BIOINORGANIC

Effect

CHEMiSTR Y 9,23-34

of Isonicotinic

on Rous Sarcoma

A. ANTONY

Acid

(I 978)

Hydrazide-Copper

Virus and its Genome

Complex

RNA

and T. RAMAKRISHNAN.

Microbiology and Cdl Biology Laboratoq, Rangalore, 560012, Irtdia PETER

23

MIKELENS,

JEAN

JACKSON,

Department of Microbiology, California 94143

/n&an Institute of Science,

and WARREN

LEVlNSON*

University of California, San Francisco,

ABSTRACT The

complex of the antituberculous drug, isonicotinic acid CINH). inhibits the RNA-dependent DNA polymerase of ROW sarcoma virus and inactivates its ability to malignantly transform chick embryo cells. The INH-copper complex binds to the 70s genome RNA of ROW sarcoma virus (RSV). which may account for its ability to inhibit the RNA-dependent DNA polymelxe. The complex binds RNA more effectively than DNA in contrast to M-IBT-copper complexes. which bind both types of nucleic acids equally. The bon-I~,polymers, poly rA and poly rU, are bound by the INH-copper complex to a greater extent than poly rC. Isonicotinic acid hydrazide alone and CuSO, alone bind neither DNA. RNA, poly (rA), poly (rU). nor poly (rC). However. CuS04 alone binds poly Crl); INH alone does not. In addition to viral DNA synthesis, chick-embryo cell DNA synthesis is inhibited by the INH~opper complex. The extent of inhibition of cellular DNA synthesis is greater than that of cellular RNA and protein synthesis. No selective inhibition of transformation in cells previously infected \\iith Rous sarcoma virus is observed. Copper

hydrazide

INTRODUCTION lsonicotinic acid hydrazide (INH) [ 11 and thiacetazone [2] . a thiosemicarbazone. inhibit the growth of Mvcohacterium tuberculosis and are effective in combined therapy against tuberculosis [3]. Thiosemlcarbazones. such as N-methyl isatin P-thiosemicarba~one (M-IBT). possess antiviral activity, also. For example, M-IBT inhibits vaccinia virus replication i-11. inhibits the RNAdependent DNA pulyrnerase of Rous sarcoma virus (RSV). and inactivates the ability of RSV to malignantly transform chick-e:nbryo cells [5]. N-methyl * Person 0 tilsevier

to vxhom correspondence North-Holland.

Inc.. 1976

should be addressed. OOOS-3061/‘78[0009-0023$1.75

24

A. AXTONY

ET AL.

isatin @-thiosemicarbazone also inactivates herpes simplex virus types 1 and 2 (61, RN.4 slow viruses [7] ) and arena viruses [S] _ The acriviry of 3%IBT against RSV is a function of copper since EDT-4 inhibits the action of the drug [5], the addition of copper sulfate enhances it purified. M-IBT-copper chelate complex is as active as ISI ’ and a preformed, ;%I-IBT itself (unpublished results). In addition. M-IBT copper complexes. but not M-IBT alone, bind to purified D-VA and R-X-4 [!9] and inhibit transfection by purified A phage DNA in E coli [IO] . Isonicotinic acid hydrazide aIso chelates copper [I I] 1 and there is evidence that the INH-copper comples inhibits the growth of M trrberczL-.si~ [ 12, I;]_ In view of these relationships we undertook experiments to determine whether INII or an ISH-copper comples inhibits the RSA-dependent DSA polymerase and the transforming ability of RSV rend whether the complex binds to nucieic acids in a manner similar to XI-IBT.

