In vitro and in vivo evaluation of cysteine and site specific conjugated herceptin antibody-drug conjugates.

Antibody drug conjugates (ADCs) are monoclonal antibodies designed to deliver a cytotoxic drug selectively to antigen expressing cells. Several components of an ADC including the selection of the antibody, the linker, the cytotoxic drug payload and the site of attachment used to attach the drug to the antibody are critical to the activity and development of the ADC. The cytotoxic drugs or payloads used to make ADCs are typically conjugated to the antibody through cysteine or lysine residues. This results in ADCs that have a heterogeneous number of drugs per antibody. The number of drugs per antibody commonly referred to as the drug to antibody ratio (DAR), can vary between 0 and 8 drugs for a IgG1 antibody. Antibodies with 0 drugs are ineffective and compete with the ADC for binding to the antigen expressing cells. Antibodies with 8 drugs per antibody have reduced in vivo stability, which may contribute to non target related toxicities. In these studies we incorporated a non-natural amino acid, para acetyl phenylalanine, at two unique sites within an antibody against Her2/neu. We covalently attached a cytotoxic drug to these sites to form an ADC which contains two drugs per antibody. We report the results from the first direct preclinical comparison of a site specific non-natural amino acid anti-Her2 ADC and a cysteine conjugated anti-Her2 ADC. We report that the site specific non-natural amino acid anti-Her2 ADCs have superior in vitro serum stability and preclinical toxicology profile in rats as compared to the cysteine conjugated anti-Her2 ADCs. We also demonstrate that the site specific non-natural amino acid anti-Her2 ADCs maintain their in vitro potency and in vivo efficacy against Her2 expressing human tumor cell lines. Our data suggests that site specific non-natural amino acid ADCs may have a superior therapeutic window than cysteine conjugated ADCs.

Automated 96-well purification of hexahistidine-tagged recombinant proteins on MagneHis Ni(2)+-particles.

Functional genomics and the application of high-throughput (HT) approaches to solve biological and medical questions are the main drivers behind the increasing need for HT parallel expression and purification of recombinant proteins. Automation is necessary to facilitate this complex multistep process. We describe, in detail, an HT-automated purification of hexahistidine-tagged recombinant proteins using MagneHis Ni-Particles and the Biomek FX robot. This procedure is universally applicable to hexa-histidine-tagged recombinant proteins with the tag positioned at either the N- or C-terminus. With minor modifications, the automated protein purification protocol presented in this chapter could be adapted to purify recombinant proteins bearing other tags than hexahistidine and/or other expression systems than E. coli.

High-throughput protein concentration and buffer exchange: comparison of ultrafiltration and ammonium sulfate precipitation.

High-throughput protein purification is a complex, multi-step process. There are several technical challenges in the course of this process that are not experienced when purifying a single protein. Among the most challenging are the high-throughput protein concentration and buffer exchange, which are not only labor-intensive but can also result in significant losses of purified proteins. We describe two methods of high-throughput protein concentration and buffer exchange: one using ammonium sulfate precipitation and one using micro-concentrating devices based on membrane ultrafiltration. We evaluated the efficiency of both methods on a set of 18 randomly selected purified proteins from Shewanella oneidensis. While both methods provide similar yield and efficiency, the ammonium sulfate precipitation is much less labor intensive and time consuming than the ultrafiltration.

Discovery of a human peptide sequence signaling islet neogenesis.

OBJECTIVE: To identify triggers for islet neogenesis in humans that may lead to new treatments that address the underlying mechanism of disease for patients with type 1 or type 2 diabetes. METHODS: In an effort to identify bioactive human peptide sequences that might trigger islet neogenesis, we evaluated amino acid sequences within a variety of mammalian pancreas-specific REG genes. We evaluated GenBank, the Basic Local Alignment Search Tool algorithm, and all available proteomic databases and developed large-scale protein-to-protein interaction maps. Studies of peptides of interest were conducted in human pancreatic ductal tissue, followed by investigations in mice with streptozocin-induced diabetes. RESULTS: Our team has defined a 14-amino acid bioactive peptide encoded by a portion of the human REG3a gene we termed Human proIslet Peptide (HIP), which is well conserved among many mammals. Treatment of human pancreatic ductal tissue with HIP stimulated the production of insulin. In diabetic mice, administration of HIP improved glycemic control and significantly increased islet number. Bioinformatics analysis, coupled with biochemical interaction studies in a human pancreatic cell line, identified the human exostoses-like protein 3 (EXTL3) as a HIP-binding protein. HIP enhanced EXTL3 translocation from the membrane to the nucleus, in support of a model whereby EXTL3 mediates HIP signaling for islet neogenesis. CONCLUSION: Our data suggest that HIP may be a potential stimulus for islet neogenesis and that the differentiation of new islets is a process distinct from beta cell proliferation within existing islets. Human clinical trials are soon to commence to determine the effect of HIP on generating new islets from one's own pancreatic progenitor cells.

