Search

We are developing a technology based on the combined application of riboflavin (RB) and light for inactivating pathogens in blood products while retaining the biological functions of the treated cells and proteins. Virus and bacteria reduction measured by tissue culture infectivity or colony formation with UV light alone and in combination with RB yield equivalent results. The effects of RB as a sensitizing agent on DNA in white cells, bacteria and viruses in combination with UV light exposure have been evaluated. UV-mediated DNA degradation in Jurkat T cells and leukocytes in plasma as measured by the FlowTACS assay was significantly increased in the presence of RB. Agarose gel electrophoretic analysis of DNA in Escherichia coli and leukocytes in plasma demonstrated enhanced DNA degradation in the presence of RB. UV light in combination with RB prevented the reactivation of lambda phage compared with samples irradiated in the absence of RB. UV-mediated oxidative damage in calf thymus DNA was also enhanced in the presence of RB. These observations clearly demonstrate that the presence of RB and UV light selectively enhances damage to the guanine bases in DNA. These data also suggest that the type and extent of damage to DNA for virus in the presence of RB and light make it less likely to be repaired by normal repair pathways available in host cells.

Methylene blue (MB), riboflavin (RB) and psoralen sensitizers (4' aminomethyl-4,5',8-trimethylpsoralen [AMT] and derivatives) are under study as sensitizers of viral inactivation of blood products such as plasma proteins, platelets and red cells, all of which lack genomic nucleic acid. To predict where these sensitizers accumulate in viruses and in cells, their relative affinities for calf thymus DNA, neutral and negatively charged phospholipids and albumin were determined by dialysis. MB has a strong affinity for nucleic acid and negatively charged phospholipid, but little affinity for albumin or neutral phospholipid. RB has modest affinity for nucleic acid and little affinity for albumin or either phospholipid. AMT has substantial affinity for nucleic acid, neutral and negatively charged phospholipids and albumin. Neither AMT nor RB binds to poly G, although MB has some affinity for this polymer. Evidence of association of RB with guanosine monophosphate, adenosine monophosphate and tryptophan methylester hydrochloride in PBS buffer in the presence and absence of formamide was obtained from nonlinear Stern-Volmer plots and shifts in the ground state absorption spectrum of RB.

The absorption and fluorescence spectra of 3-aminobenzo-1,2,4-triazine di-N-oxide (tirapazamine) have been recorded and exhibit a dependence on solvent that correlates with the Dimroth ET30 parameter. Time-dependent density functional theory calculations reveal that the transition of tirapazamine in the visible region is pi-->pi* in nature. The fluorescence lifetime is 98+/-2 ps in water. The fluorescence quantum yield is approximately 0.002 in water. The fluorescence of tirapazamine is efficiently quenched by electron donors via an electron-transfer process. Linear Stern-Volmer fluorescence quenching plots are observed with sodium azide, potassium thiocyanate, guanosine monophosphate and tryptophan (Trp) methyl ester hydrochloride. Guanosine monophosphate, tyrosine (Tyr) methyl ester hydrochloride and Trp methyl ester hydrochloride appear to quench the fluorescence at a rate greater than diffusion control implying that these substrates complex with tirapazamine in its ground state. This complexation was detected by absorption spectroscopy.

The photochemistry and photophysics of 3-amino-6-iodoacridine (Acr-I) was studied. Photolysis (350 nm) of Acr-I (free base) generates products consistent with a free radical intermediate in methanol, benzene and carbon tetrachloride. The Acr-I hydrochloride is shown to bind to calf thymus DNA and to the self-complementary dinucleotide cytidylyl-(3'-5')-guanosine (CpG) miniduplex in a manner similar to that of proflavine (Acr-NH2), a known DNA intercalator. The Acr-I is shown to more efficiently nick supercoiled plasmid DNA pBR322 upon 350 nm or 420 nm photolysis than Acr-NH2. The efficiency of Acr-I-sensitized DNA nicking is not oxygen dependent. Photolysis of the Acr-I/(CpG)2 complex leads to cleavage of the dinucleotide and to cytidine base release by selective damage to a specific ribose moiety. Dinucleotide cleavage occurs equally well in the presence or absence of oxygen, thereby eliminating a singlet oxygen- or peroxyl radical-mediated process. Photolysis of Acr-I in the presence of a mononucleotide (GMP) or a non-self-complementary dinucleotide (uridylyl-[3'-5']-cytidine-UpC) does not lead to fragmentation and base release. Similarly, photolysis of the Acr-NH2/(CpG)2 complex does not lead to fragmentation and base release. The data indicate that photolysis of an iodinated intercalator bound to CpG or plasmid DNA generates an intercalated aryl radical and that the reactive intermediate initiates a sequence of reactions that efficiently nick nucleic acids. The inactivation of lambda phage sensitized by Acr-I with UV (350 nm) light is oxygen independent but with visible (420 nm) light is strongly oxygen dependent. The Acr-I fluoresces more intensely when excited at 446 than at 376 nm. Thus, UV photolysis may lead to C-I bond homolysis and free radical formation, a process that is not energetically feasible with visible light. The results demonstrate the difficulty of extrapolating model studies involving simple molecules and DNA to understanding the mechanism of viral inactivation with a particular sensitizer.

