Complex Libraries Research Articles

Computer-aided drug design and virtual screening of targeted combinatorial libraries of mixed-ligand transition metal complexes of 2-butanone thiosemicarbazone

The present paper deals with in silico evaluation of 32 virtually designed transition metal complexes of 2-butanone thiosemicarbazone and N,S,O containing donor hetero-ligands namely py, bpy, furan, thiophene, 2-picoline, 1,10-phenanthroline, piperazine and liquid ammonia. The complexes were designed with a view to assess their potential anticancer, antioxidant and antibacterial activity. The absorption, distribution, metabolism, excretion and toxicity (ADMET) properties of the chosen ligands were calculated by admetSAR software. Metabolic sites of different ligands likely to undergo metabolism were predicted using Metaprint 2D. The proposed complexes were also evaluated for their drug-like quality based on Lipinski’s, Veber, Ghose and leadlikeness filters. Druglikeness and toxicity potential were predicted by OSIRIS property explorer. The pharmacokinetic properties and bioactivity scores were calculated by Molinspiration tool. Bioactivity scores of the complexes were predicted for drug targets including enzymes, nuclear receptors, kinase inhibitors, G-protein coupled receptor ligands and ion channel modulators. Molecular docking of selected Fe(II) mixed-ligand complexes was performed using AutoDock version 4.2.6 and i-GEMDOCK version 2.1 with two target proteins namely Ribonucleotide reductase (RR) and Topoisomerase II (Topo II). The results were compared with three standard reference drugs viz. Doxorubicin HCl, Letrozole (anticancer) and Tetracycline (antibiotic). Multivariate data obtained were analyzed using principal component analysis (PCA) for visualization and projection as scatter and 3D plots. Positive results obtained for hetero-ligands using admetSAR version 1.0 indicated good absorption and transport kinetics of the hetero-ligand complexes through the human intestine and blood-brain barrier. The hetero-ligands were predicted to have no associated mutagenic effect(s) also. However, none of the hetero-ligands was predicted to be Caco-2 (human colon cancer cell line) permeable. Most of the hetero-ligands and the parent ligand (2-butanone thiosemicarbazone) were predicted to undergo Phase-I metabolism prior to excretion using MetaPrint2D software. Pharmacokinetic evaluation of the proposed complexes revealed that all complexes displayed drug-like character and were predicted to have no apparent toxic side-effects. All the proposed complexes showed moderate to good biological activity scores (−5.0 to 0.0). Mixed complexes with bpy, 2-picoline and 1,10-phenanthroline showed significant bioactivity scores (as enzyme inhibitors) in the range 0.02–0.13. Likewise, good docking scores were obtained for Fe (II) complexes with the same ligands. [FeL(2-pic)2] displayed the lowest binding energy (−6.47 kcal/mol) with respect to Topo II followed by [FeL(py)2] (−6.14 kcal/mol) as calculated by AutoDock version 4.2.6. With respect to binding with RR, [FeL(2--pic)2] again displayed the lowest binding energy (−7.21 kcal/mol) followed by [FeL(py)2] (−5.96 kcal/mol). On the basis of docking predictions and various other computational evaluations, four mixed-ligand complexes of Fe in +2 oxidation state with py, bpy, 2--picoline and 1,10-phenanthroline were synthesized with 2-butanone thiosemicarbazone. All the synthesized Fe complexes were characterized using various spectroscopic techniques and tested for their potential anticancer activity in vitro against human breast carcinoma cell line MDA-MB 231 and human lung carcinoma cell line A549 cell line using MTT assay. [FeL(2-pic)2], [FeL(bpy)], and [FeL(py)2] were found to exhibit significant antiproliferative activity with IC50 values in the range of 80–100 μM against breast and lung cancer cells. The synthesized Fe complexes also displayed mild antioxidant activities. The synthesized and studied Fe complexes have the potential for development into future anticancer agents if analyzed and modified further for improvement of their ADMET, solubility and permeability criteria set for potential drug-candidates.

