Discovery-stage identification of drug-like antibodies using emerging experimental and computational methods

Prediction of Self-Association and Solution Behavior of Monoclonal Antibodies Using the QCM-D Metric of Loosely Interacting Layer

Despite the increasing availability and success of monoclonal antibodies (mAb), early identification of candidate molecules with desirable developability attributes remains challenging due to self-association and poor solution behavior. Measuring these phenomena experimentally using the available methods is complicated in mAbs development. Quartz crystal microbalance with dissipation monitoring (QCM-D) detects a loosely interacting layer on top of the irreversibly adsorbed layer of molecules, providing information about the mAbs interaction. This work aimed to explore whether the characteristics of this layer can be used as a reliable self-association metric. QCM-D experiments showed a large frequency shift (Δf) associated with loosely interacting layers for omalizumab but a small or absent layer for tocilizumab. Accordingly, the viscosity of omalizumab increased exponentially at high concentrations compared to tocilizumab. Testing eight mAbs with different self-association behaviors revealed a strong rank order correlation between the mostly used metric of self-association, i.e., diffusion interaction parameter (kD-DLS), and Δf, indicating Δf's potential for predicting mAb solution behavior. The study also highlighted the robustness of the metric to impurities and temperature variations compared to the sensitive kD-DLS. Overall, we demonstrate that the loosely interacting layer provides valuable information about mAb self-association, predicting the colloidal stability and solution behavior in therapeutic development.

Human antibody polyreactivity is governed primarily by the heavy-chain complementarity-determining regions

Although antibody variable regions mediate antigen-specific binding, they can also mediate non-specific interactions with non-cognate antigens, impacting diverse immunological processes and the efficacy, safety, and half-life of antibody therapeutics. To understand the molecular basis of antibody non-specificity, we sorted two dissimilar human naïve antibody libraries against multiple reagents to enrich for variants with different levels of polyreactivity. Sequence analysis of >300,000 paired antibody variable regions revealed that the heavy chain primarily mediates human antibody polyreactivity, and this is due to the high positive charge, high hydrophobicity, and combinations thereof in the corresponding complementarity-determining regions, which can be predicted using a machine learning model developed in this work. Notably, a subset of the most important features governing antibody non-specific interactions, namely those that contain tyrosine, also govern specific antigen recognition. Our findings are broadly relevant for understanding fundamental aspects of antibody molecular recognition and the applied aspects of antibody-drug design.

Accelerating drug discovery, development, and clinical trials by artificial intelligence

Artificial intelligence (AI) has profoundly advanced the field of biomedical research, which also demonstrates transformative capacity for innovation in drug development. This paper aims to deliver a comprehensive analysis of the progress in AI-assisted drug development, particularly focusing on small molecules, RNA, and antibodies. Moreover, this paper elucidates the current integration of AI methodologies within the industrial drug development framework. This encompasses a detailed examination of the industry-standard drug development process, supplemented by a review of medications presently undergoing clinical trials. Conclusively, the paper tackles a predominant obstacle within the AI pharmaceutical sector: the absence of AI-conceived drugs receiving approval. This paper also advocates for the adoption of large language models and diffusion models as a viable strategy to surmount this challenge. This review not only underscores the significant potential of AI in drug discovery but also deliberates on the challenges and prospects within this dynamically progressing field.

Elucidation of the Reversible Self-Association Interface of a Diabody-Interleukin Fusion Protein Using Hydrogen-Exchange Mass Spectrometry and In Silico Modeling

Reversible self-association (RSA) of therapeutic proteins presents major challenges in the development of high-concentration formulations, especially those intended for subcutaneous administration. Understanding self-association mechanisms is therefore critical to the design and selection of candidates with acceptable developability to advance to clinical trials. The combination of experiments and in silico modeling presents a powerful tool to elucidate the interface of self-association. RSA of monoclonal antibodies has been studied extensively under different solution conditions and have been shown to involve interactions for both the antigen-binding fragment and the crystallizable fragment. Novel modalities such as bispecific antibodies, antigen-binding fragments, single-chain-variable fragments, and diabodies constitute a fast-growing class of antibody-based therapeutics that have unique physiochemical properties compared to monoclonal antibodies. In this study, the RSA interface of a diabody-interleukin 22 fusion protein (FP-1) was studied using hydrogen-deuterium exchange coupled with mass spectrometry (HDX-MS) in combination with in silico modeling. Taken together, the results show that a complex solution behavior underlies the self-association of FP-1 and that the interface thereof can be attributed to a specific segment in the variable light chain of the diabody. These findings also demonstrate that the combination of HDX-MS with in silico modeling is a powerful tool to guide the design and candidate selection of novel biotherapeutic modalities.

