Rapid On-Cell Selection of High-Performance Human Antibodies.

Phage display is a critical tool for developing antibodies. However, existing approaches require many time-consuming rounds of biopanning and screening of potential candidates due to a high rate of failure during validation. Herein, we present a rapid on-cell phage display platform which recapitulates the complex in vivo binding environment to produce high-performance human antibodies in a short amount of time. Selection is performed in a highly stringent heterogeneous mixture of cells to quickly remove nonspecific binders. A microfluidic platform then separates antigen-presenting cells with high throughput and specificity. An unsupervised machine learning algorithm analyzes sequences of phage from all pools to identify the structural trends that contribute to affinity and proposes ideal candidates for validation. In a proof-of-concept screen against human Frizzled-7, a key ligand in the Wnt signaling pathway, antibodies with picomolar affinity were discovered in two rounds of selection that outperformed current gold-standard reagents. This approach, termed µCellect, is low cost, high throughput, and compatible with a wide variety of cell types, enabling widespread adoption for antibody development.

Inhibition of Cancer Cell Adhesion, Migration and Proliferation by a Bispecific Antibody that Targets two Distinct Epitopes on αv Integrins.

Members of the αv family of integrins regulate activation of transforming growth factor beta (TGFß) and are directly involved in pro-tumorigenic phenotypes. Thus, αv integrins may be therapeutic targets for fibrosis and cancer, yet the isolation of selective inhibitors is currently a challenge. We generated synthetic antibodies selective for αv integrins by phage display selections on cell lines that displayed integrin heterodimers. We identified antibodies that targeted two distinct epitopes on cell-surface αv integrins and partially inhibited cell adhesion mediated by interactions between integrins and the latency-associated peptide, part of the pro-form of TGFß. Using the isolated antibody paratope sequences we engineered a bispecific antibody capable of binding to both epitopes simultaneously; this antibody potently and completely inhibited cell adhesion mediated by integrins αvß1, αvß3 and αvß5. In addition, the bispecific antibody inhibited proliferation and migration of lung carcinoma lines, where the highest and lowest potencies observed correlated with integrin-αv cell surface expression levels. Taken together, our results demonstrate that phage display selections with live cells can yield high quality anti-integrin antibodies, which we used as biparatopic building blocks to construct a bispecific antibody that strongly inhibited integrin function and may be a therapeutic candidate for cancer and fibrosis.

Comprehensive Assessment of the Relationship Between Site-2 Specificity and Helix α2 in the Erbin PDZ Domain.

PDZ domains are key players in signalling pathways. These modular domains generally recognize short linear C-terminal stretches of sequences in proteins that organize the formation of complex multi-component assemblies. The development of new methodologies for the characterization of the molecular principles governing these interactions is critical to fully understand the functional diversity of the family and to elucidate biological functions for family members. Here, we applied an in vitro evolution strategy to explore comprehensively the capacity of PDZ domains for specific recognition of different amino acids at a key position in C-terminal peptide ligands. We constructed a phage-displayed library of the Erbin PDZ domain by randomizing the binding site-2 and adjacent residues, which are all contained in helix α2, and we selected for variants binding to a panel of peptides representing all possible position-2 residues. This approach generated insights into the basis for the common natural class I and II specificities, demonstrated an alternative basis for a rare natural class III specificity for Asp-2, and revealed a novel specificity for Arg-2 that has not been reported in natural PDZ domains. A structure of a PDZ-peptide complex explained the minimum requirement for switching specificity from class I ligands containing Thr/Ser-2 to class II ligands containing hydrophobic residues at position-2. A second structure explained the molecular basis for the specificity for ligands containing Arg-2. Overall, the evolved PDZ variants greatly expand our understanding of site-2 specificities and the variants themselves may prove useful as building blocks for synthetic biology.

CellectSeq: In silico discovery of antibodies targeting integral membrane proteins combining in situ selections and next-generation sequencing.

