Combinatorial Chemistry & High Throughput Screening - Volume 13, Issue 9, 2010

The usage of principal component analysis (PCA) method in prediction of pharmacological classification of the drugs based on high-performance liquid chromatography (HPLC) retention data and on non-empirical structural parameters was studied. A group of 36 drugs of established pharmacological classification were chromatographed in ten carefully designed HPLC systems. Additionally, twelve structural descriptors were derived by molecular modeling studies based on the structural formula of considered drugs. A matrix of 36 x 22 HPLC data together with molecular properties parameters was subjected to chemometric analysis by PCA. Although that size of the training set could be sometimes disputable, the work remains as a demonstration of the basic methodology without the straight focus primarily intended as a report on a comprehensive predictive model. Nevertheless, the obtained clustering of drugs was in accordance with their pharmacological classification as well as chemical structures classification. The PCA method of the HPLC retention data and structural descriptors allowed to segregate drugs and drug candidates according to their pharmacological properties, and may be of potential help to limit the number of biological assays in the search for new drugs.

We describe a powerful peptide microarray for profiling protein kinase substrates that combines the merits of chemoselective immobilization of peptides to achieve high density spots with the advantages of fluorescence-based analysis of phosphorylation for nonhazardous detection. For detection of on-chip phosphorylation, we used a fluorescence-labeled antiphosphotyrosine antibody to detect phosphotyrosine and a biotinylated Phostag, which was subsequently bound with a fluorescence-labeled streptavidin for phosphoserine/threonine. More than 290 kinds of Tyr peptides and over 1,100 kinds of Ser/Thr peptides were chemoselectively immobilized onto a glass surface in a highdensity format to profile a panel of protein kinases, including c-Src, c-Abl, EGFR, JNK1, ERK2, p38α, and PKA. Many novel, highly reactive and specific peptides were identified as substrates for each protein kinase. Most substrates had the consensus motifs that have been reported previously but some new motifs were also found. The identification of two designed peptides that have higher reactivity than the famous PKA substrate (Kemptide) indicates that analysis of the amino acid biases of substrates is very helpful to the design of new substrates with high reactivity. Thus, the high-density peptide microarray is expected to be a powerful approach for high-throughput discovery of potential substrates for protein kinases.

Control of NF-κB release through the inhibition of I&κB kinase β (IKKβ) has been identified as a potential target for the treatment of inflammatory and autoimmune diseases. To screen the small molecule compound library against IKKβ, a high throughput screening (HTS) campaign was carried out using immobilized metal affinity for phosphochemicals (IMAP)-based time-resolved fluorescence resonance energy transfer (TR-FRET) assay. Through serial optimization of assay conditions, the Z' value was achieved at 0.88 from the pilot library screening of the most diverse 7,243 compounds with reconfirmation rate of 63%. The results from this HTS campaign identified three novel scaffolds for the prospective IKKβ inhibitor, such as 7-benzoyl-4-phenylcyclopenta [1,2] oxazine, 1-(thiophen or furan)-2,3- dihydroimidazo[1,5] pyridine and 2-phenyloxazolo [5,4] pyridine. Particularly, 7-benzoyl-4-phenylcyclopenta [1,2] oxazine derivatives presented potent inhibitory activity and selectivity for IKKβ. These findings suggest that the current TR-FRET assay system for IKKβ was successful to identify hits for novel IKKβ inhibitors as a robust, reproducible and sensitive HTS system.

Bone remodeling is tightly controlled by the actions of osteoblast and osteoclast. Impaired osteoblast proliferation and differentiation may disrupt the balance and lead to pathological symptom such as osteoporosis. To help understand the molecular mechanism of osteoblast proliferation, we performed a phenotype-driven high throughput screening with a random siRNA library, in search of novel genes that can accelerate murine preosteoblast MC3T3-E1 cell proliferation. Three siRNAs screened from the library were able to enhance MC3T3-E1 cell proliferation significantly. One of the proliferation-enhancing siRNAs (B7) was further subjected to expression profiling to pinpoint genes that putatively act down stream of it. A number of genes were regulated in response to proliferation-enhancing siRNA B7. Among these genes, Tmed2, which has never been reported yet in cell proliferation, was verified to be able to enhance MC3T3-E1 cell proliferation when over-expressed. Our screening process with random siRNA library provided an alternative strategy in addition to gene-specific siRNA library, in search of novel functional genes in genome scale.

