Search

The polycyclic compound hypericin, a known photodynamic agent, was investigated for antiviral activity in the presence and absence of light. The three viruses tested: murine cytomegalovirus; Sindbis virus; and human immunodeficiency virus type 1, were all susceptible to hypericin; but these antiviral activities were considerably enhanced, in a dose-dependent manner, by exposure to light.

Hypocrellin, a photodynamic perylene quinonoid isolated from the Chinese medicinal fungus Hypocrella bambuase, was evaluated for antiviral activity against the human immunodeficiency virus (HIV-1). Hypocrellin was phototoxic to HIV-1, almost as good as the structurally similar plant pigment hypericin, and like hypericin its activity required visible light. In contrast peroxyhypocrellin had little or no effect on the virus.

Two thiarubrines (dithiacyclohexadienes), which were isolated recently from species of Asteraceae, were investigated for light-mediated antiviral activity against human immunodeficiency virus (HIV-1). Both compounds (thiarubrines A and D) showed good anti-HIV activity, in micromolar concentrations, which was dependent upon UVA radiation. They showed no anti-viral activity in the dark, and only weak visible light-mediated activity, despite the significant absorbance of these compounds in the 480 nm region. The resulting inactivated virus was unable to replicate in cell cultures, as indicated by loss of cytopathogenicity, the absence of cells expressing HIV-1 antigens and lack of production of HIV-1 p24 protein. Although this antiviral activity was not as efficient as that shown by the photoactive plant thiophene, alpha-terthienyl, both thiarubrines demonstrated increased activity when the serum component of the virus stock was reduced. This suggested that a component of bovine serum interfered to some extent with the anti-HIV activity.

— Six different photosensitizers were compared for their ability to form cross-links in murine cellular DNA and murine cytomegalovirus DNA, in the presence of long wave UV radiation. The viral DNA was in the form of free DNA or intact virions. The compounds consisted of the linear furanocoumarins 8-methoxypsoralen (8-MOP) and isopimpinellin; the angular furanocoumarin, angelicin; the two furanochromones, visnagin and khellin; and the β-carboline alkaloid, harmine. Cross-linking was assessed by alkaline agarose gel electrophoresis and hydroxyapatite chromatography. 8-MOP produced extensive cross-linking (as expected), as did isopimpinellin. Visnagin produced less cross-linking, such that not all DNA molecules were affected at the concentrations used. Khellin, angelicin and harmine produced no detectable cross-linking. The same result was obtained for DNA which was treated in situ in the virion. To some degree there was a correlation between the amount of cross-linking and the relative potency of anti-MCMV infectivity. But other factors evidently contribute to the phototoxic effect of these compounds.

— Hemolysis, K+ leakage and acetylcholinesterase (AChE) inhibition in human erythrocytes were observed with certain naturally occurring polyacetylenes and a thiophene derivative, α-terthienyl. K' leakage, subsequent hemolysis and AChE inactivation by phenylheptatriyne (PHT), a phototoxic compound, were considerably enhanced by UV light (312–400 nm). The same was true with α-terthienyl and with certain other polyacetylenes. Oxygen enhanced AChE inactivation and hemolysis with α-terthienyl in light. With PHT, only AChE inhibition was significantly enhanced in oxygen. Falcarindiol, a non-phototoxic polyacetylene, did not inactivate this enzyme but caused hemolysis in the dark. Inhibition of AChE and hemolysis by these compounds appear to be unrelated phenomena. These results indicate that certain polyacetylenes are capable of damaging biological membranes in light, and others in dark.

— When irradiated with UV-A liposome bilayers composed of distearoyl phosphatidylcholine and containing phenylheptatriyne produced a free radical signal detected by electron spin resonance spectroscopy. The spectrum contained one broad peak characterized by a linewidth of 19 G and a g value of 2.0017. The rate of formation of the signal amplitude was dependent upon light intensity, concentration of PHT within the bilayer and concentration of the PHT-liposome suspension. Enhancement of the signal under an anaerobic atmosphere indicated a non-photodynamic mechanism for free radical formation. Once formed however, the free radical was stable to the presence of oxygen and decayed very slowly with a half-life of 8 h. Formation of the free radical species was demonstrated to be dependent upon a highly ordered lipid environment since incorporation of lysophosphatidylcholine, stearoyl which perturbs lipid packing, decreased the levels of the free radical species. On the other hand, when PHT was present in a liposome with a more fluid bilayer such as egg yolk phosphatidylcholine, the levels of free radical species were even lower. Conversely, the levels increased when cholesterol which increases lipid order, was incorporated into the egg yolk phosphatidylcholine liposomes. Cells of E. coli, B, containing PHT, produced a similar free radical signal upon irradiation demonstrating in vivo generation of free radicals from PHT.

— The photosensitizing activity of α-terthienyl, a naturally occurring compound from Tagetes has been studied. Fluence-response curves for the effect of α-terthienyl and near-UV radiation on Escherichia coli have been prepared. Photosensitization of E. coli was enhanced in aerobic as compared to anaerobic conditions and exogenous superoxide dismutase had a slight protective effect. Sodium azide, a quencher of 1O2, protected yeast cells from photosensitization with α-terthienyl. The available evidence suggests that α-terthienyl acts as a photodynamic sensitizer in vivo.