Antigen discovery for regulatory T cells (Tregs) in type-1 diabetes (T1D)

Abstract Antigen specific Tregs are more efficient than polyclonal Tregs in their suppressive capacity, which makes them attractive tools for further targeted tolerogenic therapy. Autoreactive T cells specificity has been well characterized in type 1-diabetes but little is known about autoreactive Treg specificity. Whether Tregs recognize the same tissue-specific self-antigens as autoreactive T cell, or they recognize different self-antigens to regulate local immune responses via bystander suppression is unclear. Our goal is to identify antigen specificity from enriched islet infiltrated Treg by using previously validated peptide-major histocompatibility complex (pMHC) library displayed on yeast. We first performed single cell TCR sequencing on Tregs from the pancreatic islet of 12 weeks old NOD mice. Data show an abundance of the TCR allele TRAV5D4 usage in islet infiltrated Tregs that has been characterized to target a primary insulin peptide. Moreover, we also performed single cell RNAseq together with TCR sequencing on islet infiltrated CD4+ T cells from a T1D patient from the Network for Pancreatic Organ Donors with Diabetes. We were able to identify enriched TCRs with Treg gene expression profile that will be used to screen for antigen recognition on a DR4 pMHC yeast library. New antigen identification for pancreatic islet-infiltrated Tregs will allow us to improve our understanding of Treg deficiencies in T1D and may lead us to design new immunosuppressive therapies.

Heteroleptic Copper(I)-Based Complexes for Photocatalysis: Combinatorial Assembly, Discovery, and Optimization.

A library of 50 copper-based complexes derived from bisphosphines and diamines was prepared and evaluated in three mechanistically distinct photocatalytic reactions. In all cases, a copper-based catalyst was identified to afford high yields, where new heteroleptic complexes derived from the bisphosphine BINAP displayed high efficiency across all reaction types. Importantly, the evaluation of the library of copper complexes revealed that even when photophysical data is available, it is not always possible to predict which catalyst structure will be efficient or inefficient in a given process, emphasizing the advantages for catalyst structures with high modularity and structural variability.

Substitution-Inert Polynuclear Platinum Complexes as Metalloshielding Agents for Heparan Sulfate.

Cleavage of heparan sulfate proteoglycans (HSPGs) by the enzyme heparanase modulates tumour-related events including angiogenesis, cell invasion, and metastasis. Metalloshielding of heparan sulfate (HS) by positively charged polynuclear platinum complexes (PPCs) effectively inhibits physiologically critical HS functions. Studies using bacterial P. heparinus heparinase II showed that a library of Pt complexes varying in charge and nuclearity and the presence or absence of a dangling amine inhibits the cleavage activity of the enzyme on the synthetic pentasaccharide, Fondaparinux (FPX). Charge-dependent affinity of PPC for FPX was seen in competition assays with methylene blue and ethidium bromide. The dissociation constant (Kd ) of TriplatinNC for FPX was directly measured by isothermal titration calorimetry (ITC). The trend in DFT calculated interaction energies with heparin fragments is consistent with the spectroscopic studies. Competitive inhibition of TAMRA-R9 internalization in human carcinoma (HCT116) cells along with studies in HCT116, wildtype CHO and mutant CHO-pgsA745 (lacking HS/CS) cells confirm that HSPG-mediated interactions play an important role in the cellular accumulation of PPCs.

Heteroleptic Copper(I)‐Based Complexes for Photocatalysis: Combinatorial Assembly, Discovery, and Optimization

AbstractA library of 50 copper‐based complexes derived from bisphosphines and diamines was prepared and evaluated in three mechanistically distinct photocatalytic reactions. In all cases, a copper‐based catalyst was identified to afford high yields, where new heteroleptic complexes derived from the bisphosphine BINAP displayed high efficiency across all reaction types. Importantly, the evaluation of the library of copper complexes revealed that even when photophysical data is available, it is not always possible to predict which catalyst structure will be efficient or inefficient in a given process, emphasizing the advantages for catalyst structures with high modularity and structural variability.