Synthetic antibody (Ab) technologies are efficient and cost-effective platforms for the generation of monoclonal Abs against human antigens. Yet, they typically depend on purified proteins, which exclude integral membrane proteins that require the lipid bilayers to support their native structure and function. Here, we present an Ab discovery strategy, termed CellectSeq, for targeting integral membrane proteins on native cells in complex environment. As proof of concept, we targeted three transmembrane proteins linked to cancer, tetraspanin CD151, carbonic anhydrase 9, and integrin-α11. First, we performed in situ cell-based selections to enrich phage-displayed synthetic Ab pools for antigen-specific binders. Then, we designed next-generation sequencing procedures to explore Ab diversities and abundances. Finally, we developed motif-based scoring and sequencing error-filtering algorithms for the comprehensive interrogation of next-generation sequencing pools to identify Abs with high diversities and specificities, even at extremely low abundances, which are very difficult to identify using manual sampling or sequence abundances.

In situ antibody phage display yields optimal inhibitors of integrin α11/ß1.

Integrins are transmembrane multi-conformation receptors that mediate interactions with the extracellular matrix. In cancer, integrins influence metastasis, proliferation, and survival. Collagen-binding integrin-α11/ß1, a marker of aggressive tumors that is involved in stroma-tumor crosstalk, may be an attractive target for anti-cancer therapeutic antibodies. We performed selections with phage-displayed synthetic antibody libraries for binding to either purified integrin-α11/ß1 or in situ on live cells. The in-situ strategy yielded many diverse antibodies, and strikingly, most of these antibodies did not recognize purified integrin-α11/ß1. Conversely, none of the antibodies selected for binding to purified integrin-α11/ß1 were able to efficiently recognize native cell-surface antigen. Most importantly, only the in-situ selection yielded functional antibodies that were able to compete with collagen-I for binding to cell-surface integrin-α11/ß1, and thus inhibited cell adhesion. In-depth characterization of a subset of in situ-derived clones as full-length immunoglobulins revealed high affinity cellular binding and inhibitory activities in the single-digit nanomolar range. Moreover, the antibodies showed high selectivity for integrin-α11/ß1 with minimal cross-reactivity for close homologs. Taken together, our findings highlight the advantages of in-situ selections for generation of anti-integrin antibodies optimized for recognition and inhibition of native cell-surface proteins, and our work establishes general methods that could be extended to many other membrane proteins.

Large-scale survey and database of high affinity ligands for peptide recognition modules.

Many proteins involved in signal transduction contain peptide recognition modules (PRMs) that recognize short linear motifs (SLiMs) within their interaction partners. Here, we used large-scale peptide-phage display methods to derive optimal ligands for 163 unique PRMs representing 79 distinct structural families. We combined the new data with previous data that we collected for the large SH3, PDZ, and WW domain families to assemble a database containing 7,984 unique peptide ligands for 500 PRMs representing 82 structural families. For 74 PRMs, we acquired enough new data to map the specificity profiles in detail and derived position weight matrices and binding specificity logos based on multiple peptide ligands. These analyses showed that optimal peptide ligands resembled peptides observed in existing structures of PRM-ligand complexes, indicating that a large majority of the phage-derived peptides are likely to target natural peptide-binding sites and could thus act as inhibitors of natural protein-protein interactions. The complete dataset has been assembled in an online database (http://www.prm-db.org) that will enable many structural, functional, and biological studies of PRMs and SLiMs.

Predicted amino acid motif repeats in proteins potentially encode extensive multivalent macromolecular assemblies in the human proteome.

There are emerging interests in understanding higher order assemblies of biopolymers within and between cells, such as protein-protein and protein-RNA biomolecular condensates. These biomolecular condensates are thought to assemble/disassemble via multivalent interactions, including those mediated particularly by unique repeated amino acid motifs (URM). We asked how common are proteins with such URMs, their incidence and abundance, by exhaustively enumerating repeating motifs of length 3-10 in the human proteome. We found that URMs are very common and widely distributed across the human proteome. Moreover, the number of repetitions and intervals between them do not correlate with their lengths, which suggests that the number of repeats among proteins in the proteome is independent of length, contrary to the notion that short motifs are more abundant then long motifs. Finally, we describe two examples of URMs in proteins known to form higher order biopolymer assemblies: multi-PDZ domain-containing proteins and the FUS family of RNA binding proteins. For the FUS family, we predicted a known sequence 'grammar', specific motifs and interval sequence compositions that are essential to phase separation and material properties of condensates formed by this family of proteins. In PDZ domain-containing proteins we found a novel repeated motif that was surprisingly both within and between individual PDZ domains. We speculate that these motifs could be binding sites for multivalent interactions, a residual result of the mechanism by which PDZ-domain duplications occurred or that the linker sequences between PDZ domains may encode cryptic PDZ domains.

