Photochemical reduction of 5-bromouracil by cysteine derivatives and coupling of 5-bromouracil to cystine derivatives

New interactions between tumor suppressor Fhit protein and a nonhydrolyzable analog of its AP4 A substrate

Fragile histidine triad protein (Fhit) is a protein which primarily hydrolyses dinucleoside polyphosphates. To investigate possible interactions between the protein and a substrate, we used a nonhydrolyzable phosphorothioate analog of Ap₄ A, containing 5-bromo-2'-deoxyuridine instead of one adenosine residue. Photocrosslinking, followed by LC-MS experiments, determined a complex in which the probe was covalently linked to the NDSIYEELQK peptide (residues 110-119). The peptide was located within the 'disordered' region, which is invisible in the known crystal structures of Fhit. This invisible and flexible part seems to play a role in the stabilization of the Fhit-substrate complex, which may be important for its tumor suppressor activity.

Enzymatic synthesis of long double-stranded DNA labeled with haloderivatives of nucleobases in a precisely pre-determined sequence

Background

Restriction endonucleases are widely applied in recombinant DNA technology. Among them, enzymes of class IIS, which cleave DNA beyond recognition sites, are especially useful. We use BsaI enzyme for the pinpoint introduction of halogen nucleobases into DNA. This has been done for the purpose of anticancer radio- and phototherapy that is our long-term objective.

Results

An enzymatic method for synthesizing long double-stranded DNA labeled with the halogen derivatives of nucleobases (Hal-NBs) with 1-bp accuracy has been put forward and successfully tested on three different DNA fragments containing the 5-bromouracil (5-BrU) residue. The protocol assumes enzymatic cleavage of two Polymerase-Chain-Reaction (PCR) fragments containing two recognition sequences for the same or different class IIS restriction endonucleases, where each PCR fragment has a partially complementary cleavage site. These sites are introduced using synthetic DNA primers or are naturally present in the sequence used. The cleavage sites are not compatible, and therefore not susceptible to ligation until they are partially filled with a Hal-NB or original nucleobase, resulting in complementary cohesive end formation. Ligation of these fragments ultimately leads to the required Hal-NB-labeled DNA duplex. With this approach, a synthetic, extremely long DNA fragment can be obtained by means of a multiple assembly reaction (n × maximum PCR product length: n × app. 50 kb).

Conclusions

The long, precisely labeled DNA duplexes obtained behave in very much the same manner as natural DNA and are beyond the range of chemical synthesis. Moreover, the conditions of synthesis closely resemble the natural ones, and all the artifacts accompanying the chemical synthesis of DNA are thus eliminated. The approach proposed seems to be completely general and could be used to label DNA at multiple pre-determined sites and with halogen derivatives of any nucleobase. Access to DNAs labeled with Hal-NBs at specific position is an indispensable condition for the understanding and optimization of DNA photo- and radio-degradation, which are prerequisites for clinical trials of Hal-NBs in anticancer therapy.

Benign decay vs. photolysis in the photophysics and photochemistry of 5-bromouracil. A computational study

The excited state potential energy surface of 5-bromouracil has been studied with ab initio CASPT2//CASSCF calculations to rationalize the competition between the benign decay and the photolysis found experimentally. The surface is characterized by an extended region of degeneracy between S(1) and S(0). The access to this region has been studied with minimum energy path calculations from the FC structure, the seam of intersection has been mapped in detail, and the decay paths from different regions of the seam have been characterized. There are two decay paths with low barriers that are limiting cases for the actual decay dynamics. The first path involves the bromine elimination and leads to a region of near degeneracy between the ground and excited states, and the second one leads back to the reactant through a conical intersection between the two states. The conical intersection for benign decay is part of a seam that lies along the C(5)-Br stretching coordinate, and decay at the region of the seam with a stretched C(5)-Br bond leads to photolysis. Thus, the reactivity depends on the point of the seam at which decay to the ground state takes place. The low experimental photolysis quantum yield suggests that the energetically favored decay is the one that regenerates the reactant, while the low barriers computed to access the region of decay are in agreement with the measured picosecond excited state lifetime.

