Affinity chromatography matures as bioinformatic and combinatorial tools develop

Affinity chromatography has the reputation of a more expensive and less robust than other types of liquid chromatography. Furthermore, the technique is considered to stand a modest chance of large-scale purification of proteinaceous pharmaceuticals. This perception is changing because of the pressure for quality protein therapeutics, and the realization that higher returns can be expected when ensuring fewer purification steps and increased product recovery. These developments necessitated a rethinking of the protein purification processes and restored the interest for affinity chromatography. This liquid chromatography technique is designed to offer high specificity, being able to safely guide protein manufactures to successfully cope with the aforementioned challenges. Affinity ligands are distinguished into synthetic and biological. These can be generated by rational design or selected from ligand libraries. Synthetic ligands are generated by three methods. The rational method features the functional approach and the structural template approach. The combinatorial method relies on the selection of ligands from a library of synthetic ligands synthesized randomly. The combined method employs both methods, that is, the ligand is selected from an intentionally biased library based on a rationally designed ligand. Biological ligands are selected by employing high-throughput biological techniques, e.g. phage- and ribosome-display for peptide and microprotein ligands, in addition to SELEX for oligonucleotide ligands. Synthetic mimodyes and chimaeric dye-ligands are usually designed by rational approaches and comprise a chloro-triazinlyl scaffold. The latter substituted with various amino acids, carbocyclic, and heterocyclic groups, generates libraries from which synthetic ligands can be selected. A 'lead' compound may help to generating a 'focused' or 'biased' library. This can be designed by various approaches, e.g.: (i) using a natural ligand-protein complex as a template; (ii) applying the principle of complementarity to exposed residues of the protein structure; and (iii) mimicking directly a natural biological recognition interaction. Affinity ligands, based on the peptide structure, can be peptides, peptide-mimetic derivatives (<30 monomers) and microproteins (e.g. 25-200 monomers). Microprotein ligands are selected from biological libraries constructed of variegated protein domains, e.g. minibody, Kunitz, tendamist, cellulose-binding domain, scFv, Cytb562, zinc-finger, SpA-analogue (Z-domain).     

Lock-and-key motif as a concept for designing affinity adsorbents for protein purification

The lock-and-key (LAK) motif, a common structural moiety found in subunit interfaces of glutathione S-transferases (GSTs), plays an important role in biomolecular recognition and quaternary structure integrity. Inspection of the key structural features of the LAK motif prompted the de novo design and combinatorial synthesis of a 13-membered solid-phase ligand library, employing as a lead ligand the Phe-Trz-X structure, mimicking the LAK motif. 1,3,5-Triazine (Trz) was used as the scaffold for assembly, substituted with different LAK-mimetic amino acids. De novo ligand design was effected using bioinformatics and molecular modeling and based on mimicking the interactions of the LAK motif. The library of affinity adsorbents was assessed for binding corn and human serum proteomes and purified proteins of different structure and ligand binding specificity. The results showed remarkable differences in the binding specificity of LAK-mimetic adsorbents for a wide range of proteins, as a consequence of minor changes in ligand structure. One LAK-mimetic adsorbent was integrated in a single-step purification protocol for human monoclonal anti-human immunodeficiency virus 2F5 antibody (mAb 2F5) from spiked corn extract, affording high recovery and purity. The results demonstrate that the principle of natural recognition found in the lock-and-key motif, in combination with de novo combinatorial design, may lead to synthetic affinity ligands, useful in downstream processing and proteomic research.     

Advantages of Thermococcus kodakaraenis (KOD) DNA Polymerase for PCR-mass spectrometry based analyses

The advantages of the thermostable DNA polymerase from Thermococcus kodakaraensis (KOD) are demonstrated for PCR amplification with subsequent detection by mass spectrometry. Commonly used DNA polymerases for PCR amplification include those from Thermus aquaticus (Taq) and Pyrococcus furiosus (Pfu). A 116 base-pair PCR product derived from a vWA locus was amplified by Taq, Pfu, or KOD DNA polymerase and compared by agarose gel electrophoresis and electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI-FTICR-MS). KOD DNA polymerase demonstrated a 2- to 3-fold increase in PCR product formation compared to Pfu or Taq, respectively, and generated blunt-ended PCR product which allows facile interpretation of the mass spectrum. Additionally, we demonstrate the advantage of using high magnetic fields to obtain unit resolution of the same 116 base pair (approximately 72 kDa) PCR product at high m/z.     