Exhaustive search of linear information encoding protein-peptide recognition.

High-throughput in vitro methods have been extensively applied to identify linear information that encodes peptide recognition. However, these methods are limited in number of peptides, sequence variation, and length of peptides that can be explored, and often produce solutions that are not found in the cell. Despite the large number of methods developed to attempt addressing these issues, the exhaustive search of linear information encoding protein-peptide recognition has been so far physically unfeasible. Here, we describe a strategy, called DALEL, for the exhaustive search of linear sequence information encoded in proteins that bind to a common partner. We applied DALEL to explore binding specificity of SH3 domains in the budding yeast Saccharomyces cerevisiae. Using only the polypeptide sequences of SH3 domain binding proteins, we succeeded in identifying the majority of known SH3 binding sites previously discovered either in vitro or in vivo. Moreover, we discovered a number of sites with both non-canonical sequences and distinct properties that may serve ancillary roles in peptide recognition. We compared DALEL to a variety of state-of-the-art algorithms in the blind identification of known binding sites of the human Grb2 SH3 domain. We also benchmarked DALEL on curated biological motifs derived from the ELM database to evaluate the effect of increasing/decreasing the enrichment of the motifs. Our strategy can be applied in conjunction with experimental data of proteins interacting with a common partner to identify binding sites among them. Yet, our strategy can also be applied to any group of proteins of interest to identify enriched linear motifs or to exhaustively explore the space of linear information encoded in a polypeptide sequence. Finally, we have developed a webserver located at http://michnick.bcm.umontreal.ca/dalel, offering user-friendly interface and providing different scenarios utilizing DALEL.

Evolution of domain-peptide interactions to coadapt specificity and affinity to functional diversity.

Evolution of complexity in eukaryotic proteomes has arisen, in part, through emergence of modular independently folded domains mediating protein interactions via binding to short linear peptides in proteins. Over 30 years, structural properties and sequence preferences of these peptides have been extensively characterized. Less successful, however, were efforts to establish relationships between physicochemical properties and functions of domain-peptide interactions. To our knowledge, we have devised the first strategy to exhaustively explore the binding specificity of protein domain-peptide interactions. We applied the strategy to SH3 domains to determine the properties of their binding peptides starting from various experimental data. The strategy identified the majority (∼70%) of experimentally determined SH3 binding sites. We discovered mutual relationships among binding specificity, binding affinity, and structural properties and evolution of linear peptides. Remarkably, we found that these properties are also related to functional diversity, defined by depth of proteins within hierarchies of gene ontologies. Our results revealed that linear peptides evolved to coadapt specificity and affinity to functional diversity of domain-peptide interactions. Thus, domain-peptide interactions follow human-constructed gene ontologies, which suggest that our understanding of biological process hierarchies reflect the way chemical and thermodynamic properties of linear peptides and their interaction networks, in general, have evolved.

Fast and accurate discovery of degenerate linear motifs in protein sequences.

Linear motifs mediate a wide variety of cellular functions, which makes their characterization in protein sequences crucial to understanding cellular systems. However, the short length and degenerate nature of linear motifs make their discovery a difficult problem. Here, we introduce MotifHound, an algorithm particularly suited for the discovery of small and degenerate linear motifs. MotifHound performs an exact and exhaustive enumeration of all motifs present in proteins of interest, including all of their degenerate forms, and scores the overrepresentation of each motif based on its occurrence in proteins of interest relative to a background (e.g., proteome) using the hypergeometric distribution. To assess MotifHound, we benchmarked it together with state-of-the-art algorithms. The benchmark consists of 11,880 sets of proteins from S. cerevisiae; in each set, we artificially spiked-in one motif varying in terms of three key parameters, (i) number of occurrences, (ii) length and (iii) the number of degenerate or "wildcard" positions. The benchmark enabled the evaluation of the impact of these three properties on the performance of the different algorithms. The results showed that MotifHound and SLiMFinder were the most accurate in detecting degenerate linear motifs. Interestingly, MotifHound was 15 to 20 times faster at comparable accuracy and performed best in the discovery of highly degenerate motifs. We complemented the benchmark by an analysis of proteins experimentally shown to bind the FUS1 SH3 domain from S. cerevisiae. Using the full-length protein partners as sole information, MotifHound recapitulated most experimentally determined motifs binding to the FUS1 SH3 domain. Moreover, these motifs exhibited properties typical of SH3 binding peptides, e.g., high intrinsic disorder and evolutionary conservation, despite the fact that none of these properties were used as prior information. MotifHound is available (http://michnick.bcm.umontreal.ca or http://tinyurl.com/motifhound) together with the benchmark that can be used as a reference to assess future developments in motif discovery.