Cellular Neonatal Fc Receptor Recycling Efficiencies can Differentiate Target-Independent Clearance Mechanisms of Monoclonal Antibodies

In vivo clearance mechanisms of therapeutic monoclonal antibodies (mAbs) encompass both target-mediated and target-independent processes. Two distinct determinants of overall mAb clearance largely separate of target-mediated influences are non-specific cellular endocytosis and subsequent pH-dependent mAb recycling mediated by the neonatal Fc receptor (FcRn), where inter-mAb variability in the efficiency of both processes is observed. Here, we implemented a functional cell-based FcRn recycling assay via Madin-Darby canine kidney type II cells stably co-transfected with human FcRn and its light chain β2-microglobulin. Next, a series of pH-dependent internalization studies using a model antibody demonstrated proper function of the human FcRn complex. We then applied our cellular assays to assess the contribution of both FcRn and non-specific interactions in the cellular turnover for a panel of 8 clinically relevant mAbs exhibiting variable human pharmacokinetic behavior. Our results demonstrate that the interplay of non-specific endocytosis rates, pH-dependent non-specific interactions, and engagement with FcRn all contribute to the overall recycling efficiency of therapeutic monoclonal antibodies. The predictive capacity of our assay approach was highlighted by successful identification of all mAbs within our panel possessing clearance in humans greater than 5 mL/day/kg. These results demonstrate that a combination of cell-based in vitro assays can properly resolve individual mechanisms underlying the overall in vivo recycling efficiency and non-target mediated clearance of therapeutic mAbs.

Status and future developments for downstream processing of biological products: Perspectives from the Recovery XIX yield roundtable discussions

Governments and biopharmaceutical organizations aggressively leveraged expeditious communication capabilities, decision models, and global strategies to make a COVID-19 vaccine happen within a period of 12 months. This was an unusual effort and cannot be transferred to normal times. However, this focus on a single vaccine has also led to other treatments and drug developments being sidelined. Society expects the pharmaceutical industry to provide an uninterrupted supply of medicines. However, it is often overlooked how complex the manufacture of these compounds is and what logistics are required, not to mention the time needed to develop new drugs. The overarching theme, therefore, is patient access and how we can help ensure access and extend it to low- and middle-income countries. Despite unceasing efforts to make medications available to all patient populations, this must never be done at the expense of patient safety. A major fraction of the costs in biopharmaceutical manufacturing are for drug discovery, process development, and clinical studies. Infrastructure costs are very difficult to quantify because they often depend on whether a greenfield facility or an existing, depreciated facility is used or adapted for a new product. To accelerate process development concepts of platform process and prior knowledge are increasingly leveraged. While more traditional protein therapeutics continue to dominate the field, we are also experiencing the exciting emergence and evolution of other therapeutic formats (bispecifics, tetravalent mAbs, antibody-drug conjugates, enzymes, peptides, etc.) that offer unique treatment options for patients. Protein modalities are still dominant, but new modalities are being developed that can be learned from including advanced therapeutics-like cell and gene therapies. The industry must develop a model-based strategy for process development and technologies such as continuous integrated biomanufacturing must be adopted. The overall conclusion is that the pandemic pace was unsustainable, focused on vaccine delivery at the expense of other modalities/disease targets, and had implications for professional and personal life (work-life balance). Routinely reducing development time from 10 years to 1 year is nearly impossible to achieve. Environmental aspects of sustainable downstream processing are also described.

Prospects for the computational humanization of antibodies and nanobodies

To be viable therapeutics, antibodies must be tolerated by the human immune system. Rational approaches to reduce the risk of unwanted immunogenicity involve maximizing the 'humanness' of the candidate drug. However, despite the emergence of new discovery technologies, many of which start from entirely human gene fragments, most antibody therapeutics continue to be derived from non-human sources with concomitant humanization to increase their human compatibility. Early experimental humanization strategies that focus on CDR loop grafting onto human frameworks have been critical to the dominance of this discovery route but do not consider the context of each antibody sequence, impacting their success rate. Other challenges include the simultaneous optimization of other drug-like properties alongside humanness and the humanization of fundamentally non-human modalities such as nanobodies. Significant efforts have been made to develop in silico methodologies able to address these issues, most recently incorporating machine learning techniques. Here, we outline these recent advancements in antibody and nanobody humanization, focusing on computational strategies that make use of the increasing volume of sequence and structural data available and the validation of these tools. We highlight that structural distinctions between antibodies and nanobodies make the application of antibody-focused in silico tools to nanobody humanization non-trivial. Furthermore, we discuss the effects of humanizing mutations on other essential drug-like properties such as binding affinity and developability, and methods that aim to tackle this multi-parameter optimization problem.

The medicinal chemistry evolution of antibody-drug conjugates

Antibody-drug conjugates (ADCs) comprise 3 components of wildly differing sizes: antibody (150 000 Da), linker (typically <500 Da) and payload (typically <500 Da). While the drug-linker makes up only a small percent of the ADC it has a disproportionately massive impact on all aspects of the ADC. Replacing maleimide with bromoacetamide (BrAc) affords stable attachment of the linker to the antibody cysteine, supports total flexibility for linker design and affords a more homogenous ADC. Optimisation of the protease cleavable dipeptide reduces aggregation, facilitates moderation of the physicochemical properties of the ADC and enables long-term stability to facilitate subcutaneous self-administration. Payloads are designed specifically to afford the optimal ADC. Structural information and SAR guide design to improve both potency and selectivity to the small molecule target improving the therapeutic index of resulting ADCs. Minimising the solvent exposed hydrophobic surface area improves the drug-like properties of the ADC, the realisation that the attachment heteroatom can be more than just the site for linker attachment as it can also drive potency and selectivity of the payload and the adoption of a prodrug strategy at project initiation are key areas that medicinal chemistry drives. For an optimal ADC the symbiotic relationship of the three structurally disparate components requires they all function in unison and medicinal chemistry has a huge role to ensure this happens.