Development of recombinant protein-based influenza vaccine. Expression and affinity purification of H1N1 influenza virus neuraminidase

The influenza virus surface glycoprotein antigen neuraminidase (NA) is a crucial viral enzyme with many potential medical applications; therefore, the development of efficient upstream and downstream processing strategy for the expression and purification of NA is of high importance. In the present work the NA gene from the H1N1 influenza virus strain A/Beijing/262/95 was cloned from viral RNA and expressed in expresSF+ insect cells using the baculovirus expression vector system (BVES). A limited affinity-ligand library was synthesized and evaluated for its ability to bind and purify the recombinant H1N1 neuraminidase. Affinity-ligand design was based on mimicking the interactions of the lock-and-key (LAK) motif (Phe-Gly-Gln), a common structural moiety found in the subunit interface of glutathione S-transferase I (GST I), and plays an important structural role in subunit-subunit recognition. Solid-phase combinatorial chemistry was used to synthesize 13 variants of the lock-and-key lead ligand (Phe-Trz-X, where X was selected alpha-amino acid) using the 1,3,5-triazine moiety (Trz) as the scaffold for assembly. One immobilized ligand, bearing phenylalanine and isoleucine linked on the chlorotriazine ring (Phe-Trz-Ile), displayed high affinity for NA. Absorption equilibrium and molecular modeling studies were carried out to provide a detailed picture of Phe-Trz-Ile interaction with NA. This LAK-mimetic affinity adsorbent was exploited in the development of a facile purification protocol for NA, which led to 335-fold purification in a single-step. The present purification procedure is the most efficient reported so far for recombinant NA.     

Combining combinatorial chemistry and affinity chromatography: highly selective inhibitors of human betaine: homocysteine S-methyltransferase

A new method to find novel protein targets for ligands of interest is proposed. The principle of this approach is based on affinity chromatography and combinatorial chemistry. The proteins within a crude rat liver homogenate were allowed to interact with a combinatorial library of phosphinic pseudopeptides immobilized on affinity columns. Betaine: homocysteine S-methyltransferase (BHMT) was one of the proteins that was retained and subsequently eluted from these supports. The phosphinic pseudopeptides, which served as immobilized ligands for the isolation of rat BHMT, were then tested for their ability to inhibit human recombinant BHMT in solution. The most potent inhibitor also behaved as a selective ligand for the affinity purification of BHMT from a complex media. Further optimization uncovered Val-Phe-psi[PO(2-)-CH(2)]-Leu-His-NH(2) as a potent BHMT inhibitor that has an IC(50) of about 1 microM.     

Guanine-thymine intrastrand cross-linked lesion containing oligonucleotides: from chemical synthesis to in vitro enzymatic replication

An intrastrand cross-link lesion, in which two neighboring nucleobases are covalently tethered, has been site-specifically synthesized into defined sequence oligonucleotides in order to perform in vitro replication studies using either bacterial replicative or translesional synthesis polymerases. The investigated tandem base lesion that involves a cross-link between the methylene group of thymine and the C8 of an adjacent guanine residue has been prepared by UV-photolysis under anaerobic condition of the photolabile precursor 5-(phenylthiomethyl)-2'-deoxyuridine that has been site-specifically incorporated into a 9-mer oligonucleotide. After ligation, the lesion-containing modified oligonucleotide was used as a DNA template in primer extension reactions catalyzed by several DNA polymerases including the fragment Klenow exo-(Kf-) of E. coli polymerase I, the Thermus aquaticus polymerase (Taq pol) and the E. coli translesional DNA polymerase Pol IV (dinB). It was found that the primer extension reaction was stopped after the incorporation of the correct nucleotide dAMP opposite the 3'-thymine residue of guanine(C8-CH2) thymine lesion by Kf- and Pol IV; however it was noted that the efficiency of the nucleotide incorporation was reduced. In contrast, the Taq polymerase was totally blocked at the nucleotide preceding the tandem lesion. These results are strongly suggestive that the present intrastrand cross-link lesion, if not repaired, would constitute a blocking lesion for prokaryotic DNA polymerases, being likely lethal for the cell.     

