The Effect of Azo Dyes on Myxovirus Neuraminidase and on Virus Multiplication

Transcription

1 J. hen. Virol. (1968), z, Printed in Great Britain 26I The Effect of Azo Dyes on Myxovirus Neuraminidase and on Virus Multiplication By H. BECHT AND R. DRZENIEK Institut fiir Virologie, Justus Liebig-Universitiit, Giessen, Germany (Accepted 17 October 1967) SUMMARY Congo red and trypan red inhibited Newcastle disease virus neuraminidase totally and fowl plague virus neuraminidase partially in vitro, probably by polyanionic effects. The inhibitors did not prevent viral growth when present during the adsorption period. Congo red inhibited virus release and disturbed the synthesis of haemagglutinin, but did not interfere with the formation of ribonucleoprotein antigens. Trypan red, however, had no influence on the multiplication of either virus. INTRODUCTION Some azo dyes are known to repress the multiplication of influenza virus (Dettori, 1964) and of Newcastle disease virus (Finkelstein, I96 0. This inhibitory effect of azo compounds on myxovirus multiplication and Finkelsteirt's observation that such substances do not only interfere with haemagglutination by Newcastle disease virus but also with the elution of this virus from red blood ceils prompted us to examine the mode of action of such substances during virus replication in tissue culture and their effect on viral neuraminidase in vitro. It was hoped that ago dyes might represent a group of chemical inhibitors for myxovirus neuraminidase which could also be employed for in vivo studies. METHODS Fowl plague virus strain ROSTOC~: and Newcastle disease virus strain ITALmN were used as representatives of the influenza and. parainfluenza groups of myxoviruses. The 2 azo compounds congo red (E. Merck AG, Darmstadt, Germany) and trypan red (Fluka AG Buchs, St Gailen, Switzerland) were chosen. Stock solutions were prepared in the appropriate buffers for the in vitro tests or in Eagle's or Earle's minimal medium when used in tissue culture. Neuraminidase activity was tested as described previously (Drzeniek, I967) in o.i M-phosphate buffer at ph 7"o. The inhibitory action of congo red and trypan red was examined by mixing increasing amounts of azo dye with a standard quantity of virus purified by differential centrifugation. This amount splits off about IO/zg. of sialic acid in I5 min. at 37, ph 7"0, from bovine sialolactose, corresponding to z x io -z enzyme units (Drzeniek, Seto & Rott, 1966). The final protein concentration of Newcastle disease virus was 0.0o3 %, of fowl plague virus 0.0o4%. The percentage of inhibition was calculated from the difference in enzymic activity between the control and the test samples containing the azo dye.

2 262 H. BECHT AND R. DRZENIEK Since congo red stays in the water phase during extraction, concentrations up to 2 IO -4~did not disturb the test procedure. Trypan red, on the other hand, had a maximal absorption at 5IO m# which overlaps partially with the extinction range for neuraminic acid measured at 552 m#. If appropriate blank controls of trypan red were subtracted from the measured values, concentrations up to 2 lo -4 M could be used. To evaluate whether the inhibitor would also prevent Newcastle disease virus neuraminidase from removing sialic acid bound to complex structures, red blood cells were used as a substrate. The following conditions were chosen for the test: 0"5 ml. of virus containing five times the amount of neuraminidase stated above were mixed with o.i ml. of congo red in tenfold concentrations and preincubated for 5 min. at 37 ; 0"5 ml. of packed human erythrocytes, blood group O, were added and incubated for I5 min. at 37. The mixture was cooled to o and centrifuged at IOOO g for 5 min. The supernatant fluid was deproteinized by addition of trichloracetic acid and assayed for free sialic acid. The effect on viral multiplication was tested by infecting chick embryo fibroblasts at a multiplicity of approximately IO. After an adsorption period of 30 min. the cells were washed and incubated with different concentrations of the azo compound, for 8 hr in the case of fowl plague virus and I4 hr with Newcastle disease virus. When the activity of the inhibitors during adsorption was to be tested, appropriate dilutions of the substances were mixed with virus harvested from fibroblast tissue cultures, which had been infected with low multiplicities and incubated with minimal medium. This virus suspension contained only 0.02 % of protein, which did not affect the inhibitory activity of the dyes. The virus was kept in contact with the inhibitors during the 30 min. adsorption period. Then the mixture was removed, the cells were washed and incubated with minimal medium free of inhibitor for 6 hr (fowl plague virus) or 9 to Io hr (Newcastle disease virus). Media and cells were assayed separately for infectivity and viral components (Rott, I965). Since, when dissolved in the medium, the two compounds block haemagglutination by Newcastle disease virus, this inhibition was reversed by the addition of bovine serum albumin. It had been ascertained in preliminary experiments that after the addition of one drop of a 20 % (w/v) albumin solution to medium containing congo red and of three drops to medium containing trypan red, haemagglutination patterns could be read. Sindbis virus, strain AR 86 (Mussgay & Rott, I964) was grown in chick embryo fibroblasts and used for infecting fibroblast monolayers at a multiplicity of approximately Io. The conditions for cultivation in the presence of congo red were the same as described above for the myxoviruses. Infectivity of the medium was assayed by plaque tests on chick fibroblasts. RESULTS Effect on viral neuraminidase Both azo dyes proved to be potent inhibitors of Newcastle disease virus neuraminidase. A final concentration of 4 x Io -~ M-(2"8 #g./ml.) congo red and 4 x ~o -5 M- (40/zg./ml.) trypan red was required for a complete inhibition of enzymic activity and correspondingly Io -~ M-(0"7 #g./ml.) of congo red and Io -5 M-00 #g./ml.) of trypan red for 50 % inhibition (Fig. I). In contrast to Newcastle disease virus the maximal inhibition obtained with fowl

