Combinatorial Approach for Large-scale Identification of Linked Peptides from Tandem Mass Spectrometry Spectra

The combination of chemical cross-linking and mass spectrometry has recently been shown to constitute a powerful tool for studying protein–protein interactions and elucidating the structure of large protein complexes. However, computational methods for interpreting the complex MS/MS spectra from linked peptides are still in their infancy, making the high-throughput application of this approach largely impractical. Because of the lack of large annotated datasets, most current approaches do not capture the specific fragmentation patterns of linked peptides and therefore are not optimal for the identification of cross-linked peptides. Here we propose a generic approach to address this problem and demonstrate it using disulfide-bridged peptide libraries to (i) efficiently generate large mass spectral reference data for linked peptides at a low cost and (ii) automatically train an algorithm that can efficiently and accurately identify linked peptides from MS/MS spectra. We show that using this approach we were able to identify thousands of MS/MS spectra from disulfide-bridged peptides through comparison with proteome-scale sequence databases and significantly improve the sensitivity of cross-linked peptide identification. This allowed us to identify 60% more direct pairwise interactions between the protein subunits in the 20S proteasome complex than existing tools on cross-linking studies of the proteasome complexes. The basic framework of this approach and the MS/MS reference dataset generated should be valuable resources for the future development of new tools for the identification of linked peptides.The study of protein–protein interactions is crucial to understanding how cellular systems function because proteins act in concert through a highly organized set of interactions. Most cellular processes are carried out by large macromolecular assemblies and regulated through complex cascades of transient protein–protein interactions (1). In the past several years numerous high-throughput studies have pioneered the systematic characterization of protein–protein interactions in model organisms (2–4). Such studies mainly utilize two techniques: the yeast two-hybrid system, which aims at identifying binary interactions (5), and affinity purification combined with tandem mass spectrometry analysis for the identification of multi-protein assemblies (6–8). Together these led to a rapid expansion of known protein–protein interactions in human and other model organisms. Patche and Aloy recently estimated that there are more than one million interactions catalogued to date (9).But despite rapid progress, most current techniques allow one to determine only whether proteins interact, which is only the first step toward understanding how proteins interact. A more complete picture comes from characterizing the three-dimensional structures of protein complexes, which provide mechanistic insights that govern how interactions occur and the high specificity observed inside the cell. Traditionally the gold-standard methods used to solve protein structures are x-ray crystallography and NMR, and there have been several efforts similar to structural genomics (10) aiming to comprehensively solve the structures of protein complexes (11, 12). Although there has been accelerated growth of structures for protein monomers in the Protein Data Bank in recent years (11), the growth of structures for protein complexes has remained relatively small (9). Many factors, including their large size, transient nature, and dynamics of interactions, have prevented many complexes from being solved via traditional approaches in structural biology. Thus, the development of complementary analytical techniques with which to probe the structure of large protein complexes continues to evolve (13–18).Recent developments have advanced the analysis of protein structures and interaction by combining cross-linking and tandem mass spectrometry (17, 19–24). The basic idea behind this technique is to capture and identify pairs of amino acid residues that are spatially close to each other. When these linked pairs of residues are from the same protein (intraprotein cross-links), they provide distance constraints that help one infer the possible conformations of protein structures. Conversely, when pairs of residues come from different proteins (interprotein cross-links), they provide information about how proteins interact with one another. Although cross-linking strategies date back almost a decade (25, 26), difficulty in analyzing the complex MS/MS spectrum generated from linked peptides made this approach challenging, and therefore it was not widely used. With recent advances in mass spectrometry instrumentation, there has been renewed interest in employing this strategy to determine protein structures and identify protein–protein interactions. However, most studies thus far have been focused on purified protein complexes. With today''s mass spectrometers being capable of analyzing tens of thousands of spectra in a single experiment, it is now potentially feasible to extend this approach to the analysis of complex biological samples. Researchers have tried to realize this goal using both experimental and computational approaches. Indeed, a plethora of chemical cross-linking reagents are now available for stabilizing these complexes, and some are designed to allow for easier peptide identification when employed in concert with MS analysis (20, 27, 28). There have also been several recent efforts to develop computational methods for the automatic identification of linked peptides from MS/MS spectra (29–36). However, because of the lack of large annotated training data, most approaches to date either borrow fragmentation models learned from unlinked, linear peptides or learn the fragmentation statistics from training data of limited size (30, 37), which might not generalize well across different samples. In some cases it is possible to generate relatively large training data, but it is often very labor intensive and involves hundreds of separate LC-MS/MS runs (36). Here, employing disulfide-bridged peptides as an example, we propose a novel method that uses a combinatorial peptide library to (a) efficiently generate a large mass spectral reference dataset for linked peptides and (b) use these data to automatically train our new algorithm, MXDB, which can efficiently and accurately identify linked peptides from MS/MS spectra.     