Opposed steric constraints in human DNA polymerase beta and E. coli DNA polymerase I

DNA polymerase selectivity is crucial for the survival of any living species, yet varies significantly among different DNA polymerases. Errors within DNA polymerase-catalyzed DNA synthesis result from the insertion of noncanonical nucleotides and extension of misaligned DNA substrates. The substrate binding characteristics among DNA polymerases are believed to vary in properties such as shape and tightness of the binding pocket, which might account for the observed differences in fidelity. Here, we employed 4'-alkylated nucleotides and primer strands bearing 4'-alkylated nucleotides at the 3'-terminal position as steric probes to investigate differential active site properties of human DNA polymerase beta (Pol beta) and the 3'-->5'-exonuclease-deficient Klenow fragment of E. coli DNA polymerase I (KF(exo-)). Transient kinetic measurements indicate that both enzymes vary significantly in active site tightness at both positions. While small 4'-methyl and -ethyl modifications of the nucleoside triphosphate perturb Pol beta catalysis, extension of modified primer strands is only marginally affected. Just the opposite was observed for KF(exo-). Here, incorporation of the modified nucleotides is only slightly reduced, whereas size augmentation of the 3'-terminal nucleotide in the primer reduces the catalytic efficiency by more than 7000- and 260,000-fold, respectively. NMR studies support the notion that the observed effects derive from enzyme substrate interactions rather than inherent properties of the modified substrates. These findings are consistent with the observed differential capability of the investigated DNA polymerases in fidelity such as processing misaligned DNA substrates. The results presented provide direct evidence for the involvement of varied steric effects among different DNA polymerases on their fidelity.     

Purification on poly(U)-sepharose 4B of human breast cancer cell line T-47D DNA polymerases

The purification of DNA polymerases (RNA-directed DNA polymerases and DNA-directed DNA polymerases) on poly(U)-Sepharose 4B from a breast tumour cell line (T-47D) is reported. The elution of these enzymes was followed in each fraction by activity measurements with the four primer-templates poly(rA)-oligo(dT)12-18, poly(dA) oligo(dT)12-18, poly(rC)-oligo(dG)12-18 and poly(rCm)-oligo (dG)12-18. The control of the polymerase purification by chromatography was performed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis of the pooled active enzymatic fractions.     

Dissecting the differences between the alpha and beta anomers of the oxidative DNA lesion FaPydG

The oxidative DNA lesion, FaPydG rapidly anomerizes to form a mixture of the alpha and beta anomer. To investigate the mutagenic potential of both forms, we prepared stabilized bioisosteric analogues of both configurational isomers and incorporated them into oligonucleotides. These were subsequently used for thermodynamic melting-point studies and for primer-extension experiments. While the beta compound, in agreement with earlier data, prefers cytidine as the pairing partner, the alpha compound is not able form a stable base pair with any natural base. In primer-extension studies with the high-fidelity polymerase Bst Pol I, the polymerase was able to read through the lesion. The beta compound showed no strong mutagenic potential. The alpha compound, in contrast, strongly destabilized DNA duplexes and also blocked all of the tested DNA polymerases, including two low-fidelity polymerases of the Y-family.     

Incorporation of 4'-C-aminomethyl-2'-O-methylthymidine into DNA by thermophilic DNA polymerases

The dual modified nucleotide 4'-C-aminomethyl-2'-O-methylthymidine 5'-triphosphate was synthesized and enzymatically incorporated into DNA by the thermophilic DNA polymerases Pfu and Therminator III. The dual ribose modification imparted increased exonuclease resistance to DNA compared to the well-known 2'-O-methyl modification.     