— Harmine, a β-carboline alkaloid, was found to be an efficient photoactive agent against the DNA-containing murine cytomegalovirus (MCMV) and the RNA containing Sindbis virus, a togavirus, both of which contain membranes. In contrast the host cells themselves were relatively resistant. The MCMV was investigated in more detail. Virus which had been inactivated by harmine plus UVA retained its structural integrity and was fully capable of penetrating host cells, following which the viral genome entered the cell nucleus. The viral genes which normally code for the switch-off in thymidine kinase activity arid for the normal cytopathic effects were not expressed; however the viral geneis responsible for the cytotoxic effect due to high multiplicities of infection was still expressed, indicating that there may be a differential block of some viral genes

The photodynamic effects of sc-terthienyl (scT) in near-UV light (UV-A) on Escherichiu coli showed close agreement with the light absorption of scT at different wavelengths suggesting that rT is the primary absorbing molecule responsible for the photosensitized reaction. Studies with DNA repair deficient mutants of E. coli indicated that the bactericidal action of otT/UV-A was not mediated by DNA damage, in direct contrast to the well-known photosensitizer, 8-methoxypsoralen. By using a closed borosilicate glass reaction vessel and various gas mixtures, it was demonstrated that photosensitization of both E. co(i and a more resistant bacterium, Pseudomonas aeruginosa, was absolutely dependent on the presence of oxygen. The rate of killing by otT/UV-A showed a rather small dependence on preincuba- tion temperatures. with quite rapid killing at 5°C. suggesting that the movement of otT across the cytoplasmic membrane of E. c-oli is not the rate limiting step in killing and perhaps is not even necessary for killing. Sodium dodecyl sulphate-polyacrylamide gels of cell membrane proteins after 15 and 30 min of treatment with rT/UV-A showed substantial random crosslinking of these proteins. The results taken overall suggest that rT is a photodynamic photosensitizer which exerts its primary effect at the level of the cytoplasmic membrane.

— The naturally occurring furoquinoline alkaloid, dictamnine, is phototoxic to Gram-positive bacteria, yeasts and filamentous fungi in the presence of long-wave UV light. Prior to photobinding to DNA dictamnine forms a complex with the macromolecule in the dark. Photobinding to DNA is shown with [3H]-dictamnine. In contradiction to an earlier finding, dictamnine proved to be incapable of forming interstrand crossiinkages with native DNA. This fact was ascertained by the use of methods additional to hydroxyapatite column chromatography, e.g. alkaline sucrose gradient centrifugation and hyperchromicity tests. The alkaloid therefore represents a new monofunctional photoreagent towards DNA in vitro. It is assumed that it binds through its furyl-C=C to DNA. The inability to form crosslinks may be attributed to the stability of the lateral aromatic nucleus. As no photoreactivity against human erythrocytes could be observed, it is possible that nucleic acids represent the major target of damage in vivo.

— Two analogues of α-terthienyl, namely 2, 5di(2'-thienyl) selenophene and 2, 5di(2'-thienyl) furan have been prepared and their phototoxicities toward several microorganisms have been compared. Dose response studies with Escherichia coli indicate that α-terthienyl is more active than these analogues. α-Terthienyl was shown to be the most effective of the three compounds in the photoinactivation of yeast alcohol dehydrogenase. Diagnostic tests showed the participation of singlet oxygen in the photosensitization to different extents by these three thiophenes.

— Murine cytomegalovirus, a herpes virus, was used as a model virus to investigate the mechanism of the anti-viral action of phenylheptatriyne in long wave ultraviolet light. The genome and proteins of the inactivated virus penetrated the nuclei of susceptible cells normally. Furthermore, the viral genome did not contain extra single strand breaks or cross-links. However, cells infected with the treated virus did not synthesize late viral proteins, as determined by polyacrylamide gel electrophoresis. nor did they synthesize late viral RNA and viral DNA according to nucleic acid hybridization tests. Thus the compound may interfere with an early viral function so that the replication cycle cannot proceed.

— We have studied the antiviral properties of three different groups of photo-sensitizers, viz. (i) various furyl compounds; (ii) β-carboline alkaloids; (iii) thiophenes and their acetylene derivatives. In general the antiviral potency of the furyl compounds correlated with their ability to produce DNA photoadducts. Among the naturally occurring β-carboline alkaloids, harmine was considerably more potent (in the presence of long wavelength UV radiation, UVA) than several other harmane-related compounds. Slight alterations in chemical structure had profound effects on their antiviral activities. Harmine was shown to inactivate the DNA-virus murine cytomegalovirus (MCMV) by inhibiting viral gene expression, although other targets may also exist. Several eudistomins, carboline derivatives isolated from a tunicate, were also photoactive against viruses. Various plant thiophenes and polyacetylenes were studied in detail. These compounds also required UVA for antiviral activity, and some of them were extremely potent against viruses with membranes, e.g. α-terthienyl, which showed significant activity at only 10-5μg/ml. When MCMV had been treated with α-terthienyl plus UVA, the virus retained its integrity and penetrated cells normally; but the virus did not replicate. More than 30 additional thiophenes have recently been evaluated, including many synthetic ones, and some of these are even more potent than a-terthienyl. We believe that certain thiophenes possess potential therapeutic value and should be tested against model virus infections in animals.