Searching for new redox-complexes\\in organic flow batteries

The study of redox couples based on Fe(III)/(II) and Co(II)/(I) organic complexes has demonstrated chemically reversible redox processes as well as good stability in organic solvents. These active complexes, obtained with polypyridine ligands, present low cost, low toxicity and good chemical stability. Moreover, they demonstrated fast redox kinetics and for that, they are candidate for active species in redox flow cells. A wide library of polypyridine complexes have been prepared and tested as acceptor ligands to reach an open circuit voltage up to 2 V, in a mixture of ethylene carbonate and propylene carbonate (EC/PC) chosen for their low volatility and electrochemical stability. Solubility data are presented after tuning ligand design to optimize metal-complex solubility. The best compounds were [Fe(bpy)3]Tf2 (Tf = CF3SO3−, bpy = 2.2′-bipyridine) and [Co(bpy)3]Tf2 which generated current densities of the order of 30 mA/cm2 in thin layer static cells. These complexes were also preliminary tested in a complete flow cell equipped with a Nafion membrane, with LiTf electrolyte, and ca. 90% coulombic efficiency was observed. The decrease of performance observed after 8 h is under investigation and assigned, for now, to membrane degradation. A change of membrane characteristics should be considered to exploit the full potentiality of these redox mediators.

How to Bend the Uranyl Cation via Crystal Engineering.

Bending the linear uranyl (UO22+) cation represents both a significant challenge and opportunity within the field of actinide hybrid materials. As part of related efforts to engage the nominally terminal oxo atoms of uranyl cation in noncovalent interactions, we synthesized a new uranyl complex, [UO2(C12H8N2)2(C7H2Cl3O2)2]·2H2O (complex 2), that featured both deviations from equatorial planarity and uranyl linearity from simple hydrothermal conditions. Based on this complex, we developed an approach to probe the nature and origin of uranyl bending within a family of hybrid materials, which was done via the synthesis of complexes 1-3 that display significant deviations from equatorial planarity and uranyl linearity (O-U-O bond angles between 162° and 164°) featuring 2,4,6-trihalobenzoic acid ligands (where Hal = F, Cl, and Br) and 1,10-phenanthroline, along with nine additional "nonbent" hybrid materials that either coformed with the "bent" complexes (4-6) or were prepared as part of complementary efforts to understand the mechanism(s) of uranyl bending (7-12). Complexes were characterized via single crystal X-ray diffraction and Raman, infrared (IR), and luminescence spectroscopy, as well as via quantum chemical calculations and density-based quantum theory of atoms in molecules (QTAIM) analysis. Looking comprehensively, these results are compared with the small library of bent uranyl complexes in the literature, and herein we computationally demonstrate the origin of uranyl bending and delineate the energetics behind this process.

Darwin Assembly: fast, efficient, multi-site bespoke mutagenesis.

Engineering proteins for designer functions and biotechnological applications almost invariably requires (or at least benefits from) multiple mutations to non-contiguous residues. Several methods for multiple site-directed mutagenesis exist, but there remains a need for fast and simple methods to efficiently introduce such mutations – particularly for generating large, high quality libraries for directed evolution. Here, we present Darwin Assembly, which can deliver high quality libraries of >108 transformants, targeting multiple (>10) distal sites with minimal wild-type contamination (<0.25% of total population) and which takes a single working day from purified plasmid to library transformation. We demonstrate its efficacy with whole gene codon reassignment of chloramphenicol acetyl transferase, mutating 19 codons in a single reaction in KOD DNA polymerase and generating high quality, multiple-site libraries in T7 RNA polymerase and Tgo DNA polymerase. Darwin Assembly uses commercially available enzymes, can be readily automated, and offers a cost-effective route to highly complex and customizable library generation.