ProtT5 and random forests-based viscosity prediction method for therapeutic mAbs

Viscosity is a key characteristic of therapeutic antibodies for subcutaneous administration which requires low volume and high concentration formulations. It would be highly beneficial to accurately predict the viscosity of newly developed therapeutic antibodies in the early stages of development. In this work, a ProtT5-XL-UniRef50 (ProtT5) and Random Forests (RF)-based prediction method was proposed for accurately predicting the viscosity of monoclonal antibodies, with only corresponding sequences needed. Starting from the given heavy and light chain V-region sequences, corresponding features were first extracted from the ProtT5 pretrained model. Kernel principal analysis (Kernel-PCA) was then used for reducing the extracted 2048-D (1024-D for each sequence) feature vector to a reasonable level for efficient training of the RF-regressor. Then, the RF model was constructed on 40 commercially available therapeutic antibodies and tested with 3-folds cross-validation. Test results show that the model could reproduce the viscosity value at a high level (Pearson correlation coefficient (PCC) = 0.928). Performance on classifying high (>30 cP) and low (<30 cP) viscosity is much more satisfactory, the Accuracy (ACC) and the area under precision-recall curve (AUC) of the classification model from validation tests are 0.975 and 1.000, respectively. Compared to 5 existing state-of-the-art viscosity prediction methods, the proposed method performs best which would facilitate high concentration antibody viscosity high-throughput screening.

Engineering hydrophobicity and manufacturability for optimized biparatopic antibody-drug conjugates targeting c-MET

Optimal combinations of paratopes assembled into a biparatopic antibody have the capacity to mediate high-grade target cross-linking on cell membranes, leading to degradation of the target, as well as antibody and payload delivery in the case of an antibody-drug conjugate (ADC). In the work presented here, molecular docking suggested a suitable paratope combination targeting c-MET, but hydrophobic patches in essential binding regions of one moiety necessitated engineering. In addition to rational design of HCDR2 and HCDR3 mutations, site-specific spiking libraries were generated and screened in yeast and mammalian surface display approaches. Comparative analyses revealed similar positions amendable for hydrophobicity reduction, with a broad combinatorial diversity obtained from library outputs. Optimized variants showed high stability, strongly reduced hydrophobicity, retained affinities supporting the desired functionality and enhanced producibility. The resulting biparatopic anti-c-MET ADCs were comparably active on c-MET expressing tumor cell lines as REGN5093 exatecan DAR6 ADC. Structural molecular modeling of paratope combinations for preferential inter-target binding combined with protein engineering for manufacturability yielded deep insights into the capabilities of rational and library approaches. The methodologies of in silico hydrophobicity identification and sequence optimization could serve as a blueprint for rapid development of optimal biparatopic ADCs targeting further tumor-associated antigens in the future.

Reduction of monoclonal antibody viscosity using interpretable machine learning

Early identification of antibody candidates with drug-like properties is essential for simplifying the development of safe and effective antibody therapeutics. For subcutaneous administration, it is important to identify candidates with low self-association to enable their formulation at high concentration while maintaining low viscosity, opalescence, and aggregation. Here, we report an interpretable machine learning model for predicting antibody (IgG1) variants with low viscosity using only the sequences of their variable (Fv) regions. Our model was trained on antibody viscosity data (>100 mg/mL mAb concentration) obtained at a common formulation pH (pH 5.2), and it identifies three key Fv features of antibodies linked to viscosity, namely their isoelectric points, hydrophobic patch sizes, and numbers of negatively charged patches. Of the three features, most predicted antibodies at risk for high viscosity, including antibodies with diverse antibody germlines in our study (79 mAbs) as well as clinical-stage IgG1s (94 mAbs), are those with low Fv isoelectric points (Fv pIs < 6.3). Our model identifies viscous antibodies with relatively high accuracy not only in our training and test sets, but also for previously reported data. Importantly, we show that the interpretable nature of the model enables the design of mutations that significantly reduce antibody viscosity, which we confirmed experimentally. We expect that this approach can be readily integrated into the drug development process to reduce the need for experimental viscosity screening and improve the identification of antibody candidates with drug-like properties.

Optimization of therapeutic antibodies for reduced self-association and non-specific binding via interpretable machine learning

Antibody development, delivery, and efficacy are influenced by antibody-antigen affinity interactions, off-target interactions that reduce antibody bioavailability and pharmacokinetics, and repulsive self-interactions that increase the stability of concentrated antibody formulations and reduce their corresponding viscosity. Yet identifying antibody variants with optimal combinations of these three types of interactions is challenging. Here we show that interpretable machine-learning classifiers, leveraging antibody structural features descriptive of their variable regions and trained on experimental data for a panel of 80 clinical-stage monoclonal antibodies, can identify antibodies with optimal combinations of low off-target binding in a common physiological-solution condition and low self-association in a common antibody-formulation condition. For three clinical-stage antibodies with suboptimal combinations of off-target binding and self-association, the classifiers predicted variable-region mutations that optimized non-affinity interactions while maintaining high-affinity antibody-antigen interactions. Interpretable machine-learning models may facilitate the optimization of antibody candidates for therapeutic applications.