New dsDNA binding unnatural oligopeptides with pyrimidine selectivity

Solid phase peptide library screening followed by extension of a lead recognition element for binding to a dsDNA sequence (NF binding site of IL6) using solution phase screening, delivered a new DNA binding peptide, Ac-Arg-Ual-Sar-Chi-Chi-Tal-Arg-CONH2. In the present research, the contribution of the different amino acid side chains to the binding strength of the peptide to dsDNA was investigated using an ethidium bromide displacement test. Based on these results, the lead structure was optimized by deconvolution. Eight new unnatural amino acids were evaluated at two positions of the heptapeptide replacing the Ual-Sar fragment. The strongest dsDNA binding was observed using ([(3-chlorophenyl)methyl]amino)acetic acid (Cbg) and beta-cyclohexyl-l-alanine (Cha) respectively, at those two positions. A 10-fold increase in affinity compared to the Ual-Sar sequence was obtained. Further enhancement of dsDNA binding was obtained with hybrid molecules linking the newly developed peptide fragment to an acridine derivative with a flexible spacer. This resulted in ligands with affinities in the microM range for the dsDNA target (K(d) of 2.1 x 10(-6) M). DNase I footprinting with the newly developed oligopeptide motifs showed the presence of a pronounced pyrimidine specificity, while conjugation to an intercalator seems to redirect the interaction to mixed sequences. This way, new unnatural oligopeptide motifs and hybrid molecules have been developed endowed with different sequence selectivities. The results demonstrate that the unnatural peptide library approach combined with subsequent modification of selected amino acid positions, is very suited for the discovery of novel sequence-specific dsDNA binding ligands.     

Use of an electronic nose system for diagnoses of urinary tract infections

The use of volatile production patterns produced by bacterial contaminants in urine samples were examined using electronic nose technology. In two experiments 25 and 45 samples from patients were analysed for specific bacterial contaminants using agar culture techniques and the major UTI bacterial species identified. These samples were also analysed by incubation in a volatile generation test tube system for 4-5 h. The volatile production patterns were then analysed using an electronic nose system with 14 conducting polymer sensors. In the first experiment analysis of the data using a neural network (NN) enabled identification of all but one of the samples correctly when compared to the culture information. Four groups could be distinguished, i.e. normal urine, Escherichia coli infected, Proteus spp. and Staphylococcus spp. In the second experiment it was again possible to use NN systems to examine the volatile production patterns and identify 18 of 19 unknown UTI cases. Only one normal patient sample was mis-identified as an E. coli infected sample. Discriminant function analysis also differentiated between normal urine samples, that infected with E. coli and with Staphylococcus spp. This study has shown the potential for early detection of microbial contaminants in urine samples using electronic nose technology for the first time. These findings will have implications for the development of rapid systems for use in clinical practice.     