A Universal Protocol for Large-scale gRNA Library Production from any DNA Source.

The popularity of the CRISPR/Cas9 system for both genome and epigenome engineering stems from its simplicity and adaptability. An effector (the Cas9 nuclease or a nuclease-dead dCas9 fusion protein) is targeted to a specific site in the genome by a small synthetic RNA known as the guide RNA, or gRNA. The bipartite nature of the CRISPR system enables its use in screening approaches since plasmid libraries containing expression cassettes of thousands of individual gRNAs can be used to interrogate many different sites in a single experiment. To date, gRNA sequences for the construction of libraries have been almost exclusively generated by oligonucleotide synthesis, which limits the achievable complexity of sequences in the library and is relatively cost-intensive. Here, a detailed protocol for CORALINA (comprehensive gRNA library generation through controlled nuclease activity), a simple and cost-effective method for the generation of highly complex gRNA libraries based on enzymatic digestion of input DNA, is described. Since CORALINA libraries can be generated from any source of DNA, plenty of options for customization exist, enabling a large variety of CRISPR-based screens.

A Universal Protocol for Large-scale gRNA Library Production from any DNA Source

The popularity of the CRISPR/Cas9 system for both genome and epigenome engineering stems from its simplicity and adaptability. An effector (the Cas9 nuclease or a nuclease-dead dCas9 fusion protein) is targeted to a specific site in the genome by a small synthetic RNA known as the guide RNA, or gRNA. The bipartite nature of the CRISPR system enables its use in screening approaches since plasmid libraries containing expression cassettes of thousands of individual gRNAs can be used to interrogate many different sites in a single experiment. To date, gRNA sequences for the construction of libraries have been almost exclusively generated by oligonucleotide synthesis, which limits the achievable complexity of sequences in the library and is relatively cost-intensive. Here, a detailed protocol for CORALINA (comprehensive gRNA library generation through controlled nuclease activity), a simple and cost-effective method for the generation of highly complex gRNA libraries based on enzymatic digestion of input DNA, is described. Since CORALINA libraries can be generated from any source of DNA, plenty of options for customization exist, enabling a large variety of CRISPR-based screens.

Inverting the Swelling Trends in Modular Self‐Oscillating Gels Crosslinked by Redox‐Active Metal Bipyridine Complexes

AbstractThe developing field of active, stimuli‐responsive materials is in need for new dynamic architectures that may offer unprecedented chemomechanical switching mechanisms. Toward this goal, syntheses of polymerizable bipyridine ligands, bis(4‐vinylbenzyl)[2,2′‐bipyridine]‐4,4′‐dicarboxylate and N4,N4′‐bis(4‐vinylphenyl)‐2,2′‐bipyridine‐4,4′‐dicarboxamide, and a number of redox‐active Ruthenium(II) and Iron(II) complexes with them are reported. Detailed characterizations by NMR, Fourier transform infrared spectroscopy, high‐resolution mass‐spectrometry, X‐ray, and cyclic voltammetry show that the topology of these molecules allows them to serve as both comonomers and crosslinkers in polymerization reactions. Electronic properties of the ligands are tunable by choosing carboxylate‐ or carboxamido‐linkages between the core and the vinylaryl moieties, leading to a library of Ru and Fe complexes with the M(III)/M(II) standard redox potentials suitable for catalyzing self‐oscillating Belousov–Zhabotinskii (BZ) reaction. New poly(N‐isopropylacrylamide)‐based redox‐responsive functional gels containing hydrophilic comonomers, which have been prepared using representative Ru bpy complexes as both a crosslinker and redox‐active catalyst, exhibit a unique feature: their swelling/contraction mode switches its dependence on the oxidation state of the Ru center, upon changing the ratio of comonomers in the hybrid gel network. The BZ self‐oscillations of such crosslinked hydrogels have been observed and quantified for both supported film and free‐standing gel samples, demonstrating their potential as chemomechanically active modules for new functional materials.