Recent advances in anti-inflammatory active components and action mechanisms of natural medicines

Inflammation is a series of reactions caused by the body's resistance to external biological stimuli. Inflammation affects the occurrence and development of many diseases. Anti-inflammatory drugs have been used widely to treat inflammatory diseases, but long-term use can cause toxic side-effects and affect human functions. As immunomodulators with long-term conditioning effects and no drug residues, natural products are being investigated increasingly for the treatment of inflammatory diseases. In this review, we focus on the inflammatory process and cellular mechanisms in the development of diseases such as inflammatory bowel disease, atherosclerosis, and coronavirus disease-2019. Also, we focus on three signaling pathways (Nuclear factor-kappa B, p38 mitogen-activated protein kinase, Janus kinase/signal transducer and activator of transcription-3) to explain the anti-inflammatory effect of natural products. In addition, we also classified common natural products based on secondary metabolites and explained the association between current bidirectional prediction progress of natural product targets and inflammatory diseases.

Deep learning in preclinical antibody drug discovery and development

Antibody drugs have become a key part of biotherapeutics. Patients suffering from various diseases have benefited from antibody therapies. However, its development process is rather long, expensive and risky. To speed up the process, reduce cost and improve success rate, artificial intelligence, especially deep learning methods, have been widely used in all aspects of preclinical antibody drug development, from library generation to hit identification, developability screening, lead selection and optimization. In this review, we systematically summarize antibody encodings, deep learning architectures and models used in preclinical antibody drug discovery and development. We also critically discuss challenges and opportunities, problems and possible solutions, current applications and future directions of deep learning in antibody drug development.

AI/ML combined with next-generation sequencing of VHH immune repertoires enables the rapid identification of de novo humanized and sequence-optimized single domain antibodies: a prospective case study

Introduction: In this study, we demonstrate the feasibility of yeast surface display (YSD) and nextgeneration sequencing (NGS) in combination with artificial intelligence and machine learning methods (AI/ML) for the identification of de novo humanized single domain antibodies (sdAbs) with favorable early developability profiles. Methods: The display library was derived from a novel approach, in which VHH-based CDR3 regions obtained from a llama (Lama glama), immunized against NKp46, were grafted onto a humanized VHH backbone library that was diversified in CDR1 and CDR2. Following NGS analysis of sequence pools from two rounds of fluorescence-activated cell sorting we focused on four sequence clusters based on NGS frequency and enrichment analysis as well as in silico developability assessment. For each cluster, long short-term memory (LSTM) based deep generative models were trained and used for the in silico sampling of new sequences. Sequences were subjected to sequence- and structure-based in silico developability assessment to select a set of less than 10 sequences per cluster for production. Results: As demonstrated by binding kinetics and early developability assessment, this procedure represents a general strategy for the rapid and efficient design of potent and automatically humanized sdAb hits from screening selections with favorable early developability profiles.

Discovery of ABBV-154, an anti-TNF Glucocorticoid Receptor Modulator Immunology Antibody-Drug Conjugate (iADC)

Stable attachment of drug-linkers to the antibody is a critical requirement, and for maleimide conjugation to cysteine, it is achieved by ring hydrolysis of the succinimide ring. During ADC profiling in our in-house property screening funnel, we discovered that the succinimide ring open form is in equilibrium with the ring closed succinimide. Bromoacetamide (BrAc) was identified as the optimal replacement, as it affords stable attachment of the drug-linker to the antibody while completely removing the undesired ring open-closed equilibrium. Additionally, BrAc also offers multiple benefits over maleimide, especially with respect to homogeneity of the ADC structure. In combination with a short, hydrophilic linker and phosphate prodrug on the payload, this afforded a stable ADC (ABBV-154) with the desired properties to enable long-term stability to facilitate subcutaneous self-administration.

Simplifying complex antibody engineering using machine learning

Machine learning is transforming antibody engineering by enabling the generation of drug-like monoclonal antibodies with unprecedented efficiency. Unsupervised algorithms trained on massive and diverse protein sequence datasets facilitate the prediction of panels of antibody variants with native-like intrinsic properties (e.g., high stability), greatly reducing the amount of subsequent experimentation needed to identify specific candidates that also possess desired extrinsic properties (e.g., high affinity). Additionally, supervised algorithms, which are trained on deep sequencing datasets obtained after enrichment of in vitro antibody libraries for one or more specific extrinsic properties, enable the prediction of antibody variants with desired combinations of extrinsic properties without the need for additional screening. Here we review recent advances using both machine learning approaches and how they are impacting the field of antibody engineering as well as key outstanding challenges and opportunities for these paradigm-changing methods.