IMGTERIALS

AND METHODS

Virus and CelIs The B-77 strain (group C) of RSV was propagated described by Levinson et ai. [5] and its transforming formation as described by Levinson rind Rubin [ 141.

in chick-embryo ability assayed

cells as by focus

Chemicais The SOLC~S of the chemicals used in this study are: t.1) Sigma: IXII; (3) K K bboratories; hl-IBT; and (3) Baker; CuSO~. The IXH-copper comples of ISH ~3s synthesized by misin g equal volumes of 100 mM aqueous solutions and CUSO~ at room temperature. The resulting precipitate was wtished eshnustively with water to remove free IKH and CuSO1. The opticai spectrum of the comples \v’;fs similar to that previously described [IS] ivith an absorption maximum of 760 nip. Stock solutions of IKH*copper rzinples (I m,u/mi) in 100% dimethylsulfoside, of INH (IS-7 rng~ml or IO0 1~131) in water. and of copper su!fate (10 mg/ml or 30 mbl) in water were made and appropriate dimtions used as stock solution for the individual experiments. and

RNA-I’?pendent

DNA Polymerase

Assay

Five microliters of purified B-77 strain of RSV. suspended in 200 i.rl of solution 4 (see below) was exposed to the compounds by the addition of 1 h of stock SoiutiOn and incubated at 37= for 30 min_ A 5~1 sample was taken for transformation assay. and 20 ~1 ::f solutirrn B (see below) was added. The reaction misture was incubated at 37’ for 60 min and acid-precipitable counts were obtained by the addition of 0.5 ml of 0.01 M Na pyrophosphate and 2 ml

INH-Cu COMPLEX

AND ROUS SARCOMA

VIRUS

25

of 5% perchloric acid in 0.01M Na pyrophosphate. The precipitate was filtered, dried, and counted in a liquid scintillation spectrometer. Solution A contained 10 mM phosphate buffer of pH 7.4, 150 mM NaCI, 10 mM MgC12, and 20 k&i dATP, dGTP, dCTP. Solution B contained 90 iul of aH TTP (500 Fe/ml; sp. act. 17 cu/mmol). 45 ~1 of 1% NP40, and 5 ~1 of 3mercaptoethanol. The final concentrations of these reagents in the reaction mixture were 7.5 cu/ml, 00.3% and 0.3%. respectively. Additional information is described by Levinson et al. [5]. CelIular

DNA. RNA. and Protein

Synthesis

Chick cells were seeded at 1 X lo6 cells per 50-mm dish in medium 199 plus 4% calf serum. After overnight incubation at 38*. the cells were exposed to the drug for 10 min and then the appropriate isotope was added and incubated for an additional 30 min The radioactive metabolites (Schwartz/Mann) used were at 0.2 I-cc/ml. 3H-uridine (sp. act. 23 3H-thymidine (sp. act. 40 cu/mmol) at 2 &ml. The cultures cu/mmol) at 0.2 &ml, and 3H protein hy drolysate were processed for counting by solubilization in 1% SDS and acid precipitation with 10% TCA. The precipitate was trapped on glass fiber filters, dried. and counted in a liquid scintillation spectrometer. Nucleic-acid

Binding

Assay

ce!l DNA (2 1.8 cpm/pg) and 32P-SV Nucleic acids used were 3H-HeLa 70s RNA (300 cpm/pg) prepared in this laboratory and 3H ribohomopolymers rA (40.0 cpm/,ug), rC (13.3 cpm/&, rU (132 cpm/&, and rI (13.6 cpm/& purchased from Schwartz-Mann. Reactions contained nucleic acid in 0.5 ml of 0.05 M phosphate buffer, pH 8.0. to which appropriate volumes of stock solutions of either INH=copper comp1z.x or M-IBT in 100% DMSO or either 100 mM I-NH or 40 mM copper sulfate in HZ0 were then added. After incubation at 3” for 5 min, the solutions were filtered through 0.45 millipore filters. dried. and counted in a Beckman liquid scintillation spectrometer. No filter retention was seen with 10 ,ul DMSO. the largest volume of stock solution added. A background of 2% of counts was retained in the untreated control_ -4dditional information is described by Mikelens et al. [9] -

RESULTS Effect

of INH-Copper

Complex

on RSV

Exposure of purified RSV to INHmcopper complex (-2.5 ,ug/ml or 12.5 PM calculated on the basis of a 1:l comples) inhibits the polymerase activity by 91% and inactivates the transforming ability by greater than 99% (Tablel).