Genetic Constructor: An Online DNA Design Platform.

Genetic Constructor is a cloud Computer Aided Design (CAD) application developed to support synthetic biologists from design intent through DNA fabrication and experiment iteration. The platform allows users to design, manage, and navigate complex DNA constructs and libraries, using a new visual language that focuses on functional parts abstracted from sequence. Features like combinatorial libraries and automated primer design allow the user to separate design from construction by focusing on functional intent, and design constraints aid iterative refinement of designs. A plugin architecture enables contributions from scientists and coders to leverage existing powerful software and connect to DNA foundries. The software is easily accessible and platform agnostic, free for academics, and available in an open-source community edition. Genetic Constructor seeks to democratize DNA design, manufacture, and access to tools and services from the synthetic biology community.

Protein interaction perturbation profiling at amino-acid resolution.

The identification of genomic variants in healthy and diseased individuals continues to rapidly outpace our ability to functionally annotate these variants. Techniques that both systematically assay the functional consequences of nucleotide-resolution variation and can scale to hundreds of genes are urgently required. We designed a sensitive yeast two-hybrid-based 'off switch' for positive selection of interaction-disruptive variants from complex genetic libraries. Combined with massively parallel programmed mutagenesis and a sequencing readout, this method enables systematic profiling of protein-interaction determinants at amino-acid resolution. We defined >1,000 interaction-disrupting amino acid mutations across eight subunits of the BBSome, the major human cilia protein complex associated with the pleiotropic genetic disorder Bardet-Biedl syndrome. These high-resolution interaction-perturbation profiles provide a framework for interpreting patient-derived mutations across the entire protein complex and thus highlight how the impact of disease variation on interactome networks can be systematically assessed.

Greener solvents for solid-phase organic synthesis

Abstract Solid-phase organic synthesis is an essential method for the rapid synthesis of complex biological structures and libraries of small organic molecules. However, it is often associated with the use of large quantities of problematic solvents for the removal of excess reagents and reaction by-products. Given that solvent will often be the biggest contributor to waste generated in the average pharmaceutical/fine-chemical process, its exchange for a more desirable alternative often presents the biggest gains in terms of reducing environmental impact. This review aims to explore recent approaches to performing solid-phase organic synthesis, and associated solid-phase peptide synthesis, in neoteric solvents and reaction media that present greener alternatives.

Sequence homology between HLA-bound cytomegalovirus and human peptides: A potential trigger for alloreactivity.

Human cytomegalovirus (hCMV) reactivation may often coincide with the development of graft-versus-host-disease (GVHD) in stem cell transplantation (SCT). Seventy seven SCT donor-recipient pairs (DRP) (HLA matched unrelated donor (MUD), n = 50; matched related donor (MRD), n = 27) underwent whole exome sequencing to identify single nucleotide polymorphisms (SNPs) generating alloreactive peptide libraries for each DRP (9-mer peptide-HLA complexes); Human CMV CROSS (Cross-Reactive Open Source Sequence) database was compiled from NCBI; HLA class I binding affinity for each DRPs HLA was calculated by NetMHCpan 2.8 and hCMV- derived 9-mers algorithmically compared to the alloreactive peptide-HLA complex libraries. Short consecutive (≥6) amino acid (AA) sequence homology matching hCMV to recipient peptides was considered for HLA-bound-peptide (IC50<500nM) cross reactivity. Of the 70,686 hCMV 9-mers contained within the hCMV CROSS database, an average of 29,658 matched the MRD DRP alloreactive peptides and 52,910 matched MUD DRP peptides (p<0.001). In silico analysis revealed multiple high affinity, immunogenic CMV-Human peptide matches (IC50<500 nM) expressed in GVHD-affected tissue-specific manner. hCMV+GVHD was found in 18 patients, 13 developing hCMV viremia before GVHD onset. Analysis of patients with GVHD identified potential cross reactive peptide expression within affected organs. We propose that hCMV peptide sequence homology with human alloreactive peptides may contribute to the pathophysiology of GVHD.