Generation of Thiyl Radicals by the Photolysis of 5-Iodo-4-thiouridine

The photochemistry of 2',3',5'-tri-O-acetyl-5-iodo-4-thiouridine (3) in deoxygenated 1:1 CH(3)CN-H(2)O pH 5.8 (phosphate buffer) solution has been studied by means of steady-state and nanosecond laser flash photolysis methods. Under steady-state irradiation (lambda > or = 334 nm), the stable photoproducts were iodide ion, 2',3',5'-tri-O-acetyl-4-thiouridine (4), and two disulfides. The disulfides were the symmetrical bis-(2',3',5'-tri-O-acetyl-5-iodo-4-thiouridine) (5) and unsymmetrical 6, which contains both 4-thiouridine and 5-iodo-4-thiouridine residues. The formation of the dehalogenated photoproduct suggests that C(5)-I bond cleavage is a primary photochemical step. Attempts to scavenge the resulting C(5)-centered radical by suitable addends, bis-(N-alpha-acetyl)cystine-bis-N-ethylamide or benzene, were unsuccessful. Analysis of the photoproducts formed under these conditions showed that the S-atom is the reactive center. The photoproduct 4, obtained by irradiation of 3 in CD(3)CN-H(2)O, followed by reversed-phase HPLC isolation using nonlabeled eluents, did not contain deuterium. An analogous experiment performed in CH(3)CN-D(2)O gave deuterated product 4-d with 88% of the deuterium incorporated at C(5). Transient absorption observed upon laser excitation (lambda= 308 nm) of 3 was assigned to the 4-uridinylthiyl radical on the basis of the similarity of this spectrum with that obtained upon laser photolysis of the disulfide: bis-(2',3',5'-tri-O-acetyl-4-thiouridine) 14. On the basis of the results of steady-state and laser photolysis studies, a mechanism of the photochemical reaction of 3 is proposed. The key mechanistic step is a transformation of the C(5)-centered radical formed initially by C(5)-I bond cleavage into a long-lived S-centered radical via a 1,3-hydrogen shift. Theoretical calculations confirmed that the long-lived S-centered radical is the most stable radical derived from the 4-thiouracil residue.

Kinetic analysis of site-specific photoaptamer-protein cross-linking

ssDNA oligonucleotides containing bromodeoxyuridine, BrdU-photoaptamers, are rapidly emerging as specific protein capture reagents in protein microarray technologies. A mathematical model for the kinetic analysis of photoaptamer-protein photocross-linking reactions is presented. The model is based on specific aptamer/protein binding followed by laser excitation that can lead to either covalent cross-linking of the photoaptamer and protein in the complex or irreversible photodamage to the aptamer. Two distinct kinetic regimes, (1) frozen and (2) rapid equilibrium, are developed analytically to model binding kinetics between laser pulses. The models are used to characterize the photocross-linking between three photoaptamers and their cognate protein targets; photoaptamers 0650 and 0615 cross-link human basic fibroblast growth factor and 0518 cross-links HIV MN envelope glycoprotein. Data for cross-linking reaction yields as a function of both laser energy dose and target protein concentration are analyzed for affinity constants and cross-link reaction rates. The binding dissociation constants derived from the cross-linking data are in good accord with independent measurements; the rapid equilibrium model appears to produce results more consistent with the experimental observations, although there is significant overlap between the two models for most conditions explored here. The rate of photodamage for 0615 and 0518 is 3.5 and 2.5 times that of the specific cross-link, giving low maximum reaction yields of approximately 20% and approximately 30%, whereas 0650 cross-links with a rate over five times higher than its photodamage rate and has a maximum reaction yield exceeding 80%. Quantum yields for the three systems are estimated from the data; photoaptamer 0650 has a reasonably high quantum yield of approximately 0.2 for protein cross-linking, while 0518 and 0615 have quantum yields of 0.07 and 0.02. The work presented here provides a useful set of metrics that allow for refinement of photoaptamer properties.