Allosteric peptides bind a caspase zymogen and mediate caspase tetramerization

The caspases are a family of cytosolic proteases with essential roles in inflammation and apoptosis. Drug discovery efforts have focused on developing molecules directed against the active sites of caspases, but this approach has proved challenging and has not yielded any approved therapeutics. Here we describe a new strategy for generating inhibitors of caspase-6, a potential therapeutic target in neurodegenerative disorders, by screening against its zymogen form. Using phage display to discover molecules that bind the zymogen, we report the identification of a peptide that specifically impairs the function of caspase-6 in vitro and in neuronal cells. Remarkably, the peptide binds at a tetramerization interface that is uniquely present in zymogen caspase-6, rather than binding into the active site, and acts via a new allosteric mechanism that promotes caspase tetramerization. Our data illustrate that screening against the zymogen holds promise as an approach for targeting caspases in drug discovery.     

High-throughput crystallization of membrane proteins using the lipidic bicelle method

Membrane proteins (MPs) play a critical role in many physiological processes such as pumping specific molecules across the otherwise impermeable membrane bilayer that surrounds all cells and organelles. Alterations in the function of MPs result in many human diseases and disorders; thus, an intricate understanding of their structures remains a critical objective for biological research. However, structure determination of MPs remains a significant challenge often stemming from their hydrophobicity. MPs have substantial hydrophobic regions embedded within the bilayer. Detergents are frequently used to solubilize these proteins from the bilayer generating a protein-detergent micelle that can then be manipulated in a similar manner as soluble proteins. Traditionally, crystallization trials proceed using a protein-detergent mixture, but they often resist crystallization or produce crystals of poor quality. These problems arise due to the detergent's inability to adequately mimic the bilayer resulting in poor stability and heterogeneity. In addition, the detergent shields the hydrophobic surface of the MP reducing the surface area available for crystal contacts. To circumvent these drawbacks MPs can be crystallized in lipidic media, which more closely simulates their endogenous environment, and has recently become a de novo technique for MP crystallization. Lipidic cubic phase (LCP) is a three-dimensional lipid bilayer penetrated by an interconnected system of aqueous channels. Although monoolein is the lipid of choice, related lipids such as monopalmitolein and monovaccenin have also been used to make LCP. MPs are incorporated into the LCP where they diffuse in three dimensions and feed crystal nuclei. A great advantage of the LCP is that the protein remains in a more native environment, but the method has a number of technical disadvantages including high viscosity (requiring specialized apparatuses) and difficulties in crystal visualization and manipulation. Because of these technical difficulties, we utilized another lipidic medium for crystallization-bicelles (Figure 1). Bicelles are lipid/amphiphile mixtures formed by blending a phosphatidylcholine lipid (DMPC) with an amphiphile (CHAPSO) or a short-chain lipid (DHPC). Within each bicelle disc, the lipid molecules generate a bilayer while the amphiphile molecules line the apolar edges providing beneficial properties of both bilayers and detergents. Importantly, below their transition temperature, protein-bicelle mixtures have a reduced viscosity and are manipulated in a similar manner as detergent-solubilized MPs, making bicelles compatible with crystallization robots. Bicelles have been successfully used to crystallize several membrane proteins (Table 1). This growing collection of proteins demonstrates the versatility of bicelles for crystallizing both alpha helical and beta sheet MPs from prokaryotic and eukaryotic sources. Because of these successes and the simplicity of high-throughput implementation, bicelles should be part of every membrane protein crystallographer's arsenal. In this video, we describe the bicelle methodology and provide a step-by-step protocol for setting up high-throughput crystallization trials of purified MPs using standard robotics.     