Thiosemicarbazone Dynamic Combinatorial Chemistry: Equilibrator-Induced Dynamic State, Formation of Complex Libraries, and a Supramolecular On/Off Switch.

Dynamic combinatorial libraries that equilibrate under thermodynamic control and can be trapped kinetically when desired are key to creating complex systems that can mimic dynamic biological systems, such as the biochemical system of life. A much-sought-after feature is the ability to turn off the dynamic exchange of the system, in order to investigate a transient state away from thermodynamic equilibrium, and then turn on the dynamic exchange again. We describe here the first use of thiosemicarbazone exchange to form dynamic combinatorial libraries. The libraries were found to require a nucleophilic catalyst, or equilibrator, in order to reach thermodynamic equilibrium. This equilibrator approach adds a supramolecular level of control over the dynamic system and allows the dynamic exchange to be turned off by addition of 18-crown-6, which binds the equilibrator in a nonnucleophilic complex. The dynamic exchange can be restarted by addition of potassium ions that competitively bind 18-crown-6, thus liberating the equilibrator. The highly complex thiosemicarbazone-based macrocyclic libraries contain both [2]catenanes and sequence isomers, which can be distinguished by HPLC-MS/MS.

Abstract 871: [18F]BMS-986192 as a novel PET imaging agent for assessment of PD-L1 expression in vivo

Abstract Objectives Inhibition of the Programmed Death Ligand-1 (PD-L1)/PD-1 interaction allows for potent anti-tumor activity and antibodies that disrupt this interaction have been approved for the treatment of multiple cancer types. PD-L1 expression has been investigated clinically as a potential biomarker to predict response to anti-PD-1/PD-L1 therapy. BMS-986192, an Adnectin with high affinity and specificity for human PD-L1, was selected in vitro from a complex library. Here we report the discovery and first preclinical evaluation of [18F]BMS-986192 as a PET imaging agent to detect PD-L1 expression in vivo. Methods [18F]BMS-986192 was radiolabeled via copper-free click chemistry and assessed for its ability to detect PD-L1 expression. Tracer binding to human L2987 (PD-L1+) and HT-29 (PD-L1-) xenografts as well as human non-small cell lung cancer (NSCLC) tissue samples was assessed by autoradiography (ARG). Tracer binding was compared to PD-L1 expression assessed independently with anti-PD-L1 immunohistochemistry (IHC). In vivo performance of the tracer was also assessed by PET imaging in mice bearing bilateral L2987 and HT-29 xenografts, and tracer biodistribution was further assayed in these animals ex vivo by gamma counter. Finally, initial in vivo biodistribution and radiation dosimetry was measured by PET in cynomolgus monkey. Results ARG studies showed increased [18F]BMS-986192 total binding to PD-L1(+) L2987 xenograft compared to PD-L1(-) HT-29 xenograft tissue. Radiotracer binding was higher in all tested human NSCLC tissue samples compared to xenografts. Dose-dependent blockade was seen in all PD-L1(+) tissues co-incubated with cold BMS-986192, and binding was unaffected by co-incubation with cold non-PD-L1 binding control. Visual comparison of tracer binding aligns closely with PD-L1 IHC both spatially as well as in intensity. Preferential accumulation of [18F]BMS-986192 was noted in PD-L1(+) L2987 compared to PD-L1(-) HT-29 xenografts in tumor-bearing mice. PET studies in cynomolgus monkeys confirmed binding to PD-L1(+) tissue (e.g. spleen) with minimal nonspecific background signal exclusive of primary clearance organs. Radiation dosimetry of [18F]BMS-986192 indicates an estimated single administration dose limit of 6.2 mCi for an average human subject. Conclusions ARG, PET studies, and ex vivo measurements in rodent and cynomolgus monkey demonstrated sensitive and specific [18F]BMS-986192 binding to PD-L1. Low background signal in cynomolgus monkey in the context of endogenous PD-L1 expression further supports the potential of this tracer for sensitive detection of PD-L1(+) lesions in vivo. Radiation dosimetry suggests that [18F]BMS-986192 can be safely administered in human trials, with estimated absorbed radiation doses well within safe parameters for human administration. [18F]BMS-986192 has potential as a sensitive PD-L1 imaging agent for same-day imaging in patients. Citation Format: Ralph A. Smith, David Donnelly, Paul E. Morin, Dasa Lipovsek, Jochem Gokemeijer, Daniel Cohen, Joonyoung Kim, Adrienne Pena, Olufemi Adelakun, Xi-Tao Wang, Patrick Chow, Samuel J. Bonacorsi, Wendy Hayes. [18F]BMS-986192 as a novel PET imaging agent for assessment of PD-L1 expression in vivo [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 871. doi:10.1158/1538-7445.AM2017-871