Sensitivity and specificity of photoaptamer probes

The potential of photoaptamers as proteomic probes was investigated. Photoaptamers are defined as aptamers that bear photocross-linking functionality, in this report, 5-bromo-2'-deoxyuridine. A key question regarding the use of photoaptamer probes is the specificity of the cross-linking reaction. The specificity of three photoaptamers was explored by comparing their reactions with target proteins and non-target proteins. The range of target/non-target specificity varies from 100- to >10(6)-fold with most values >10(4)-fold. The contributions of the initial binding step and the photocross-linking step were evaluated for each reaction. Photocross-linking never degraded specificity and significantly increased aptamer specificity in some cases. The application of photoaptamer technology to proteomics was investigated in microarray format. Immobilized anti-human immunodeficiency virus-gp120 aptamer was able to detect subnanomolar concentrations of target protein in 5% human serum. The levels of sensitivity and specificity displayed by photoaptamers, combined with other advantageous properties of aptamers, should facilitate development of protein chip technology.

Diagnostic potential of PhotoSELEX-evolved ssDNA aptamers

High sensitivity and specificity of two modified ssDNA aptamers capable of photocross-linking recombinant human basic fibroblast growth factor (bFGF((155))) were demonstrated. The aptamers were identified through a novel, covalent, in vitro selection methodology called photochemical systematic evolution of ligands by exponential enrichment (PhotoSELEX). The aptamers exhibited high sensitivity for bFGF((155)) comparable with commercially available ELISA monoclonal antibodies with an absolute sensitivity of at least 0.058 ppt bFGF((155)) under prevailing test conditions. The aptamers exquisitely distinguished bFGF((155)) from consanguine proteins, vascular endothelial growth factor (VEGF) and platelet derived growth factor (PDGF). A commercially viable diagnostic system incorporating PhotoSELEX-evolved aptamers capable of simultaneous quantification of a large number of analyte molecules is also described. Such a system benefits from covalent bonding of aptamer to target protein allowing vigorous washing with denaturants to improve signal to noise.

Mass spectral characterization of a protein-nucleic acid photocrosslink

A photocrosslink between basic fibroblast growth factor (bFGF155) and a high affinity ssDNA oligonucleotide was characterized by positive ion electrospray ionization mass spectrometry (ESIMS). The DNA was a 61-mer oligonucleotide photoaptamer bearing seven bromodeoxyuridines, identified by in vitro selection. Specific photocrosslinking of the protein to the oligonucleotide was achieved by 308 nm XeCl excimer laser excitation. The cross-linked protein nucleic acid complex was proteolyzed with trypsin. The resulting peptide crosslink was purified by PAGE, eluted, and digested by snake venom phosphodiesterase/alkaline phosphatase. Comparison of the oligonucleotide vs. the degraded peptide crosslink by high performance liquid chromatography coupled to an electrospray ionization triple quadrupole mass spectrometer showed a single ion unique to the crosslinked material. Sequencing by collision induced dissociation (MS/MS) on a triple quadrupole mass spectrometer revealed that this ion was the nonapeptide TGQYKLGSK (residues 130-138) crosslinked to a dinucleotide at Tyr133. The MS/MS spectrum indicated sequential fragmentation of the oligonucleotide to uracil covalently attached to the nonapeptide followed by fragmentation of the peptide bonds. Tyr133 is located within the heparin binding pocket, suggesting that the in vitro selection targeted this negative ion binding region of bFGF155.

Specific regions of contact between human T-cell leukemia virus type I Tax protein and DNA identified by photocross-linking

The human T-cell leukemia virus type I Tax protein forms a ternary complex on DNA in association with a host factor, the cyclic AMP response element-binding protein (CREB). An understanding of the precise geometry of this complex has been elusive. We have used photocross-linking to investigate Tax-DNA contacts. Our data show that Tax contacts the DNA at two symmetric positions 14 nucleotides apart on either side of the Tax responsive element. The presence of symmetric, widely separated regions of contact suggests that at least two molecules of Tax are present in the complex. Mapping the contacts onto a three-dimensional model of the CREB-DNA binary complex shows that they lie on the same face of the DNA near the regions where the N termini of the CREB bZIP domains enter the major groove. This location correlates well with previous evidence that CREB amino acid residues immediately N-terminal to the bZIP domain are crucial for the formation of the ternary complex. The limited number of cross-links observed suggests that contacts are primarily with the phosphate backbone and does not support the idea that a major structural element of the Tax protein inserts into the major or minor grooves of the DNA.

