flowers, D.G., Kitareewan, S., Krabill, K., and Tsal, C.-c. (1989)Potentiation ofthe antiviral activity of poly r{A-U)by riboflavin, FAD and FMN. Cell Biollnt Rep 13: 215-222. Le Pecq, J.B. (1971) Use of ethidium bromide for separation and determination of nucleic acids of various conformational forms

60 J. M. Jamison, P. J. Bonilla and c-c. Tsai and the measurement of their associated enzymes. In Methods of BiochemicalAnalysis. Vol. 20. Glick, D. (ed.). New York: John Wiley and Sons, pp. 41-86. Levy, H.B. (1980) Induction of interferon by polynucleotides. In Interferons and Interferon Inducers - Clinical Applications. Stringfellow, D.A. (ed.). New York: Dekker, pp. 167-186. Patel, D.J., and Canuel, L.L. (1977) Biphasic helix-coil transition of the ethidium bromide-poly (dA-dT) and the propidium diiodidepoly (dA-dT) complexes. Stabilization of base-pair regions about the intercalation site. Biopolymers 16: 857-873. Research Reference Reagents Note no. 13. (1975) Research Resources Branch, National Institute of Allergy and Infectious Diseases, National Institutes of Health (1975). Research Reference Reagents Note no. 16A. (1979) J Bioi Stand 7: 383-395. Sehgal, P.B., Tamm, I., and Vilcek, J. (1975) Enhancement of human interferon production by neutral red and chloroquine: analysis of inhibition of protein degradation and macromolecular synthesis. J Exp Med 142: 1283-1300.

Survanarayana, J., and Burma, D.P. (1975) Effects of intercalating agents on the structure of the ribosome. Biochem Biophys Res Commun 65: 708-713. Tsai, C.-c. (1978) Stereochemistry of drug-nucleic acid interactions and its biological implications. In Annual Reports in Medicinal Chemistry. Clarke, F.H. (ed.). New York: Academic Press, Inc., pp. 318-326. Tsai, C.-c., Jain, S.C., and Sobell, H.M. (1975) Drug-nucleic acid interaction: X-ray crystallographic determination of an ethidium-dinucleoside monophosphate crystalline complex: ethidium 5-iodo-uridyl(3'-5')adenosine. Phil Trans R Soc London B 272: 137-146. Waring, M.J. (1964)Complex formation with DNA and inhibition of Escherichia coli RNA polymerase by ethidium bromide. Biochim Biophys Acta 87: 356--358. Waring, M.J. (1965) Complex formation between ethidium bromide and nucleic acids. J Mol Bioi 13: 269-282.