Statistically inferring protein-protein associations with affinity isolation LC-MS/MS assays.

Affinity isolation of protein complexes followed by protein identification by LC-MS/MS is an increasingly popular approach for mapping protein interactions. However, systematic and random assay errors from multiple sources must be considered to confidently infer authentic protein-protein interactions. To address this issue, we developed a general, robust statistical method for inferring authentic interactions from protein prey-by-bait frequency tables using a binomial-based likelihood ratio test (LRT) coupled with Bayes' Odds estimation. We then applied our LRT-Bayes' algorithm experimentally using data from protein complexes isolated from Rhodopseudomonas palustris. Our algorithm, in conjunction with the experimental protocol, inferred with high confidence authentic interacting proteins from abundant, stable complexes, but few or no authentic interactions for lower-abundance complexes. The algorithm can discriminate against a background of prey proteins that are detected in association with a large number of baits as an artifact of the measurement. We conclude that the experimental protocol including the LRT-Bayes' algorithm produces results with high confidence but moderate sensitivity. We also found that Monte Carlo simulation is a feasible tool for checking modeling assumptions, estimating parameters, and evaluating the significance of results in protein association studies.

Solvent-accessible cysteines in human cystathionine beta-synthase: crucial role of cysteine 431 in S-adenosyl-L-methionine binding.

Cystathionine beta-synthase (CBS) is a tetrameric heme protein that catalyzes the PLP-dependent condensation of serine and homocysteine to cystathionine. CBS occupies a crucial regulatory position between the methionine cycle and transsulfuration. Human CBS contains 11 cysteine residues that are highly conserved in mammals but completely absent in the yeast enzyme, which catalyzes an identical reaction, suggesting a possible regulatory role for some of these residues. In this report, we demonstrate that in both the presence and absence of the CBS allosteric regulator S-adenosyl-l-methionine (AdoMet), only C15 and C431 of human CBS are solvent accessible. Mutagenesis of C15 to serine did not affect catalysis or AdoMet activation but significantly reduced aggregation of the purified enzyme in vitro. Mutagenesis of C431 resulted in a constitutively activated form of CBS that could not be further activated by either AdoMet or thermal activation. We and others have previously reported a number of C-terminal CBS point mutations that result in a decreased or abolished response to AdoMet. In contrast to all of these previously investigated CBS mutants, the C431 mutant form of CBS was unable to bind AdoMet, indicating that either this residue is directly involved in AdoMet binding or its absence induces a conformational change that destroys the integrity of the binding site for this regulatory ligand.

"Zero-length" cross-linking in solid state as an approach for analysis of protein-protein interactions.

We have developed a new approach for the analysis of interacting interfaces in protein complexes and protein quaternary structure based on cross-linking in the solid state. Protein complexes are freeze-dried under vacuum, and cross-links are introduced in the solid phase by dehydrating the protein in a nonaqueous solvent creating peptide bonds between amino and carboxyl groups of the interacting peptides. Cross-linked proteins are digested into peptides with trypsin in both H2(16)O and H(2)18O and then readily distinguished in mass spectra by characteristic 8 atomic mass unit (amu) shifts reflecting incorporation of two 18O atoms into each C terminus of proteolytic peptides. Computer analysis of mass spectrometry (MS) and MS/MS data is used to identify the cross-linked peptides. We demonstrated specificity and reproducibility of our method by cross-linking homo-oligomeric protein complexes of glutathione-S-transferase (GST) from Schistosoma japonicum alone or in a mixture of many other proteins. Identified cross-links were predominantly of amide origin, but six esters and thioesters were also found. The cross-linked peptides were validated against the GST monomer and dimer X-ray structures and by experimental (MS/MS) analyses. Some of the identified cross-links matched interacting peptides in the native 3D structure of GST, indicating that the structure of GST and its oligomeric complex remained primarily intact after freeze-drying. The pattern of oligomeric GST obtained in solid state was the same as that obtained in solution by Ru (II) Bpy(3)2+ catalyzed, oxidative "zero-length" cross-linking, confirming that it is feasible to use our strategy for analyzing the molecular interfaces of interacting proteins or peptides.