A. ANTONY ET AL.

26 TABLE 1 Inhibition

of RSV Polymerase and Transform@ IAN Copper Complex

Ability by

l

flg/mJ’

Inhibitionb Polymerase Activity (%)

Cu

2.5

91

Cu

1.25

Compound INH INH

l

l

INH Cllsoa

Inactivationb Transforming Activity (%) >99

0

0

137.0

40

70

2.5

30

0

0

0

DMSO (0.5%)

a The unit yg/ml is presented since the precise molecular weight of the complex is unknown. For a I : 1 complex, 2.5 Irg/ml INH copper is 12.5 MM. b Untreated controls had approximately 10,000 cpm incorporated in the polymerase assay and approximately 200 foci in the transformation assay in replicate experiments. l

Exposure to 1.25 &ml of the complex has no inhibitory effect. A sample of INH*copper complex (prepared by Dr. John Cymmerman Craig of this institution), which inhibited IIL tuberculosis growth, has similar activity. The formation of a copper complex clearly potentiates the activity of INH since exposure to 137 pg/ml of INH (10-a M) is required to inhibit the polymerase activity by 40% and focus formation by 70%. Little, if any, effect of copper sulfate at 2.5 /cg/ml or of the solvent, DMSO, is seen. Since previous studies with M-IBT [S] showed that EDTA prevented inactivation, we tested the effect of exposure of the virus to the complex in the presence of either 10v3 M EDTA. This compound prevents the action of the drug on both enzyme activity and transformation also (data not shown). The kinetics of inactivation of RSV by INH.copper complex are depicted in Fig. 1. The 37% survival dose for transforming ability is approximately 1 min. 30 s using 5 kg/ml of the drug. A small proportion (- 4%) of the population may be more resistant to the drug since the slope of the curve decreases after 96% of virus has been inactivated. In view of the significant inactivation of the transforming ability of RSV by the INH*copper complex, its ability to inhibit transformation of chick-embryo cells after infection was tested. Two types of experiments were carried out, in which the drug was added only during the: (1) first 12 h after infection and

INH-CuCOMPLEXAND

ROUSSARCOMAVIRUS

27

r

FIG. 1. Kinetics of inhibition of PSV by INH*copper complex. The virus was exposed to the drug as described in the Material and Methods section. At varying times after drug addition, a 5X sample was removed and diluted to 1: 100 prior to assay for transformation by focus formation. No heat inactivation of the virus occurred during the 1 S-min incubation time.

6.day assay period. The data in Table 2 show that there is no concentration of the INH*copper complex or of INH itself that inhibits focus formation that does not visibly decrease cell growth. (2) entire

Effect of INH*Copper Complex on Ceil Synthesis In view of the toxic effects observed in Table 2. we investigated the effect of INH*copper C.I cell DNA, RNA, and protein synthesis. The data in Fig. 2 demonstrate t lat exposure of the cell to 20 ~g/ml of INH*copper complex !‘or 10 min inhibi s DNA synthesis by 72% but has little effect on RNA and protein synthesis. Mgher concentrations oi compiex do inhibit RNA and protein synthesis, but DY.4 synthesis is consistently more sensitive. A 10 He/ml complex for as long as 16 h has no effect on synthesis of any of the three macrqnolecules. No effect of 137 pg/ml of INH, 40 pg/ml of CuS04. or 2% DMSO on these parameters was seen.

A. ANTONY ET AL.

28 TABLE Effect

of INH-Copper

Complex

2 on Focus

Formation

by RSV”

12-h Exposure

Compound

Inhibition Transforming Activity (5%)

Pg/mlb

None

-

DMSO (4%)

I5 98 25 0 0

!NH-Cu Ii\!H*Cu C&O4 Ii’.