3 Azo dyes and myxoviruses 263 plague virus was 64% at aconcentration of 6 Io -5 M-(42 Fg./ml.) of congo red. The 50 % end point was reached with 2 x Io -5 M-(I4 Fg./ml.) of congo red. With trypan red only 28 % inhibition could be achieved employing a 2 x io -4 M-(200 #g./ml.) solution. Higher concentrations of the dyes could not be employed, since they were poorly soluble or interfered with the test procedure. In the haemagglutination tests, on the other hand, concentrations of congo red and trypan red that bring about neuraminidase inhibition, block haemagglutination of chick erythrocytes by Newcastle disease virus but have no effect on haemagglutination by fowl plague virus. I I I o.=. E e~ o o I 30.& _ ,~ lo-s 10-= lo-a Molarity Fig. i. Inhibition of Newcastle disease virus and fowl plague virus neuraminidases by congo red and trypan red. 2 x io -s viral enzyme units were mixed with the azo dyes to give the final concentrations indicated on the abscissa in moles/1, and incubated with Ioo/zg. of bovine sialolactose for I5 min. at 37. The final volume was 0"5 ml.; o-i u-phosphate buffer ph 7"0 was used. The amount of sialic acid liberated in the presence and absence of the inhibitor was determined and plotted as per cent of inhibition of the test samples containing the azo dye compared to the controls. A, NDV plus congo red; A, NDV plus trypan red;, FPV plus congo red; 0, FPV plus trypan red. When erythrocytes were employed as a substrate, a final concentration of Io -~ M- (70 Fg./ml.) congo red inhibited 50 % of the enzymic activity of Newcastle disease virus neuraminidase. If, however, erythrocytes were incubated with congo red before virus was added, there was no inhibition of enzyme activity. In this case the inhibitor was bound by the proteins of the erythrocytes and thus did not prevent the enzyme acting on the substrate. Since polyanions are potent inhibitors of neuraminidase and their inhibitory effect can be reversed by proteins (Drzeniek, I966) and polycations

4 264 H. BECHT AND R. DRZENIEK (Drzeniek, in preparation) analogous conditions were tried in the azo dye + neuraminidase system. A minimal amount of x mg./ml, of egg albumin had to be added to the reaction mixture in order to reverse completely the inhibitory effect of congo red on Newcastle disease virus neuraminidase. For fowl plague virus 5 mg./ml, had to be used. With 0-2 mg./ml, no reversion of either enzyme was observed. Inhibition by trypan red was only slightly affected by 5 mg./ml, of ovalbumin. The polycationic substance protamine sulphate reactivated the neuraminidases at much lower concentrations than ovalbumin (Newcastle disease virus at 2 #g./ml. and fowl plague virus at 20 #g./ml). The higher amount of protamine sulphate needed for reactivation of fowl plague virus was due to the relatively high concentrations of congo red used for blocking this enzyme. 9~- I I I I I I I I ~ sf i o I I I I I Congo red (Fg./ml.) Fig. z. Effect of congo red on the multiplication of fowl plague virus. Chick fibroblasts were infected with allantoic fluid. After adsorption for 3o rain. the infected cells were incubated for 8 hr with medium containing the amount of congo red indicated on the abscissa in/zg./ ml. of medium. A, Plaque-forming units (medium); O, haemagglutinating units haemagglutinating units (cell-associated); A, ribonucleoprotein antigen, complementfixing units. Effect on myxovirus multiplication Congo red in concentrations above IOO #g./ml. had some toxic effect on fibroblast monolayers; after 5 hr some cells had rounded, were stained red, and the cell sheet had lost its coherence. Under the same experimental conditions cellular RNA and protein synthesis were not disturbed as measured by the incorporation rate of [~4C]uridine and [14C]leucine. With trypan red no morphological changes of the cells could be observed. The dose-response curves for fowl plague virus (Fig. 2) and Newcastle disease virus (Fig. 3) demonstrate that congo red interferes with the multiplication of both viruses. Increasing concentrations between to and 50 #g./ml. gradually reduced the haemag-

