Characterization of benzodiazepine "combinatorial" chemical libraries by on-line immunoaffinity extraction, coupled column HPLC-ion spray mass spectrometry-tandem mass spectrometry.

To characterize combinatorial chemical libraries of small drug compounds, an automated column switching system incorporating an immunoaffinity extraction (IAE) column and two reversed-phase HPLC columns was coupled to a triple-quadrupole mass spectrometer. A Protein G column and antibodies to benzodiazepines were used to screen library components. A pH change in the mobile phase eluted the benzodiazepine-antibody complexes onto a C-18 restricted access media (RAM) column, thereby separating the selected benzodiazepines from the antibody. In a final step, backflushing the RAM column eluted the benzodiazepines onto a C-8 analytical reversed-phase column for separation before detection and preliminary structural characterization using ion spray mass spectrometry (MS) and tandem mass spectrometry (MS/ MS). A known 19-component library and an unknown 20-component library were analyzed. Full-scan IAE/LC/ LC/MS and IAE/LC/LC/MS/MS chromatograms suggested the feasibility of this combination of techniques, although the antibodies used were not highly specific. Inspection of MS/MS spectra of components in the unknown library compared to the MS/MS spectrum of a known standard (chlordiazepoxide) identified a subclass of benzodiazepines. Productions of the known standard and an unknown benzodiazepine were successively captured and fragmented (MSn experiments) using an iontrap mass spectrometer off-line, which confirmed that the unknown was an analogue of chlordiazepoxide.

Proline pipe helix: structure of the tus proline repeat determined by 1H NMR.

The structure of a 22 amino acid peptide, TPPI [Nedved, M. L., Gottlieb, P. A., & Moe, G. R. (1994) Nucleic Acids Res. 22, 5024-5030], that is similar to the proline repeat segment of the replication arrest protein, Tus, has been determined by 1H NMR in 50% trifluroethanol. The structure is a novel left-handed helix having 5.56 residues per turn and a regular hydrogen bonding network that is limited to one side of the helix and contains a channel that runs down the helix axis. The latter feature gives the structure an overall pipe-like appearance; hence, the structure has been designated a proline pipe helix. The Tus proline pipe is also amphiphilic with one side consisting of proline and other nonpolar residues while the other side contains mostly basic and other polar residues. Tus and several other proteins that contain a similar proline repeat sequence are DNA binding proteins. It is shown here that the proline pipe helix of TPPI can be accommodated within the major grove of B-form DNA in a manner that positions nearly all of the basic residues near phosphate groups in the DNA backbone. The proline pipe helical motif may be a structural element of many other proteins including integral membrane receptor proteins.

The use of affinity coelectrophoresis to characterize cooperative, nonspecific nucleic acid binding peptides.

Affinity coelectrophoresis (ACE) is a technique for characterizing ligand/nucleic acid binding interactions under equilibrium conditions. It is used here to characterize the protamine clupeine Z binding to several DNA fragments in order to define the use and limitations of ACE for ligands that bind cooperatively and nonspecifically to nucleic acids. The results demonstrate that the ACE data for cooperative, nonspecific ligands can be analyzed using the McGhee-von Hippel model and that binding-site sizes can be accurately determined using lattices containing as few as one site. However, binding constants can be greatly underestimated for some cooperative ligands with large-site sizes if small lattices are used. The salt dependence of the binding constant can also be determined but is limited to salt concentrations less than approximately 300 mM. Given the simplicity and reproducibility of the ACE assay, it should find many applications for studying binding interactions for a variety of cooperative and noncooperative nucleic acid binding peptides and proteins.

Cooperative, non-specific binding of a zinc finger peptide to DNA.

The DNA binding and structural properties of Xfin-31 (Lee, M.S., Gippert, G.P., Soman, K.V., Case, D.A. and Wright, P.E., 1989, Science 245, 635-637), a twenty five amino acid zinc finger peptide, in the reduced, oxidized and zinc complex forms, as well as the fourteen residue helical segment of the zinc finger (residues 12-25) have been compared using affinity coelectrophoresis (ACE) and circular dichroism (CD) spectroscopy. The zinc complex and oxidized peptides bind cooperatively to DNA although the cooperativity factor, omega, is more than 15-fold greater for the zinc complex. The reduced peptide in the absence of zinc and the helical segment do not bind cooperatively (omega = 1). Hence, the binding constant for singly contiguous sites (K omega) ranges over 100-fold for the various peptides even though the intrinsic binding constants (K) are similar. An increase in binding order and affinity for the other forms of Xfin-31 is correlated with an increasing similarity of the CD spectrum to that of the Xfin-31 zinc complex. The surprising DNA binding activity of the oxidized peptide may result from hydrophobic interactions between the amino-terminal loop formed by the Cys3-Cys6 disulfide bond and conserved hydrophobic residues in the carboxyl-terminal segment. Xfin-31 may be a particularly useful model for studying several poorly understood aspects of cooperative, non-specific DNA binding since it is small, has a stable, well-defined structure, and structures of zinc fingers bound to DNA have been determined.

CD and DNA binding studies of a proline repeat-containing segment of the replication arrest protein Tus.

Tus is a sequence-specific DNA binding protein that regulates its own transcription and can arrest Escherichia coli replication when bound to Ter sites on the chromosome. In order to identify segments of Tus that may be involved in DNA binding interactions we have analyzed the Tus amino acid sequence with respect to secondary structure motifs and similarity to other protein sequences. A twenty amino acid segment containing several basic residues and a proline repeat motif with a periodicity of five residues was identified. The motif was common to several other nucleic acid binding proteins, including histone H1-3, Xenopus laevis ribosomal protein L1, and the single-stranded DNA binding protein (DBP) from adenovirus. A 22 amino acid peptide, TPPI, having a sequence similar to the Tus segment binds non-specifically and non-cooperatively to double- and single-stranded DNA with a binding constant of 1.5 +/- 0.2 x 10(6) M-1. The estimated binding site size was 4.3 +/- 0.5 base pairs. Circular dichroism studies indicated that the peptide was a random coil in buffer but adopted a helical structure in 50% trifluroethanol and in sodium dodecyl sulfate at concentrations above the critical micellar concentration. Several helical models of the TPPI sequence were constructed graphically and minimized. One of them, an amphiphilic, left-handed, 5.1(19) helical model was best able to account for the observed structural properties of TPPI in the presence of structure-promoting additives.