Automated purification of recombinant proteins: combining high-throughput with high yield.

Protein crystallography, mapping protein interactions, and other functional genomic approaches require purifying many different proteins, each of sufficient yield and homogeneity, for subsequent high-throughput applications. To fill this requirement efficiently, there is a need to develop robust, automated, high-throughput protein expression, and purification processes. We developed and compared two alternative workflows for automated purification of recombinant proteins based on expression of bacterial genes in Escherichia coli (E. coli). The first is a filtration separation protocol in which proteins of interest are expressed in a large volume, 800 ml of E. coli cultures, then isolated by filtration purification using Ni-NTA-Agarose (Qiagen). The second is a smaller scale magnetic separation method in which proteins of interest are expressed in a small volume, 25 ml, of E. coli cultures then isolated using a 96-well purification system with MagneHis Ni2+ Agarose (Promega). Both workflows provided comparable average yields of proteins, about 8 microg of purified protein per optical density unit of bacterial culture measured at 600 nm. We discuss advantages and limitations of these automated workflows, which can provide proteins with more than 90% purity and yields in the range of 100 microg to 45 mg per purification run, as well as strategies for optimizing these protocols.

Simple protein complex purification and identification method for high-throughput mapping of protein interaction networks.

Most current methods for purification and identification of protein complexes use endogenous expression of affinity-tagged bait, tandem affinity tag purification of protein complexes followed by specific elution of complexes from beads, and gel separation and in-gel digestion prior to mass spectrometric analysis of protein interactors. We propose a single affinity tag in vitro pull-down assay with denaturing elution, trypsin digestion in organic solvent, and LC-ESI MS/MS protein identification using SEQUEST analysis. Our method is simple and easy to scale-up and automate, making it suitable for high-throughput mapping of protein interaction networks and functional proteomics.

Ampicillin/penicillin-binding protein interactions as a model drug-target system to optimize affinity pull-down and mass spectrometric strategies for target and pathway identification.

The identification and validation of the targets of active compounds identified in cell-based assays is an important step in preclinical drug development. New analytical approaches that combine drug affinity pull-down assays with mass spectrometry (MS) could lead to the identification of new targets and druggable pathways. In this work, we investigate a drug-target system consisting of ampicillin- and penicillin-binding proteins (PBPs) to evaluate and compare different amino-reactive resins for the immobilization of the affinity compound and mass spectrometric methods to identify proteins from drug affinity pull-down assays. First, ampicillin was immobilized onto various amino-reactive resins, which were compared in the ampicillin-PBP model with respect to their nonspecific binding of proteins from an Escherichia coli membrane extract. Dynal M-270 magnetic beads were chosen to further study the system as a model for capturing and identifying the targets of ampicillin, PBPs that were specifically and covalently bound to the immobilized ampicillin. The PBPs were identified, after in situ digestion of proteins bound to ampicillin directly on the beads, by using either one-dimensional (1-D) or two-dimensional (2-D) liquid chromatography (LC) separation techniques followed by tandem mass spectrometry (MS/MS) analysis. Alternatively, an elution with N-lauroylsarcosine (sarcosyl) from the ampicillin beads followed by in situ digestion and 2-D LC-MS/MS analysis identified proteins potentially interacting noncovalently with the PBPs or the ampicillin. The in situ approach required only little time, resources, and sample for the analysis. The combination of drug affinity pull-down assays with in situ digestion and 2-D LC-MS/MS analysis is a useful tool in obtaining complex information about a primary drug target as well as its protein interactors.