5 Azo dyes and myxoviruses 265 glutination and infectivity titres. At 5 /zg./ml. (Newcastle disease virus) and 75 #g./ml. (fowl plague virus) no haemagglutinating activity could be detected in the medium and the infectivity titres were reduced to their minimal levels. There was also a marked repression of the formation of haemagglutinin in the cells and consequently of cell- 81, 7 I I I I I I! 7 e~ o.2o A &,Ig.Am 6 5 "4 ~Q 4.= ~ 5 I I I I ~ I Congo red (#g./ml.) Fig. 3. Effect of congo red on the multiplication of Newcastle disease virus. Experimental conditions and graphical details are the same as for Fig. 2; the incubation period was I4 hr. I 200 Table I. Effect of trypan red on the multiplication of fowl plague virus and Newcastle disease viruses Haemagglutinating units (HAD-) and plaque-forming units (p.f.u.) were measured under experimental conditions analogous to those for Fig. 2 and 3. Fowl plague virus Newcastle disease virus Trypan red r ~ (#g.]ml.) HAU p.f.u. HAU p.f.u. O I28 4"0 X IO 7 8 3"05 X IO Z'7 X IO 7 8 2"8 X IO e "0 x io T 8 3"6 x io T "75 IOT 8 4"25 lo T x io T 8 2"4 x lo T I'6 x IO ~ 8 3"35 x to ~ associated infectivity. In contrast to fowl plague virus, which maintained a constant minimal haemagglutination, approximately I5o/zg./ml. of congo red completely suppressed the formation of Newcastle disease virus haemagglutinin in the cells. With both viruses, however, a practically constant amount of complement-fixing ribonucleoprotein antigen was found over the whole range of concentrations tested. In a further experiment multiplication of fowl plague virus was allowed to proceed

6 266 H. BECHT AND R. DRZENIEK normally for at least 4 hr. During this period all viral subunits and some infectious virus have been produced. When the medium was exchanged at this point of the growth cycle for a medium containing IOO/zg./ml. of congo red and this medium was assayed 4 hr later for haemagglutination and infectivity, it was found to contain none of the preformed virus. With trypan red, on the other hand, a normal or only slightly reduced amount of virus was released into the medium containing the inhibitor. Even concentrations up to 5oo #g./ml. (roughly ten times the amount that inhibits neuraminidase activity totally in vitro) did not interfere with the production of virus and its liberation into the medium (Table I). When either inhibitor in the same range of concentrations as in the dose-effect curves was kept in contact with the inoculum during the adsorption period, normal yields of both viruses were obtained upon incubation with normal medium. 9 I I i I I I I I I 7 ~ ~ I,, Q ~ 3 I I I I I I 7S s Congo red (#g./ml.) I I I Fig. 4. Effect of congo red on the multiplication of Sindbis virus. Dose-response expressed as p.f.u./ml, of medium. Experimental conditions as for Fig. 2; incubation period 8 hr. Inhibition of Sindbis virus multiplication Release of Sindbis virus, a representative of another group of lipid-containing RNA viruses, was also very sensitive to congo red (Fig. 4). The profile of the inhibition curve was very similar to the ones obtained with fowl plague and Newcastle disease viruses; 5o/zg./ml. of congo red reduced plaque formation by more than Ioo-fold. DISCUSSION The inhibition of neuraminidase by azo dyes can be explained as being a similar phenomenon to the inhibition of these enzymes by different polyanions, reversible by proteins and polycations. Since the two azo dyes tested do not pass through a dialysis membrane, one can assume that these substances are not in solution as monomers, but as micelles with polyanionic properties. The efficient reversion of the inhibited enzymic activity by proteins and particularly by the polycation protamine sulphate supports such a concept. Another common characteristic is that smaller amounts of polyanions and of the azo dyes were required to inhibit Newcastle disease virus neuraminidase