DNAc: A clustering method for identifying kinship relations between DNA profiles using a novel similarity measure.

After decades of refinement, DNA testing methods have become essential tools in forensic sciences. They are essentially based on likelihood ratio test principle, which is utilized specifically, by using as prior knowledge the allele frequencies in the population, to confirm or refute a given kinship hypothesis made on two genotypes. This makes these methods ill suited when allele frequencies or kinship hypotheses are unavailable. In this paper, we introduce DNAc, a new clustering methodology for DNA testing based on a new similarity measure that allows an accurate retrieval of the degree of relatedness among two or more genotypes, without relying on kinship hypotheses or allele frequencies in the population. We used DNAc in analyzing microsatellite DNA sequences distributed among 12 genotypes from normal individuals from two distinct families. The results show that DNAc accurately determines kinship among genotypes and further gathers them in the appropriate kinship groups.

CLUSS2: an alignment-independent algorithm for clustering protein families with multiple biological functions.

CLUSS is an algorithm proposed for clustering both alignable and non-alignable protein sequences. However, CLUSS tends to be ineffective on protein datasets that include a large number of biochemical activities. To overcome this difficulty, we propose in this paper a new algorithm, named CLUSS2 that scales better with the increase of the number of biochemical activities. CLUSS2 differs from CLUSS in many ways including protein sequences representation, conserved motifs extraction and time efficiency. Our experiments show that CLUSS2 more accurately highlights the functional characteristics of the clustered families, especially for those with a large number of biochemical activities.

CLUSS: clustering of protein sequences based on a new similarity measure.

BACKGROUND: The rapid burgeoning of available protein data makes the use of clustering within families of proteins increasingly important. The challenge is to identify subfamilies of evolutionarily related sequences. This identification reveals phylogenetic relationships, which provide prior knowledge to help researchers understand biological phenomena. A good evolutionary model is essential to achieve a clustering that reflects the biological reality, and an accurate estimate of protein sequence similarity is crucial to the building of such a model. Most existing algorithms estimate this similarity using techniques that are not necessarily biologically plausible, especially for hard-to-align sequences such as proteins with different domain structures, which cause many difficulties for the alignment-dependent algorithms. In this paper, we propose a novel similarity measure based on matching amino acid subsequences. This measure, named SMS for Substitution Matching Similarity, is especially designed for application to non-aligned protein sequences. It allows us to develop a new alignment-free algorithm, named CLUSS, for clustering protein families. To the best of our knowledge, this is the first alignment-free algorithm for clustering protein sequences. Unlike other clustering algorithms, CLUSS is effective on both alignable and non-alignable protein families. In the rest of the paper, we use the term "phylogenetic" in the sense of "relatedness of biological functions". RESULTS: To show the effectiveness of CLUSS, we performed an extensive clustering on COG database. To demonstrate its ability to deal with hard-to-align sequences, we tested it on the GH2 family. In addition, we carried out experimental comparisons of CLUSS with a variety of mainstream algorithms. These comparisons were made on hard-to-align and easy-to-align protein sequences. The results of these experiments show the superiority of CLUSS in yielding clusters of proteins with similar functional activity. CONCLUSION: We have developed an effective method and tool for clustering protein sequences to meet the needs of biologists in terms of phylogenetic analysis and prediction of biological functions. Compared to existing clustering methods, CLUSS more accurately highlights the functional characteristics of the clustered families. It provides biologists with a new and plausible instrument for the analysis of protein sequences, especially those that cause problems for the alignment-dependent algorithms.