Flow cytometric isolation of drug-like conformational antibodies specific for amyloid fibrils

Antibodies that recognize specific protein conformational states are broadly important for research, diagnostic and therapeutic applications, yet they are difficult to generate in a predictable and systematic manner using either immunization or in vitro antibody display methods. This problem is particularly severe for conformational antibodies that recognize insoluble antigens such as amyloid fibrils associated with many neurodegenerative disorders. Here we report a quantitative fluorescence-activated cell sorting (FACS) method for directly selecting high-quality conformational antibodies against different types of insoluble (amyloid fibril) antigens using a single, off-the-shelf human library. Our approach uses quantum dots functionalized with antibodies to capture insoluble antigens, and the resulting quantum dot conjugates are used in a similar manner as conventional soluble antigens for multi-parameter FACS selections. Notably, we find that this approach is robust for isolating high-quality conformational antibodies against tau and α-synuclein fibrils from the same human library with combinations of high affinity, high conformational specificity and, in some cases, low off-target binding that rival or exceed those of clinical-stage antibodies specific for tau (zagotenemab) and α-synuclein (cinpanemab). This approach is expected to enable conformational antibody selection and engineering against diverse types of protein aggregates and other insoluble antigens (e.g., membrane proteins) that are compatible with presentation on the surface of antibody-functionalized quantum dots.

CRISPR/Cas9 Genome Editing for Tissue-Specific In Vivo Targeting: Nanomaterials and Translational Perspective

Clustered randomly interspaced short palindromic repeats (CRISPRs) and its associated endonuclease protein, i.e., Cas9, have been discovered as an immune system in bacteria and archaea; nevertheless, they are now being adopted as mainstream biotechnological/molecular scissors that can modulate ample genetic and nongenetic diseases via insertion/deletion, epigenome editing, messenger RNA editing, CRISPR interference, etc. Many Food and Drug Administration-approved and ongoing clinical trials on CRISPR adopt ex vivo strategies, wherein the gene editing is performed ex vivo, followed by reimplantation to the patients. However, the in vivo delivery of the CRISPR components is still under preclinical surveillance. This review has summarized the nonviral nanodelivery strategies for gene editing using CRISPR/Cas9 and its recent advancements, strategic points of view, challenges, and future aspects for tissue-specific in vivo delivery of CRISPR/Cas9 components using nanomaterials.

Optimization of Drug-Linker to Enable Long-term Storage of Antibody-Drug Conjugate for Subcutaneous Dosing

To facilitate subcutaneous dosing, biotherapeutics need to exhibit properties that enable high-concentration formulation and long-term stability in the formulation buffer. For antibody-drug conjugates (ADCs), the introduction of drug-linkers can lead to increased hydrophobicity and higher levels of aggregation, which are both detrimental to the properties required for subcutaneous dosing. Herein we show how the physicochemical properties of ADCs could be controlled through the drug-linker chemistry in combination with prodrug chemistry of the payload, and how optimization of these combinations could afford ADCs with significantly improved solution stability. Key to achieving this optimization is the use of an accelerated stress test performed in a minimal formulation buffer.

Physiologically Based Pharmacokinetic Modeling to Characterize the Effect of Molecular Charge on Whole-Body Disposition of Monoclonal Antibodies

Motivated by a series of work demonstrating the effect of molecular charge on antibody pharmacokinetics (PK), physiological-based pharmacokinetic (PBPK) models are emerging that relate in silico calculated charge or in vitro measures of polyspecificity to antibody PK parameters. However, only plasma data has been used for model development in these studies, leading to unvalidated assumptions. Here, we present an extended platform PBPK model for antibodies that incorporate charge-dependent endothelial cell pinocytosis rate and nonspecific off-target binding in the interstitial space and on circulating blood cells, to simultaneously characterize whole-body disposition of three antibody charge variants. Predictive potential of various charge metrics was also explored, and the difference between positive charge patches and negative charge patches (i.e., PPC-PNC) was used as the charge parameter to establish quantitative relationships with nonspecific binding affinities and endothelial cell uptake rate. Whole-body disposition of these charge variants was captured well by the model, with less than 2-fold predictive error in area under the curve of most plasma and tissue PK data. The model also predicted that with greater positive charge, nonspecific binding was more substantial, and pinocytosis rate increased especially in brain, heart, kidney, liver, lung, and spleen, but remained unchanged in adipose, bone, muscle, and skin. The presented PBPK model contributes to our understanding of the mechanisms governing the disposition of charged antibodies and can be used as a platform to guide charge engineering based on desired plasma and tissue exposures.

Artificial intelligence in cancer immunotherapy: Applications in neoantigen recognition, antibody design and immunotherapy response prediction

Cancer immunotherapy is a method of controlling and eliminating tumors by reactivating the body's cancer-immunity cycle and restoring its antitumor immune response. The increased availability of data, combined with advancements in high-performance computing and innovative artificial intelligence (AI) technology, has resulted in a rise in the use of AI in oncology research. State-of-the-art AI models for functional classification and prediction in immunotherapy research are increasingly used to support laboratory-based experiments. This review offers a glimpse of the current AI applications in immunotherapy, including neoantigen recognition, antibody design, and prediction of immunotherapy response. Advancing in this direction will result in more robust predictive models for developing better targets, drugs, and treatments, and these advancements will eventually make their way into the clinical setting, pushing AI forward in the field of precision oncology.