Photocross-linking of nucleic acids to associated proteins

Photocross-linking is a useful technique for the partial definition of the nucleic acid-protein interface of nucleoprotein complexes. It can be accomplished by one or two photon excitations of wild-type nucleoprotein complexes or by one photon excitation of nucleoprotein complexes bearing one or more substitutions with photoreactive chromophores. Chromophores that have been incorporated into nucleic acids for this purpose include aryl azides, 5-azidouracil, 8-azidoadenine, 8-azidoguanine, 4-thiouracil, 5-bromouracil, 5-iodouracil, and 5-iodocytosine. The various techniques and chromophores are described and compared, with attention to the photochemical mechanism.

An RNA-protein contact determined by 5-bromouridine substitution, photocrosslinking and sequencing

An analogue of the replicase translational operator of bacteriophage R17, that contains a 5-bromouridine at position -5 (RNA 1), complexes with a dimer of the coat protein and photocrosslinks to the coat protein in high yield upon excitation at 308 nm with a xenon chloride excimer laser. Tryptic digestion of the crosslinked nucleoprotein complex followed by Edman degradation of the tryptic fragment bearing the RNA indicates crosslinking to tyrosine 85 of the coat protein. A control experiment with a Tyr 85 to Ser 85 variant coat protein showed binding but no photocrosslinking at saturating protein concentration. This is consistent with the observation from model compound studies of preferential photocrosslinking of BrU to the electron rich aromatic amino acids tryptophan, tyrosine, and histidine with 308 nm excitation.

Recombinant photoreactive tRNA molecules as probes for cross-linking studies

Photoreactive tRNA derivatives have been used extensively for investigating the interaction of tRNA molecules with their ligands and substrates. Recombinant RNA technology facilitates the construction of such tRNA probes through site-specific incorporation of photoreactive nucleosides. The general strategy involves preparation of suitable tRNA fragments and their ligation either to a photoreactive nucleotide or to each other. tRNA fragments can be prepared by site-specific cleavage of native tRNAs, or synthesized by enzymatic and chemical means. A number of photoreactive nucleosides suitable for incorporation into tRNA are presently available. Joining of tRNA fragments is accomplished either by RNA ligase or by DNA ligase in the presence of a DNA splint. The application of this methodology to the study of tRNA binding sites on the ribosome is discussed, and a model of the tRNA-ribosome complex is presented.

Determination of the DNA-interacting region of the archaebacterial chromosomal protein MC1. Photocrosslinks with 5-bromouracil-substituted DNA

Protein MC1 is the major chromosomal protein in methanosarcinaceae. Using photochemical crosslinking on 5-bromouracil-substituted DNA, we identified the region of the protein that interacts with it. This region is located in the C-terminal part of the polypeptide chain, and the crosslinked amino-acids are in the region 74-86. Tryptophan 74 is one of the amino-acids crosslinked to DNA.

A specific, UV-induced RNA-protein cross-link using 5-bromouridine-substituted RNA

The well-characterized RNA binding site of the bacteriophage R17 coat protein has been used to investigate the cross-linking of protein to 5-bromouridine (BrU)-substituted RNA using medium-wavelength UV light. We have demonstrated a specific RNA-protein cross-link and identified the site on the RNA of protein attachment. Formation of the covalent complex is dependent upon the presence of BrU at position -5 of the RNA and specific binding of the RNA by coat protein. The amount of cross-linking increases with time and depends on the light source and conditions used. Irradiations using a broad-spectrum UV transilluminator (peak at 312 nm) or monochromatic XeCl excimer laser (308 nm) gave levels of cross-linking exceeding 20 and 50%, respectively. The quantum yield of photo-cross-linking, determined with 308-nm excitation, was 0.003. While little strand breakage or debromination of the RNA occurred, significant protein photodamage was observed.