Abstract 5391: Low frequency variant detection and tissue-of-origin exploration using liquid biopsies

Abstract The promise of liquid biopsy assays lies in the non-invasive monitoring of diseases, such as cancer, through circulating, cell-free DNA (cfDNA) or circulating tumor cell DNA. This may assist in advancing early-stage diagnosis while simultaneously monitoring treatment response over time. Since these materials are often limited, most liquid biopsy assays incorporate targeted sequencing to enable cost-effective deep coverage of target loci for detection of low frequency pathogenic variants. Yet a critical aspect in attaining the necessary sensitivity is an assay that produces uniform, comprehensive coverage from low DNA input quantities. We have developed a liquid biopsy workflow to enable low frequency variant detection from a 10 mL blood draw using the Promega Maxwell RSC combined with Swift Biosciences Accel-NGS 2S® library preparation methodologies. In addition, we explore whether methylation patterns of the extracted cfDNA possess information of the tissue-of-origin. Whole blood samples were collected in Streck cell-free DNA BCT vials from patients with late stage cancer and cfDNA was extracted with the Promega Maxwell RSC. This instrument yielded DNA outputs ranging from 8-32 ng, with a size profile defined by a predominant peak of ~170bp and a mean Alu repeat qPCR integrity score of 0.22, characteristic of high quality cfDNA lacking cellular DNA content. A total of 20 ng cfDNA was used to make an Accel-NGS 2S Hyb library followed by hybridization capture using Agilent SureSelect Human All Exon probes. The Accel-NGS 2S Hyb Kit exhibits a 90% library conversion rate with cfDNA and provides high complexity libraries with uniform target coverage. In addition, molecular barcodes were incorporated to label each library molecule uniquely prior to PCR amplification. These molecular barcodes were utilized for accurate removal of PCR duplicates while simultaneously preserving naturally occurring fragmentation and strand duplicates to maximize data recovery. Secondly, these barcoded molecules were grouped to generate consensus sequences after removal of false positives originating from PCR and sequencing errors. Variant calling was performed using Vardict and Lofreq enabling highly sensitive and precise detection of variants down to a 0.5% allele frequency. In parallel, we have developed a workflow to determine if the epigenetic status of cfDNA can identify tissue-of-origin. This workflow utilizes the Accel-NGS Methyl-Seq DNA Library Kit to enable unbiased characterization from low (5 ng) cfDNA inputs. Through whole genome bisulfite sequencing, using a priori knowledge of differentially methylated regions characteristic of different human tissues, we can identify the predominant tissue source of cfDNA in blood. Citation Format: Justin S. Lenhart, Ashley Wood, Sukhinder Sandhu, Cassie Schumacher, Laurie Kurihara, Vladimir Makarov, Tim Harkins. Low frequency variant detection and tissue-of-origin exploration using liquid biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5391. doi:10.1158/1538-7445.AM2017-5391

Abstract 2422: Simultaneous detection of activating somatic DNA mutations and expressed fusion transcripts from lung tumor FFPE samples