Expression screen by enzyme-linked immunofiltration assay designed for high-throughput purification of affinity-tagged proteins.

High-throughput purification of affinity-tagged fusion proteins is currently one of the fastest developing areas of molecular proteomics. A prerequisite for success in protein purification is sufficient soluble protein expression of the target protein in a heterologous host. Hence, a fast and quantitative evaluation of the soluble-protein levels in an expression system is one of the key steps in the entire process. Here we describe a high-throughput expression screen for affinity-tagged fusion proteins based on an enzyme linked immunofiltration assay (ELIFA). An aliquot of a crude Escherichia coli extract containing the analyte, an affinity-tagged protein, is adsorbed onto the membrane. Subsequent binding of specific antibodies followed by binding of a secondary antibody horseradish peroxidase (HRP) complex then allows quantitative evaluation of the analyte using tetramethylbenzidine as the substrate for HRP. The method is accurate and quantitative, as shown by comparison with results from western blotting and an enzymatic glutathione S-transferase (GST) assay. Furthermore, it is a far more rapid assay and less cumbersome than western blotting, lending itself more readily to high-throughput analysis. It can be used at the expression level (cell lysates) or during the subsequent purification steps to monitor yield of specific protein.

Deletion mutagenesis of human cystathionine beta-synthase. Impact on activity, oligomeric status, and S-adenosylmethionine regulation.

Cystathionine beta-synthase is a tetrameric hemeprotein that catalyzes the pyridoxal 5'-phosphate-dependent condensation of serine and homocysteine to cystathionine. We have used deletion mutagenesis of both the N and C termini to investigate the functional organization of the catalytic and regulatory regions of this enzyme. Western blot analysis of these mutants expressed in Escherichia coli indicated that residues 497-543 are involved in tetramer formation. Deletion of the 70 N-terminal residues resulted in a heme-free protein retaining 20% of wild type activity. Additional deletion of 151 C-terminal residues from this mutant resulted in an inactive enzyme. Expression of this double-deletion mutant as a glutathione S-transferase fusion protein generated catalytically active protein (15% of wild type activity) that was unaffected by subsequent removal of the fusion partner. The function of the N-terminal region appears to be primarily steric in nature and involved in the correct folding of the enzyme. The C-terminal region of human cystathionine beta-synthase contains two hydrophobic motifs designated "CBS domains." Partial deletion of the most C-terminal of these domains decreased activity and caused enzyme aggregation and instability. Removal of both of these domains resulted in stable constitutively activated enzyme. Deletion of as few as 8 C-terminal residues increased enzyme activity and abolished any further activation by S-adenosylmethionine indicating that the autoinhibitory role of the C-terminal region is not exclusively a function of the CBS domains.

High homocysteine and thrombosis without connective tissue disorders are associated with a novel class of cystathionine beta-synthase (CBS) mutations.

Cystathionine beta-synthase (CBS) is a crucial regulator of plasma levels of the thrombogenic amino acid homocysteine (Hcy). Homocystinuria due to CBS deficiency confers a dramatically increased risk of thrombosis. Early diagnosis usually occurs after the observation of ectopia lentis, mental retardation, or characteristic skeletal abnormalities. Homocystinurics with this phenotype typically carry mutations in the catalytic region of the protein that abolish CBS activity. We describe a novel class of missense mutations consisting of I435T, P422L, and S466L that are located in the non-catalytic C-terminal region of CBS that yield enzymes that are catalytically active but deficient in their response to S-adenosylmethionine (AdoMet). The P422L and S466L mutations were found in patients suffering premature thrombosis and homocystinuric levels of Hcy but lacking any of the connective tissue disorders typical of homocystinuria due to CBS deficiency. The P422L and S466L mutants demonstrated a level of CBS activity comparable to that of the AdoMet stimulated wild-type CBS but could not be further induced by the addition of AdoMet. In terms of temperature stability, oligomeric organization, and heme saturation the I435T, P422L, and S466L mutants are indistinguishable from wild-type CBS. Our findings illustrate the importance of AdoMet for the regulation of Hcy metabolism and are consistent with the possibility that the characteristic connective tissue disturbances observed in homocystinuria due to CBS deficiency may not be due to elevated Hcy.