Biomaterials-assisted construction of neoantigen vaccines for personalized cancer immunotherapy

INTRODUCTION

Cancer vaccine represents a promising strategy toward personalized immunotherapy, and its therapeutic potency highly relies on the specificity of tumor antigens. Among these extensively studied tumor antigens, neoantigens, a type of short synthetic peptides derived from random somatic mutations, have been shown to be able to elicit tumor-specific antitumor immune response for tumor suppression. However, challenges remain in the efficient and safe delivery of neoantigens to antigen-presenting cells inside lymph nodes for eliciting potent and sustained antitumor immune responses. The rapid advance of biomaterials including various nanomaterials, injectable hydrogels, and macroscopic scaffolds has been found to hold great promises to facilitate the construction of efficient cancer vaccines attributing to their high loading and controllable release capacities.

AREAS COVERED

In this review, we will summarize and discuss the recent advances in the utilization of different types of biomaterials to construct neoantigen-based cancer vaccines, followed by a simple perspective on the future development of such biomaterial-assisted cancer neoantigen vaccination and personalized immunotherapy.

EXPERT OPINION

These latest progresses in biomaterial-assisted cancer vaccinations have shown great promises in boosting substantially potentiated tumor-specific antitumor immunity to suppress tumor growth, thus preventing tumor metastasis and recurrence.

Approaches to expand the conventional toolbox for discovery and selection of antibodies with drug-like physicochemical properties

Antibody drugs should exhibit not only high-binding affinity for their target antigens but also favorable physicochemical drug-like properties. Such drug-like biophysical properties are essential for the successful development of antibody drug products. The traditional approaches used in antibody drug development require significant experimentation to produce, optimize, and characterize many candidates. Therefore, it is attractive to integrate new methods that can optimize the process of selecting antibodies with both desired target-binding and drug-like biophysical properties. Here, we summarize a selection of techniques that can complement the conventional toolbox used to de-risk antibody drug development. These techniques can be integrated at different stages of the antibody development process to reduce the frequency of physicochemical liabilities in antibody libraries during initial discovery and to co-optimize multiple antibody features during early-stage antibody engineering and affinity maturation. Moreover, we highlight biophysical and computational approaches that can be used to predict physical degradation pathways relevant for long-term storage and in-use stability to reduce the need for extensive experimentation.

A library approach for the de novo high-throughput isolation of humanized VHH domains with favorable developability properties following camelid immunization

In this study, we generated a novel library approach for high throughput de novo identification of humanized single-domain antibodies following camelid immunization. To achieve this, VHH-derived complementarity-determining regions-3 (CDR3s) obtained from an immunized llama (Lama glama) were grafted onto humanized VHH backbones comprising moderately sequence-diversified CDR1 and CDR2 regions similar to natural immunized and naïve antibody repertoires. Importantly, these CDRs were tailored toward favorable in silico developability properties, by considering human-likeness as well as excluding potential sequence liabilities and predicted immunogenic motifs. Target-specific humanized single-domain antibodies (sdAbs) were readily obtained by yeast surface display. We demonstrate that, by exploiting this approach, high affinity sdAbs with an optimized in silico developability profile can be generated. These sdAbs display favorable biophysical, biochemical, and functional attributes and do not require any further sequence optimization. This approach is generally applicable to any antigen upon camelid immunization and has the potential to significantly accelerate candidate selection and reduce risks and attrition rates in sdAb development.

Analytical Workflows to Unlock Predictive Power in Biotherapeutic Developability

PURPOSE

Forming accurate data models that assist the design of developability assays is one area that requires a deep and practical understanding of the problem domain. We aim to incorporate expert knowledge into the model building process by creating new metrics from instrument data and by guiding the choice of input parameters and Machine Learning (ML) techniques.

METHODS

We generated datasets from the biophysical characterisation of 5 monoclonal antibodies (mAbs). We explored combinations of techniques and parameters to uncover the ones that better describe specific molecular liabilities, such as conformational and colloidal instability. We also employed ML algorithms to predict metrics from the dataset.

RESULTS

We found that the combination of Differential Scanning Calorimetry (DSC) and Light Scattering thermal ramps enabled us to identify domain-specific aggregation in mAbs that would be otherwise overlooked by common developability workflows. We also found that the response to different salt concentrations provided information about colloidal stability in agreement with charge distribution models. Finally, we predicted DSC transition temperatures from the dataset, and used the order of importance of different metrics to increase the explainability of the model.

CONCLUSIONS

The new analytical workflows enabled a better description of molecular behaviour and uncovered links between structural properties and molecular liabilities. In the future this new understanding will be coupled with ML algorithms to unlock their predictive power during developability assessment.

Antibody apparent solubility prediction from sequence by transfer learning

Developing therapeutic monoclonal antibodies (mAbs) for the subcutaneous administration requires identifying mAbs with superior solubility that are amenable for high-concentration formulation. However, experimental screening is often material and labor intensive. Here, we present a strategy (named solPredict) that employs the embeddings from pretrained protein language modeling to predict the apparent solubility of mAbs in histidine (pH 6.0) buffer. A dataset of 220 diverse, in-house mAbs were used for model training and hyperparameter tuning through 5-fold cross validation. solPredict achieves high correlation with experimental solubility on an independent test set of 40 mAbs. Importantly, solPredict performs well for both IgG1 and IgG4 subclasses despite the distinct solubility behaviors. This approach eliminates the need of 3D structure modeling of mAbs, descriptor computation, and expert-crafted input features. The minimal computational expense of solPredict enables rapid, large-scale, and high-throughput screening of mAbs using sequence information alone during early antibody discovery.