H

40 20 40 137 6-day

Xone DhfSO (2%) iNH-Cu INHaCu INH.Cu cuso‘$ INH INH

Condition of Cells NOlTIld

Normal Sparse Normal NOXTfld

Normal

Exposure 10 98 80 0 0 98 0

40 20 10 IO 137 69

NOrItld NOlTfld

Sparse Sparse NOl-I-Ild

Normal Sparse NORTlal

o Chick-embryo cells were infected with approximately 200 FFU RSV in 0.2 m!, which was ailowed to absorb for 30 min at 37”_ In the 12-h exposure experiment the drug was added to 5 ml of fluid growth medium and the cultures incubated for I2 h at 38O. The medium was then discarded and the cultures overlaid with S-ml agar containing growth medium without drug. In the B-day exposure the drug was inchtdcd in the S-ml agar containing growth r*,::diurn that was added to the cultures di:._.= ;,:tion of INH.copper complexes in the inactivation of RSV was investigated. Since we have shown [9] that M-IBT copper complexes bind to DNA and RNA. we tested the possibility that INH or its copper complex could bind to nucleic acids_ As shown in Tabie 3, INHmcopper complex binds to 70s RSV RNA but not to HeLa cell DNA. isonicotinic acid hydrazide itself does not

INH-Cu

COhlPLEX

AND ROUS

SARCOMA

29

VIRUS

FIG. 2. Inhibition of cell synthesis by NH-copper complex. The cells were exposed to 20 pgjml, 30 pg/ml. and 40 ,ug/ml ISH*copper complex and the extent of DNA, RSA, and protein synthesis determined as described in the Material and Methods section. Untreated celIs incorported approximately 43,000 cpm 3H thymidinc. Ii,000 cpm 3H uridine, and 9900 cpm 3H amino acids (line A. RNA: line B. protein; line C. DNA).

bind

to

either

more

active

DNA

or

RNA.

in their nucleic-acid

The

M-IBT-copper

binding

ability

complexes

are significantly

than are INH*copper

complexes.

As described in Table 3, the M-IBTcopper complexes cause the binding of 100% of acid-precipitable counts added. and DNA and RNA are retained equally. In view of differential binding of RNA and DNA by the NH-copper complex, we explored the nature of this specificity in more detail. The data in Fig. 3 demonstrate that over a wide range of concentrations of preformed I;\I’H*copper complex. RSA is bound to a greater extent than is DNA. The base specificity of the interaction with ribohomopolymers \vas esamined also. The data in Table 4 demonstrate that poly (rA), poly (ri). and poly (rU) are bound at 86-8870. whereas poly (rC) is bound only 51%_ Isonicotinic acid hydrazide alone and CuS04 at greater than IO-fold higher concentration have no binding abihty except for the effective binding of poly (rIj by CuS04_

A. ANTONY ET AL.

30 TABLE 3 Binding of DNA and RNA by INHCoppet

Complex

Percent cpm Bound’ ccglmla

DNA

RNA

1s

2

16

137

2

2

M-IBT + CuSOl

10

100

94

M-IBT

10

2

2

cuso*

IO; 100

2

2

Compound INH*Cu INH

a The unit pg/ml is presented since the precise molecular weight of the complex is unknown. For a 1: 1 complex, IS MB/ml of INHvopper is 75 PM. For INH, 137 yg/ml is 10e3 M, M-IBT 10 fig/ml is 40 PM, and for CuS04, 10 pg/ml is 40 PM. b Total (100%)

binding represents ‘Lie number of 10% TCA precipitable counts (e.g., ca. 6000 cpm of 3H DNA and 6000 cpm of 32P-RNA).

TABLE 4 Binding of Ribohomopolymers

by INH-Copper

Complex

Percent cpm Boundb Compound

rA

r1

rC

rU

30

88

86

51

87

INH

37s

2

2

2

L

CUSO, cuso,

375 75

2

100 50

2

-

INH-Cu

Ccg/mla

-

2 -

a The unit Pg/ml is presented since the precise molecular weight of the complex is unknown. For a 1: 1 complex, 30 I_cg/ml is 150 PM. For INH, 375 pg/ml is 2.7 X low3 M, and for CUSO~, 375 pg./ml is 1.5 X 10s3 M. b Total (100%) binding represents the amount of cpm bound by 40 PM

M-IBT plus 40 PM CuSO, (e.g., ca. 4800 cpm rA, 4850 cpm r1, 4200 cpm rC, and 3300 cpm rU).