Directed evolution of ATP binding proteins from a zinc finger domain by using mRNA display

Antibodies have traditionally been used for isolating affinity reagents to new molecular targets, but alternative protein scaffolds are increasingly being used for the directed evolution of proteins with novel molecular recognition properties. We have designed a combinatorial library based on the DNA binding domain of the human retinoid-X-receptor (hRXRalpha). We chose this domain because of its small size, stable fold, and two closely juxtaposed recognition loops. We replaced the two loops with segments of random amino acids, and used mRNA display to isolate variants that specifically recognize adenosine triphosphate (ATP), demonstrating a significant alteration of the function of this protein domain from DNA binding to ATP recognition. Many novel independent sequences were recovered with moderate affinity and high specificity for ATP, validating this scaffold for the generation of functional molecules.     

Synthesis and biological evaluation of a beauveriolide analogue library

Synthesis of beauveriolide III (1b), which is an inhibitor of lipid droplet accumulation in macrophages, was achieved by solid-phase assembly of linear depsipeptide using a 2-chlorotrityl linker followed by solution-phase cyclization. On the basis of this strategy, a combinatorial library of beauveriolide analogues was carried out by radio frequency-encoded combinatorial chemistry. After automated purification using preparative reversed-phase HPLC, the library was tested for inhibitory activity of CE synthesis in macrophages to determine structure-activity relationships of beauveriolides. Among them, we found that diphenyl derivative 7{9,1} is 10 times more potent than 1b.     

Amplification of bifunctional ligands for calmodulin from a dynamic combinatorial library

A well known strategy to prepare high affinity ligands for a biological receptor is to link together low affinity ligands. DCC (dynamic combinatorial chemistry) was used to select bifunctional protein ligands with high affinity relative to the corresponding monofunctional ligands. Thiol to disulfide linkage generated a small dynamic library of bifunctional ligands in the presence of calmodulin, a protein with two independently mobile domains. The binding constant of the bifunctional ligand (disulfide) most amplified by the presence of calmodulin is nearly two orders of magnitude higher than that of the corresponding monofunctional ligand (thiol).     

Thermodynamic and Mechanistic Insights into Translesion DNA Synthesis Catalyzed by Y‐Family DNA Polymerase Across a Bulky Double‐Base Lesion of an Antitumor Platinum Drug

To determine how the Y‐family translesion DNA polymerase η (Polη) processes lesions remains fundamental to understanding the molecular origins of the mutagenic translesion bypass. We utilized model systems employing a DNA double‐base lesion derived from 1,2‐GG intrastrand cross‐links of a new antitumor Pt^II complex containing a bulky carrier ligand, namely [PtCl₂(cis‐1,4‐dach)] (DACH=diaminocyclohexane). The catalytic efficiency of Polη for the insertion of correct dCTP, with respect to the other incorrect nucleotides, opposite the 1,2‐GG cross‐link was markedly reduced by the DACH carrier ligand. This reduced efficiency of Polη to incorporate the correct dCTP could be due to a more extensive DNA unstacking and deformation of the minor groove induced in the DNA by the cross‐link of bulky [PtCl₂(cis‐1,4‐dach)]. The major products of the bypass of this double‐base lesion produced by [PtCl₂(cis‐1,4‐dach)] by Polη resulted from misincorporation of dATP opposite the platinated G residues. The results of the present work support the thesis that this misincorporation could be due to sterical effects of the bulkier 1,4‐DACH ligand hindering the formation of the Polη–DNA–incoming nucleotide complex. Calorimetric analysis suggested that thermodynamic factors may contribute to the forces that governed enhanced incorporation of the incorrect dATP by Polη as well.     

An artificial protein L for the purification of immunoglobulins and fab fragments by affinity chromatography

The development and characterization of an artificial protein L (PpL) for the affinity purification of antibodies is described. Ligand 8/7, which emerged as the lead from a de novo designed combinatorial library of ligands, inhibits the interaction of PpL with IgG and Fab by competitive ELISA and shows negligible binding to Fc. The ligand 8/7 adsorbent (Ka approximately 10(4) M(-1)) compared well with PpL in binding to immunoglobulins from different classes and sources and, in addition, bound to IgG1 with K and lambda isotypes (92% and 100% of loaded protein) and polyclonal IgG from sheep, cow, goat and chicken. These properties were also reflected in the efficient isolation of immunoglobulins from crude samples.     