7 Azo dyes and myxoviruses 267 than the enzyme of fowl plague virus. In contrast to the typical polyanions, which are strong inhibitors in acidic solutions and practically ineffective at neutral ph, congo red and trypan red blocked the enzymes efficiently at ph 7"o. Although fowl plague virus neuraminidase is not completely inhibited by congo red and the results in this system would therefore not be fully conclusive for the enzyme's function, the complete blocking effect of minute amounts of both inhibitors on Newcastle disease virus neuraminidase in vitro and the lack of any influence during the adsorption period fully support the conclusion of Seto & Rott (I966) that viral neuraminidase does not play an essential part during the initial phase of myxovirus replication. The fact that congo red blocks the liberation of preformed virus from the ceils and in the presence of the inhibitor neither virus was released into the medium suggests that one of the functions of the enzyme is to assist in the release of virus particles. This would be in agreement with Seto & Rott's finding that viral release was blocked by a specific neuraminidase antibody. A specific inhibition of neuraminidase, however, cannot be the only mode of action of congo red since trypan red, another potent inhibitor of neuraminidase, which was present in concentrations completely inhibiting enzymic activity in vitro, did not interfere with the detachment of either virus from the cell membrane. This is again emphasized by an equally strong inhibition by congo red, of the multiplication of Sindbis virus, another lipid-containing virus which does not have a neuraminidase. The host cell is influenced by congo red to such an extent that besides viral release the formation of the haemagglutinating components of both viruses is disturbed. Complement-fixing ribonucleoprotein antigens, however, even Newcastle disease virus nucleoprotein, which is synthesized in the cytoplasm (Reda, Rott & Sch/ifer, 1964) can be demonstrated in equal amounts, regardless of the dose of congo red employed. This finding is a particularly strong indication that the effect of congo red is specific and not merely a toxic action. Different rates and sites of synthesis of the ribonucleoprotein antigens and the haemagglutinating subunit of the viral coat, which is finally assembled in the membrane region and needs lipids as binding material to establish its biologically active form, might explain why the formation of the inner component is not markedly affected. Our experimental results indicate that congo red and perhaps other members of the azo dye group of substances interfere with the final phase of the multiplication cycle of lipid-containing viruses and can cause an abortive type of infection in myxoviruses. We would like to thank Professor Rott for his stimulating interest. We are obliged to Dr Scholtissek for his help with the radio-isotope incorporation studies and to Miss U. Majer, Miss H. Krombach and Miss M. Orlich for their excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft. REFERENCES DETrom, R. 0964). Attivit~t antivirale di coloranti azoici. G. MicrobioL x2, 135. DP.ZENnm, R. (1966). Inhibition of neuraminidase by polyanions. Nature, Lond. 2ix, 12o5. D~ZENmlC, R. (I967). Differences in splitting capacity of virus and V. cholerae neuraminidase on sialic acid type substrates. Biochem. biophys. Res. Commun. 26, 63 I.

8 268 H. BECHT AND R. DRZENIEK DRZE~K, R., SETO, J. T. & ROar, R. (I966). Characterization of neuraminidases from myxoviruses. Biochim. biophys. Acta x28, 547. F~Y~Lsr~rN, R.A. (196I). Protection of chick embryos against Newcastle disease virus by trypan blue and other compounds. J. Immun. 87, 707. MUSSGAY, M. & Roar, R. (I964). Studies on the structure of a haemagglutinating component of a group A arbo virus (Sindbis). Virology 23, 573. ILEDA, I. M., Roar, R. & S ~rer, W. (I964). Fluorescent antibody studies with NDV-infected cell systems. Virology 22, 422. Roar, R. (I965). Untersuchingen tiber die Feinstruktur des infekti6sen Partikels der Newcastle disease und tiber die neben ibm auftretenden, nichtinfektit~sen, virusspeeifischen Einheiten. ZentbL VetMed. x2, 74. SErO, J. T. & Roar, R. (I966). Functional significance of sialidase during influenza virus multiplication. Virology 30, 73 I. (Received I5 July I967)