The mechanism of action of the phototoxic polyacetylene phenylheptatriyne (PHT) against viruses was investigated. For this purpose the membrane containing mouse cytomegalovirus (MCMV) was treated with PHT in long wave UV light, and several functions of the virus were then examined. Following this treatment MCMV retained its structural integrity, as revealed by electron microscopy, resistance to exogenous deoxyribonuclease, and sedimentation-velocity. The treated virus was also normal in its ability to penetrate into the nuclei of susceptible mouse cells. Replication of the virus was defective, however, in that infectious centers could not be established and no infectious progeny resulted. Two RNA-containing viruses, Sindbis and infectious hematopoietic necrosis virus (IHNV, a fish virus), both of which contain membranes, were inactivated by PHT in long wave UV as efficiently as MCMV. In contrast, the fish virus IPNV (infectious pancreatic necrosis virus), which does not possess a membrane, was resistant to PHT. Thus, it appears that PHT, in the presence of long wave UV, exerts its action via the viral membrane and as a consequence the viral genome, DNA or RNA, is unable to replicate.

— Five photoactive furanyl compounds were investigated for their activities against viruses. The two furanocoumarins used were 8-methoxypsoralen (8-MOP) and angelicin; two furanochromones, visnagin and khellin, and the furanoquinoline, dictamnine, were also used. The DNA-containing herpes virus murine cytomegalovirus (MCMV) and the RNA-containing togavirus, Sindbis virus, were the target viruses. All five compounds inactivated both viruses in the presence of UVA, although Sindbis virus was much less sensitive. The relative order of antiviral potency was 8-MOP > dictamnine > visnagin > angelicin > khellin. Dictamnine however was slightly more effective than 8-MOP against Sindbis virus. None of the treatments affected the structural integrity of MCMV, nor did they interfere with the normal transit of the virus into host cells or the localisation of the viral genome in the cell nucleus. Some early viral gene functions were expressed but the viruses did not replicate.

The UV-A mediated antiviral effect of harmine was investigated using murine cytomegalovirus (MCMV) as the target. Virus, which had been inactivated by harmine + UVA, was used to infect cultured mouse cells, and various stages in the viral replication cycle were examined. No viral protein synthesis or RNA synthesis (as measured by polyacrylamide gel electrophoresis or DNA-RNA hybridization) could be detected, and the viral DNA did not replicate (measured by DNA-DNA hybridization). In contrast virus which had been treated with harmine in the dark promoted a complete growth cycle in mouse cells. An attempt was made to identify the primary target of harmine + UVA activity by examining the bacteriophages T4 and M13, which unlike MCMV do not possess membranes. Both bacteriophages were sensitive, but the single-stranded DNA phage M13 was considerably more so. These results, together with others discussed in the text, suggest that DNA and possibly other macromolecules can serve as targets for harmine photoactivity.

The wide range of naturally occurring compounds from microorganisms and plants which are phototoxic in UV-A light (320-400 nm) includes cinnamyl esters, coumarins and furanocoumarins, furanochromones, benzofurans, alkaloids, based on tryptophan or phenylalanine, extended naptho-and anthraquinones, polyacetylenes and their thiophene derivatives. The cellular targets are cell membranes, e.g., acetylenes, or the nucleus, e.g., furanocoumarins. Compounds reacting with cell membranes in light either generate singlet oxygen, e.g., alpha-terthienyl, or react by a free radical mechanism, e.g., phenylheptatriyne, or do both. Those that react with nucleic acids intercalate with these macromolecules forming photoadducts, e.g., furanocoumarins, furanochromones and furanoquinolines. With others, such as the beta-carbolines, the explanation for their photogenotoxicity is unknown. A number of these natural photosensitizers have been examined with bacteria, yeasts, and viruses and the mechanisms of phototoxicity have been elucidated. Some of the sulfur-containing acetylenes may be useful in photochemotherapy.

— Five eudistomins, β-carboline derivatives isolated from a Caribbean tunicate, were tested for phototoxicity against several viruses, bacteria, yeast, and mammalian cells. The five compounds showed varying degrees of UVA dependant phototoxicity (i.e. long wavelength UV dependant) against murine cytomegalovirus (MCMV), Sindbis virus (SV) and mouse 3T3 cells, although the relative order of potency was the same for these three organisms. Eudistomin N was the most active (approximately the same as the β-carboline, harmine), while eudistomins M and O were moderately phototoxic, and H and I had little activity. To some degree the relative phototoxicity was correlated with fewer side chain substituents. A similar relative order of phototoxic potency was seen against phage T4, but in this case the magnitude of the effect was considerably reduced, in contrast to harmine. The antibacterial and antifungal activities were not correlated with antiviral effects, and some UVA-independent activities were seen. Thus the eudistomins may possess different mechanisms of action against different organisms, depending upon the presence or absence of UVA.