INH-Cu COMPLEX AND ROUS SARCOMA VIRUS

31

i

,,__--415

20

I

I

1

I

40

60

80

100

IIIH.Cu

t pcj/fi~

CONCENTPATlON

1

1

FIG. 3. Binding of DNA and RNA hy INH*copper complex. Approximately 6000 cpm of 3H HeLa cell DNA and 32P-RSV 70s RNA was exposed to the indicated concentrations of INH*copper complex and binding measured as described in the Material and Meth+!ds section. (Line A, RNA; line B, DNA).

Since INH alone is the effective antitubercular agent, it was of interest to dete-mine the nucleic-acid binding of INH to which CUSO~ was added rather than the preformed complex. It can be seen in Fig. 4 that poly (rA) and poly (rU), but not poly (rC), are bound by INH to which 1500 PM CuS04 was added. Rous sarcoma virus 70s RNA was bound as expected; however, preformed INH=copper complexes are more effective in binding DNA since 500 PM (100 Yg/ml) preformed complex bound 80 % of the DNA (Fig. 3), whereas 1400 PM (196 pg/ml) INH plus ‘00 .uM (375 ugjml) bound no DNA (Fig. 4). The binding of poly rI is not described since control experiments with 1500 I.AM CuSO4 only showed 100% binding of his homopolyer. In contrast to INH snd CuSO.,, M-IBT binding of DNA and RNA is equally effective whether added separately as decribed in Table 4 or as a preformed M-IBT-copper complex Anpublished observation).

A. ANTONY ET AL.

32

CONCENTRdTIUN

INN

(

ptl 1

FIG. 4. Binding of RNA and ribohomopolymers by INH and CuSO, added separately_ “*P-ROW sarcoma virus 70s RNA, 3H-HeLa cell DNA, and 3H ribohomopolymers, r-4, rC, and rU were exposed to the indicated concentration of INH followed by 1500 yM CuS04 and the binding measured as described in the Material Methods section The cpm used are described in second footnotes of Tables 3 and 4 (line A, rA; line B, RNA; line C, rU; line D, r-C; line E, DNA).

DISCUSSlON In this report we demonstrate RNA-dependent DNA polymerase

that the copper complex of RSV and inactivates

of INH inhibits focus formation

the by

this virus in ce;i culture. One possible mechanism by which the drugs could inactivate RSV is by binding to their nucleic acid. A second possible mechanism for the activity of these drugs relates to the observation that INH reacts with oxygen to form hydrogen peroxide [IS, 161 and free radicals that inactivate the transforming DNA of f3. subtilis [ 153 . It is difficult to evaluate the importance of ?his possibility for the INH=copper complex since the products of the reaction of the copper complex with oxygen have not been determined. We have found that catalase has no effect on the inactivation of these viruses, which makes thi; hypothesis less likely (unpublished results). In any case, the mechanism of action against the viruses is probably different from that against M. ru-