Abstract Worldwide, lung cancer is the most commonly diagnosed form of cancer with a survival rate among the lowest. Combined, somatic mutations (in the form of SNVs and InDels) and gene fusions, account for the majority of interpretable and actionable genomic alterations. Importantly, this typically requires the analysis of DNA and RNA from limited amounts of FFPE-preserved specimens. Currently, these analyses typically require complex sample pre-processing for assay on separate platforms or separate complex library preparation methods for assessment by high throughput sequencing. To provide a unified and simpler alternative, NanoString’s molecular barcoding technology has been modularized to permit simultaneous digital measurement of cancer-relevant targets that span these two analyte classes. Novel ‘SNV’ probes enable sensitive and specific identification of DNA mutant allele sequences down to a level of detection of ≤ 5% from 5 ng of FFPE-extracted genomic DNA. Fusion transcripts are detected with 5’/3’ imbalance probes and toehold-mediated junction probes. This dual analyte workflow requires just a single 5-10 micron section of FFPE tissue and provides to sample-to-answer results with approximately 5 minutes of hands-on time per sample after nucleic acid extraction. To demonstrate utility, 37 lung cancer samples were assayed simultaneously with an SNV panel that targets &amp;gt;100 solid tumor somatic mutations and a lung cancer fusion gene panel that provides general evidence of ALK, RET, and ROS1 gene fusion events along with specific detection of 35 unique fusion transcripts that correspond to known break-points. In this particular cohort, 16 samples were positive for activating KRAS SNVs (one of which was also positive for an activating STK11 variant), 3 were positive for activating EGFR mutations including two SNVs and an 18-base InDel and one was positive for an activating KIF5B16:RET12 fusion transcript. Positive mutation calls obtained with the SNV panel could only be confirmed by whole-exome sequencing (average depth of 100X) for 13 of 20 variants detected; however, ultra-deep (average depth of 4400X) targeted sequencing revealed that the 7 additional panel-detected mutations were, in fact, present. Measured against the sequencing datasets, the SNV panel provided 100% sensitivity, specificity, accuracy and precision for all variants present at 5% or greater allele frequency. The KIF5B16:RET12 fusion event was also confirmed by sequencing. Combined, these results show that these two important classes of activating mutations can be readily and efficiently assayed together on a NanoString nCounter® system (for research use only). Citation Format: A. McGarry Houghton, Gavin Meredith, Julia Kargl, Jill McKay-Fleisch, P. Martin Ross, Anisha Kharkia, Afshin Mashadi-Hossein, Dae Kim, Joseph Beechem. Simultaneous detection of activating somatic DNA mutations and expressed fusion transcripts from lung tumor FFPE samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2422. doi:10.1158/1538-7445.AM2017-2422

Long-adapter single-strand oligonucleotide probes for the massively multiplexed cloning of kilobase genome regions.

As the catalogue of sequenced genomes and metagenomes continues to grow, massively parallel approaches for the comprehensive and functional analysis of gene products and regulatory elements are becoming increasingly valuable. Current strategies to synthesize or clone complex libraries of DNA sequences are limited by the length of the DNA targets, throughput and cost. Here, we show that long-adapter single-strand oligonucleotide (LASSO) probes can capture and clone thousands of kilobase DNA fragments in a single reaction. As a proof-of-principle, we simultaneously cloned >3,000 bacterial open reading frames (ORFs) from E. coli genomic DNA (spanning 400–5,000 bp targets). Targets were enriched up to a median of ~60-fold compared to non-targeted genomic regions. At a cutoff of 3 times the median non-target reads per kilobase of genetic element per million reads, ~75% of the targeted ORFs were successfully captured. We also show that LASSO probes can clone human ORFs from complementary DNA, and an ORF library from a human-microbiome sample. LASSO probes could be used for the preparation of long-read sequencing libraries and for massively multiplexed cloning.