Rapid isolation of viable circulating tumor cells from patient blood samples

Circulating tumor cells (CTC) are cells that disseminate from a primary tumor throughout the circulatory system and that can ultimately form secondary tumors at distant sites. CTC count can be used to follow disease progression based on the correlation between CTC concentration in blood and disease severity. As a treatment tool, CTC could be studied in the laboratory to develop personalized therapies. To this end, CTC isolation must cause no cellular damage, and contamination by other cell types, particularly leukocytes, must be avoided as much as possible. Many of the current techniques, including the sole FDA-approved device for CTC enumeration, destroy CTC as part of the isolation process (for more information see Ref. 2). A microfluidic device to capture viable CTC is described, consisting of a surface functionalized with E-selectin glycoprotein in addition to antibodies against epithelial markers. To enhance device performance a nanoparticle coating was applied consisting of halloysite nanotubes, an aluminosilicate nanoparticle harvested from clay. The E-selectin molecules provide a means to capture fast moving CTC that are pumped through the device, lending an advantage over alternative microfluidic devices wherein longer processing times are necessary to provide target cells with sufficient time to interact with a surface. The antibodies to epithelial targets provide CTC-specificity to the device, as well as provide a readily adjustable parameter to tune isolation. Finally, the halloysite nanotube coating allows significantly enhanced isolation compared to other techniques by helping to capture fast moving cells, providing increased surface area for protein adsorption, and repelling contaminating leukocytes. This device is produced by a straightforward technique using off-the-shelf materials, and has been successfully used to capture cancer cells from the blood of metastatic cancer patients. Captured cells are maintained for up to 15 days in culture following isolation, and these samples typically consist of >50% viable primary cancer cells from each patient. This device has been used to capture viable CTC from both diluted whole blood and buffy coat samples. Ultimately, we present a technique with functionality in a clinical setting to develop personalized cancer therapies.     

Subcutaneous bioavailability of therapeutic antibodies as a function of FcRn binding affinity in mice

The neonatal Fc receptor (FcRn) plays an important and well-known role in immunoglobulin G (IgG) catabolism; however, its role in the disposition of IgG after subcutaneous (SC) administration, including bioavailability, is relatively unknown. To examine the potential effect of FcRn on IgG SC bioavailability, we engineered three anti-amyloid β monoclonal antibody (mAb) reverse chimeric mouse IgG2a (mIgG2a) Fc variants (I253A.H435A, N434H and N434Y) with different binding affinities to mouse FcRn (mFcRn) and compared their SC bioavailability to that of the wild-type (WT) mAb in mice. Our results indicated that the SC bioavailability of mIgG2a was affected by mFcRn-binding affinity. Variant I253A.H435A, which did not bind to mFcRn at either pH 6.0 or pH 7.4, had the lowest bioavailability (41.8%). Variant N434Y, which had the greatest increase in binding affinity at both pH 6.0 and pH 7.4, had comparable bioavailability to the WT antibody (86.1% vs. 76.3%), whereas Variant N434H, which had modestly increased binding affinity at pH 6.0 to mFcRn and affinity comparable to the WT antibody at pH 7.4, had the highest bioavailability (94.7%). A semi-mechanism-based pharmacokinetic model, which described well the observed data with the WT antibody and variant I253A.H435A, is consistent with the hypothesis that the decreased bioavailability of variant I253A.H435A was due to loss of the FcRn-mediated protection from catabolism at the absorption site. Together, these data demonstrate that FcRn plays an important role in SC bioavailability of therapeutic IgG antibodies.Key words: monoclonal antibody, FcRn, binding affinity, subcutaneous bioavailability, semi-mechanism-based pharmacokinetic model     

A strategy for risk mitigation of antibodies with fast clearance

A majority of human therapeutic antibody candidates show pharmacokinetic properties suitable for clinical use, but an unexpectedly fast antibody clearance is sometimes observed that may limit the clinical utility. Pharmacokinetic data in cynomolgus monkeys collected for a panel of 52 antibodies showed broad distribution of target-independent clearance values (2.4–61.3 mL/day/kg), with 15 (29%) having clearance > 10 mL/day/kg. Alteration in the interaction with the recycling FcRn receptor did not account for the faster than expected clearance observed for the antibodies; off-target binding was presumed to account for the fast clearance. We developed an assay based on ELISA detection of non-specific binding to baculovirus particles that can identify antibodies having increased risk for fast clearance. This assay can be used during lead generation or optimization to identify antibodies with increased risk of having fast clearance in both humans and cynomolgus monkeys, and thus increase the likelihood of obtaining a suitable drug candidate.