5-Azido-8-alkoxy psoralens were synthesized. Laser flash photolysis (LFP: XeF, 351 nm, 55 mJ, 17 ns) of the azides in acetonitrile or benzene solution produces the triplet nitrene and a small amount of ketenimine. Laser flash photolysis of the azides in methanol or aqueous solution cleanly produces the triplet nitrene. In aqueous solution containing highly polymerized calf thymus DNA, LFP produces a mixture of triplet nitrene and ketenimine corresponding to photolysis of free and bound psoralen, respectively. The two transients decay slowly but at different rates. Assignment of the transient spectra were secured by matrix EPR and UV-visible spectroscopy. The triplet nitrene lifetime is the same in buffer and in the presence and absence of calf thymus DNA. The results explain why psoralen azides are unable to efficiently nick plasmid DNA pBR322 upon UV activation.

Amino or polyamino derivatives of naphthalene (N-H), anthracene (A-H) and 8-alkoxypsoralen (PSR-H) were prepared along with their monobrominated analogs (N-Br, A-Br and PSR-Br). The ammonium salts of these compounds are all water soluble and bind strongly to calf thymus DNA and to lambda phage, a double-helical DNA, protein-coated virus. Binding of the sensitizer to DNA occurs, presumably by a mixture of hydrophobic, intercalative and electrostatic interactions. Relative binding constants to calf thymus DNA and to lambda phage were measured by the ethidium bromide fluorescence quenching assay. In general the brominated analogs bind more tightly to calf thymus DNA and to the virus than to the nonhalogenated analogs. It is demonstrated that the brominated aromatics are much more effective at inactivating lambda phage upon photoactivation (lambda approximately 310 or 350 nm) than are their nonbrominated analogs. At identical sensitizer concentrations (by weight) and light flux N-Br, A-Br, and PSR-Br produce 5-6 more logs of viral inactivation than their nonbrominated counterparts (N-H, A-H and PSR-H, respectively). The bromine effect may originate from light-induced electron transfer and subsequent cleavage of the C-Br bond of the sensitizer radical anion bonds to form aryl radicals. Singlet oxygen cannot be responsible for the viral inactivation because the brominated sensitizers are equally effective in the presence and absence of oxygen. Dithiothreitol does not protect lambda phage from light-induced inactivation by the brominated sensitizer thereby demonstrating that the photogenerated reactive intermediates responsible for the effect are complexed to the virus and are not generated free in solution.

Exposure of lambda phage to 658 nm light in the presence of 5,10,15,20-tetrakis-(1-methyl-4-pyridyl)-21H,23H-porphine, tetra-p-tosylate leads to complete (greater than 7 logs) inactivation as measured by the plaque assay. The sensitizer without light and 658 nm photolysis of lambda phage in the absence of sensitizer do not lead to a measurable decrease in viral inactivity. Viral inactivation is not dependent upon the presence of oxygen.

The photolysis of para-methyl and para-thiomethylphenylazide at 77 K produces the corresponding triplet nitrenes which can be detected by electron paramagnetic resonance (EPR) spectroscopy. Photolysis of these azides in frozen toluene at 77 K leads to insertion of the nitrene into a benzylic C-H bond of the matrix in modest yields. Photolysis of iodinated aryl azides under these conditions does not produce triplet nitrenes that can be detected by EPR spectroscopy. In contrast to the para-methyl and para-thiomethyl substituted phenyl nitrenes, photo-induced coupling of iodo-substituted phenyl nitrenes to toluene proceeds in very poor yield.