INHCu COMPLEX AND ROUS SARCOMA VIRUS

33

berculosis, in which it has been shown that the Y enzyme (perioxidase) of the bacillus is involved in both the uptake [ 171 and action [ 181 of the drug. The nature of the INH=copper complex has been investigated recently by NMR, ESR, and spectrophotometric techniques [ 19]_ It was found that a strong charge-transfer complex occurs between the hydrazine side chain and the cupric ion_ A very weak interaction between the ring nitrogen and the cupric ion was also found. The role of the copper in the antiviral activity of the complex is not clear. It is possible that copper is the inactivator and that the INH serves to enhance its permeability, transport, or binding. Alternatively, the copper may serve to form a sterically correct molecule that can interact with a site on the nucleic acid. The formation of complexes of varying !igand to metal ratios may be an important factor in the antivira1 activity. In support of hypothesis that copper per se may be involved are the findings that copper can degrade polyribonucieotides 1201 and cause a rapid loss of the infectivity of tobacco mosaic virus (TMV) infectious RNA [21 J _ Copp er binds also the DNA, particularly to G-C base pairs causing local denaturation [ 22, 231. Under certain conditions, copper(H) acetate dimer has the ability to selectively interact with ribonucleotides rather than deoxynucleotides [24] _ This latter finding is in accord with the specificity of the INH-copper complex for RNA. It is interesting to note that although the complex binds to RNA more efficiently than DNA, it is cellular DNA synthesis that is most sensitive to the drug. One possible explanation of this observation is that initiation of DNA synthesis, due to the presence of the drug on the RNA primer, is inhibited rather than elongation of the DNA. These findings raise the possibility of effective antiviral chemotherapy with INH, a relatively nontoxic compound in widespread use, or with its copper complex On one hand, the prospect does not seem promising since the complex dir; not selectively inhibit transformation of RSV-infected cells. On the tither !land, it is possible that its ability to inactivate RSV on contact could be exploited. The observation that inactivation of herpes viruses by proflavin and visible light is clinically effective 1251 lends support to the possibility that inactivating agents may prove useful. In addition, it will be of interest to determine whether patients receiving INH for long-term therapy against tuberculosis have a decreased incidence of malignancy or of some viral diseases. However, the interpretation of the malignancy data may be complicated by the observation that both INH and its major metabolite in man, I-acetyl-2-isonicotinoyl hydrazine, cause an increase in lung tumors in mice [26]. To our knowledge, no information is available regarding carcinogenesis by either compound in man, but liver damage, a prominent side effect of INH therapy, may be due to this metabolite [27].

A.ANTONYETAL.

34

The senior author (W. Levinson) gratefully acknowledges the support of the United States Educational Foundation in India during his sabbatical leave at The Indian Institute of Science. In addition, this work was supported by USPHS Grant CA 12705, Contract NO1 CP33aP3 within the Virus &izncerProgram of the National Cancer Institute, NIH PHS, and b.c American Cancer Society Grant VC- 7019.

REFERENCES 1.

W. Lott,

J. Bernstein.

B. Sternberg.

and H. Yale, Annu. Rev. Tuherc. Pdmon.

357 (1952). 2. R. Dgnovick, F. Pansy, G. Stryker, and J. Bernstein, 1. Bacterial. 59,667 3. K. Citron, Br, Med. J. I, 426 (1972). 4. D. B..der and P. Sadler. Br. J. Pharmacol. 15, 101 (1960). 5.

W. Levinson,

A. Faras, B. Woodson,

J. Jackson,

and J. M. Bishop.

Proc

Dis, 65.

(19501.

/Vat. Acad. Sci.

c?jA 70, 164 (1973). V. Coleman, B. Wsudson. A. Rabson, J. Lanier, 1. Wltchcr, and C. Dnwson. AWmicroh. Agents Chemother. 5, 398 (1974). 7. A. Huasc and W. Levinson, Biochem. Bioph.vs. Res. Commun. 51. 875 (I 973). 8. J. Logap, P. Fox, J. Morgan, A. Makohon,andC. Pfau, J. Gen Viroi. 28, 271 (1975) 9. P. Mikclens, 8, Woodson, 2nd W. Levmson. Bi,xhcm. Pharm. 25. 821 f 1976). 10. W. L+:vinson and R. Helling. Antimicrob. Agents Chrmoi~hcr. 9, 160 ( 1976). 11, S. Fallab and H. Erlenmayer, flelv. Chim. Acta 36,6 ( 1953). 12. J. Cymmermann-Craig, Nature 177.480 ( 1956). 13. hl. Rcibtzr and G. Rem&i. Jrch. Biochem. f3ioph)x 131, 655 ( 1969). 14. W. Levinson and H. Rubin, l’irolog)! 28, 533 ( 1966). IS. If. 1:rccsc, S. Sklarow, and I