Clinical application of viral vectors is often hampered by the lack of selectivity of viral particles for the targeted tissue. This drawback decreases the efficiency of gene delivery and raises safety concerns. We successfully established a novel in vitro evolution protocol to engineer adeno-associated virus vectors with increased selectivity for designated target cells. Subjecting a peptide-display library of AAV capsids to negative selection cycles on human primary fibroblasts and to positive selection cycles on a human melanoma cell line, we isolated several variants with up to 3.7-fold increased specificity for malignant cells in comparison to fibroblasts and other cell types. These mutants can be used to achieve high levels of gene transfer to target cells reducing undesired transduction of neighbouring tissues.

An efficient synthesis of 4-alkyl-4-methyl-3,4-dihydro-2H,5H-pyrano[3,2-c]chromene-2,5-dione or 4-aryl-3,4- dihydro-2H,5H-pyrano[3,2-c]chromene-2,5-diones via reaction 4-hydroxycoumarin with Meldrum's acid and ketones or aldehydes is described.

We generated monoclonal scFv (single chain variable fragment) antibodies from an antibody phage display library towards three small synthetic peptides derived from the sequence of αs1-casein. Key difficulties for selection of scFv-phages against small peptides were addressed. Small peptides do not always bind efficiently to passive adsorption surfaces, and we developed a simple method to quantify the binding capacity of surfaces with the peptides. Background binding (the binding of scFvs to the background matrix) is an obstacle for successful selection, and we evaluated two methods that drastically reduced the background binding. An optimized method therefore enabled a panning procedure where the specific (peptide binding) scFv-phages were always dominant. Using 15-mer peptides immobilized on Nunc Immobilizer Streptavidin plates, we successfully generated scFvs specifically against them. The scFvs were sequenced and characterized, and specificity was characterized by ELISA. The methods developed in this study are universally applicable for antibody phage display to efficiently produce antibody fragments against small peptides.

A recombinant Haematobia irritans irritans trypsin inhibitor (HiTI - Mw 7030 kDa) phagemid library was constructed and displayed functionally on the tip of the filamentous M13 phage. A combinatorial library of 7.2 x 106 mutants was created with HiTI mutations restricted to the P1' - P3' and P5' positions of the reactive site. This combinatorial library was selected for trypsin-like Pr2 proteases of Metarhizium anisopliae fungus, and 11 HiTI mutants containing the following substitutions: K17G, S18R, D19G, S21A, among 60 sequenced clones, were obtained. In order to confirm the inhibitory activity of the selected sequences, we transferred the selected sequence to the shortest protease inhibitor, the sunflower trypsin inhibitor (SFTI - Mw 1533 Da), for inhibitory activity analysis. The hybrid peptide containing the mutated sequence (SFTI-Mut, GRCTRGRGLACFPD-NH2; Ki = 14 μM) presented an apparent inhibition constant (Kiapp) for Pr2 proteases ~20-fold lower than the control peptide containing the original HiTI sequence (SFTIHiTI, GRCTRKSDLSCFPD-NH2; Ki = 259 μM). In conclusion, the present work enabled the selection of a specific HiTI mutant for Pr2 proteases of M. anisopliae fungus using a HiTI combinatorial library on M13 phage surface. Selection of strong binders by phage display and their validation as inhibitors using synthetic hybrid peptides proved to be a powerful technique to generate specific serine protease inhibitors suitable for studies of drug design and enzyme-inhibitor interaction.