Evaluation of multi-specificity of antibody G2 using its single-chain Fv and its covalently linked antigen peptides

The antibody G2 specifically binds to four peptides with different amino acid sequences: Pep18mer, Pep8, Pep395, and PepH4P6. To elucidate the multi-specificity of G2, we generated a G2 single-chain Fv (scFv) antibody and analyzed its binding thermodynamics and kinetics to antigen peptides. Our results clearly showed that the recognition of PepH4P6 was similar to that of Pep18mer, to which G2 could obtain binding ability through the deletion of Pro95 at light chain on the affinity maturation process. The covalent linking of peptides could increase the thermal stability of G2 scFv due to intramolecular antigen binding. In the effects of respective peptides, the increased thermal stability of G2 scFv linked to Pep8 was significant, possibly due to the rapid dissociation. Binding experiments of G2 scFv linked to peptides to other peptides showed decreased association rates relative to those of antigen-free G2 scFv while the dissociation rates were almost unchanged.

Co-optimization of therapeutic antibody affinity and specificity using machine learning models that generalize to novel mutational space

Therapeutic antibody development requires selection and engineering of molecules with high affinity and other drug-like biophysical properties. Co-optimization of multiple antibody properties remains a difficult and time-consuming process that impedes drug development. Here we evaluate the use of machine learning to simplify antibody co-optimization for a clinical-stage antibody (emibetuzumab) that displays high levels of both on-target (antigen) and off-target (non-specific) binding. We mutate sites in the antibody complementarity-determining regions, sort the antibody libraries for high and low levels of affinity and non-specific binding, and deep sequence the enriched libraries. Interestingly, machine learning models trained on datasets with binary labels enable predictions of continuous metrics that are strongly correlated with antibody affinity and non-specific binding. These models illustrate strong tradeoffs between these two properties, as increases in affinity along the co-optimal (Pareto) frontier require progressive reductions in specificity. Notably, models trained with deep learning features enable prediction of novel antibody mutations that co-optimize affinity and specificity beyond what is possible for the original antibody library. These findings demonstrate the power of machine learning models to greatly expand the exploration of novel antibody sequence space and accelerate the development of highly potent, drug-like antibodies.

Machine learning prediction of antibody aggregation and viscosity for high concentration formulation development of protein therapeutics

Machine learning has been recently used to predict therapeutic antibody aggregation rates and viscosity at high concentrations (150 mg/ml). These works focused on commercially available antibodies, which may have been optimized for stability. In this study, we measured accelerated aggregation rates at 45°C and viscosity at 150 mg/ml for 20 preclinical and clinical-stage antibodies. Features obtained from molecular dynamics simulations of the full-length antibody and sequences were used for machine learning model construction. We found a k-nearest neighbors regression model with two features, spatial positive charge map on the CDRH2 and solvent-accessible surface area of hydrophobic residues on the variable fragment, gives the best performance for predicting antibody aggregation rates (r = 0.89). For the viscosity classification model, the model with the highest accuracy is a logistic regression model with two features, spatial negative charge map on the heavy chain variable region and spatial negative charge map on the light chain variable region. The accuracy and the area under precision recall curve of the classification model from validation tests are 0.86 and 0.70, respectively. In addition, we combined data from another 27 commercial mAbs to develop a viscosity predictive model. The best model is a logistic regression model with two features, number of hydrophobic residues on the light chain variable region and net charges on the light chain variable region. The accuracy and the area under precision recall curve of the classification model are 0.85 and 0.6, respectively. The aggregation rates and viscosity models can be used to predict antibody stability to facilitate pharmaceutical development.

Physiologically Based Modeling to Predict Monoclonal Antibody Pharmacokinetics in Humans from in vitro Physiochemical Properties

A model-based framework is presented to predict monoclonal antibody (mAb) pharmacokinetics (PK) in humans based on in vitro measures of antibody physiochemical properties. A physiologically based pharmacokinetic (PBPK) model is used to explore the predictive potential of 14 in vitro assays designed to measure various antibody physiochemical properties, including nonspecific cell-surface interactions, FcRn binding, thermal stability, hydrophobicity, and self-association. Based on the mean plasma PK time course data of 22 mAbs from humans reported in the literature, we found a significant positive correlation (R = 0.64, p = .0013) between the model parameter representing antibody-specific vascular to endothelial clearance and heparin relative retention time, an in vitro measure of nonspecific binding. We also found that antibody-specific differences in paracellular transport due to convection and diffusion could be partially explained by antibody heparin relative retention time (R = 0.52, p = .012). Other physiochemical properties, including antibody thermal stability, hydrophobicity, cross-interaction and self-association, in and of themselves were not predictive of model-based transport parameters. In contrast to other studies that have reported empirically derived expressions relating in vitro measures of antibody physiochemical properties directly to antibody clearance, the proposed PBPK model-based approach for predicting mAb PK incorporates fundamental mechanisms governing antibody transport and processing, informed by in vitro measures of antibody physiochemical properties, and can be expanded to include more descriptive representations of each of the antibody processing subsystems, as well as other antibody-specific information.