Highly efficient modification of DNA polymerase β under conditions of direct and sensitized activation of photoreactive DNAs. Modification of cell extract proteins

dUTP and dCTP derivatives containing a 4-azido-2,3,5,6-tetrafluorobenzylideneaminooxy group were incorporated into the 3′-end of the DNA primer within complexes with the DNA-matrix as analogs of natural dTTP by virtue of catalytic activity of DNA polymerase β or endogenous DNA polymerases of the cell extract. The photoreactive DNAs synthesized in situ were used for affinity modification of DNA polymerase β and DNA-binding proteins of the cell extract. For the photoreactive DNA based on these analogs, the efficiency of formation of covalent adducts with DNA polymerase β under the highest degree of DNA complexation with the enzyme was determined. The yield of covalent DNA adducts with the enzyme was 28–47%, depending on the type of the analog. The effect of the sequence of the DNA template near the localization of the photoreactive group on the redistribution of covalent cross-links between the possible targets was demonstrated. A possibility of increasing the efficiency of DNA polymerase β modification in the presence of a substantial excess of photoreactive DNA using a sensitizer, a dUTP derivative containing a pyrene residue, was studied. When photoreactive DNA containing a 2,3,5,6-tetrafluoro-4-azidobenzoyl (FAB) group was used, about 60% of DNA polymerase β was covalently attached to DNA. Photoreactive dNTP analogs ensuring a high level of protein modification in the cell extract were found. __________ Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 5, pp. 1273–1283, May, 2005.     

Biochemical Characterization of Translesion Synthesis by Sulfolobus acidocaldarius DNA Polymerases

To study the DNA synthesis mechanism of Sulfolobus acidocaldarius, a thermophilic species from Crenarchaeota, two DNA polymerases of B family(polB1 and polB3), and one DNA polymerase of Y family(polIV) were recombinantly expressed, purified and biochemically characterized. Both DNA polymerases polB1(Saci_1537) and polB3(Saci_0074) possessed DNA polymerase and 3' to 5' exonuclease activities; however, both the activities of B3 were very inefficient in vitro. The polIV(Saci_0554) was a polymerase, not an exonuclease. The activities of all the three DNA polymerases were dependent on divalent metal ions Mn²⁺ and Mg²⁺. They showed the highest activity at pH values ranging from 8.0 to 9.5. Their activities were inhibited by KCl with high concentration. The optimal reaction temperatures for the three DNA polymerases were between 60 and 70℃. Deaminated bases dU and dI on DNA template strongly hindered primer extension by the two DNA polymerases of B family, not by the DNA polymerase of Y family. DNA polymerase of Y Family bypassed the two AP site analogues dSpacer and propane on template more easily than DNA polymerases of B family. Our results suggest that the three DNA polymerases coordinate to fulfill various DNA synthesis in Sulfolobus acidocaldarius cell.     

Ligand-regulated peptides: a general approach for modulating protein-peptide interactions with small molecules

We engineered a novel ligand-regulated peptide (LiRP) system where the binding activity of intracellular peptides is controlled by a cell-permeable small molecule. In the absence of ligand, peptides expressed as fusions in an FKBP-peptide-FRB-GST LiRP scaffold protein are free to interact with target proteins. In the presence of the ligand rapamycin, or the nonimmunosuppressive rapamycin derivative AP23102, the scaffold protein undergoes a conformational change that prevents the interaction of the peptide with the target protein. The modular design of the scaffold enables the creation of LiRPs through rational design or selection from combinatorial peptide libraries. Using these methods, we identified LiRPs that interact with three independent targets: retinoblastoma protein, c-Src, and the AMP-activated protein kinase. The LiRP system should provide a general method to temporally and spatially regulate protein function in cells and organisms.     

G-quadruplex-specific peptide-hemicyanine ligands by partial combinatorial selection

G-quadruplex DNA-specific ligands were generated using a combinatorial selection of tetrapeptides which were then conjugated to a hemicyanine scaffold. The heterocycle enhanced peptide binding affinity by approximately 1000-fold to give ligands with near micromolar affinity and >40-fold discrimination for quadruplex DNA over duplex.