Agonist antibody discovery: Experimental, computational, and rational engineering approaches

Agonist antibodies that activate cellular signaling have emerged as promising therapeutics for treating myriad pathologies. Unfortunately, the discovery of rare antibodies with the desired agonist functions is a major bottleneck during drug development. Nevertheless, there has been important recent progress in discovering and optimizing agonist antibodies against a variety of therapeutic targets that are activated by diverse signaling mechanisms. Herein, we review emerging high-throughput experimental and computational methods for agonist antibody discovery as well as rational molecular engineering methods for optimizing their agonist activity.

Separating clinical antibodies from repertoire antibodies, a path to in silico developability assessment

Approaches for antibody discovery have seen substantial improvement and success in recent years. Yet, advancing antibodies into the clinic remains difficult because therapeutic developability concerns are challenging to predict. We developed a computational model to simplify antibody developability assessment and enable accelerated early-stage screening. To this end, we quantified the ability of hundreds of sequence- and structure-based descriptors to differentiate clinical antibodies that have undergone rigorous screening and characterization for drug-like properties from antibodies in the human repertoire that are not natively paired. This analysis identified 144 descriptors capable of distinguishing clinical from repertoire antibodies. Five descriptors were selected and combined based on performance and orthogonality into a single model referred to as the Therapeutic Antibody Developability Analysis (TA-DA). On a hold-out test set, this tool separated clinical antibodies from repertoire antibodies with an AUC = 0.8, demonstrating the ability to identify developability attributes unique to clinical antibodies. Based on our results, the TA-DA score may serve as an approach for selecting lead antibodies for further development.

Abbreviations: Affinity-Capture Self-Interaction Nanoparticle Spectroscopy (AC-SINS), Area Under the Curve (AUC), Complementary-Determining Region (CDR), Clinical-Stage Therapeutics (CST), Framework (FR), Monoclonal Antibodies (mAbs), Observed Antibody Space (OAS), Receiver Operating Characteristic (ROC), Size-Exclusion Chromatography (SEC), Structural Aggregation Propensity (SAP), Therapeutic Antibody Developability Analysis (TA-DA), Therapeutic Antibody Profiler (TAP), Therapeutic Structural Antibody Database (Thera-SAbDab), Variable Heavy (VH), Variable Light (VL).

Reduction of therapeutic antibody self-association using yeast-display selections and machine learning

Self-association governs the viscosity and solubility of therapeutic antibodies in high-concentration formulations used for subcutaneous delivery, yet it is difficult to reliably identify candidates with low self-association during antibody discovery and early-stage optimization. Here, we report a high-throughput protein engineering method for rapidly identifying antibody candidates with both low self-association and high affinity. We find that conjugating quantum dots to IgGs that strongly self-associate (pH 7.4, PBS), such as lenzilumab and bococizumab, results in immunoconjugates that are highly sensitive for detecting other high self-association antibodies. Moreover, these conjugates can be used to rapidly enrich yeast-displayed bococizumab sub-libraries for variants with low levels of immunoconjugate binding. Deep sequencing and machine learning analysis of the enriched bococizumab libraries, along with similar library analysis for antibody affinity, enabled identification of extremely rare variants with co-optimized levels of low self-association and high affinity. This analysis revealed that co-optimizing bococizumab is difficult because most high-affinity variants possess positively charged variable domains and most low self-association variants possess negatively charged variable domains. Moreover, negatively charged mutations in the heavy chain CDR2 of bococizumab, adjacent to its paratope, were effective at reducing self-association without reducing affinity. Interestingly, most of the bococizumab variants with reduced self-association also displayed improved folding stability and reduced nonspecific binding, revealing that this approach may be particularly useful for identifying antibody candidates with attractive combinations of drug-like properties.Abbreviations: AC-SINS: affinity-capture self-interaction nanoparticle spectroscopy; CDR: complementarity-determining region; CS-SINS: charge-stabilized self-interaction nanoparticle spectroscopy; FACS: fluorescence-activated cell sorting; Fab: fragment antigen binding; Fv: fragment variable; IgG: immunoglobulin; QD: quantum dot; PBS: phosphate-buffered saline; V_H: variable heavy; V_L: variable light.

Improving antibody drug development using bionanotechnology

Monoclonal antibodies are being used to treat a remarkable breadth of human disorders. Nevertheless, there are several key challenges at the earliest stages of antibody drug development that need to be addressed using simple and widely accessible methods, especially related to generating antibodies against membrane proteins and identifying antibody candidates with drug-like biophysical properties (high solubility and low viscosity). Here we highlight key bionanotechnologies for preparing functional and stable membrane proteins in diverse types of lipoparticles that are being used to improve antibody discovery and engineering efforts. We also highlight key bionanotechnologies for high-throughput and ultra-dilute screening of antibody biophysical properties during antibody discovery and optimization that are being used for identifying antibodies with superior combinations of in vitro (formulation) and in vivo (half-life) properties.