Biology subjects

Hotzel, I.

Publications and source records attributed to Hotzel, I..

9 recordsLinked to original sources

bioRxiv2025

Property Enhancer - a data efficient multi-objective approach for functional antibody optimization

In-silico antibody lead optimization remains challenging due to scarce high-quality data, costly experimental validation, and the need to jointly optimize multiple developability properties. Discovery workflows often rely on high-throughput phage, ribosome or yeast display experiments, which yield large but noisy datasets; as leads emerge, strategies shift to low-throughput assays which are precise, yet unscalable. Deep-learning and language-model approaches are hindered by such limited, unreliable measurements. We introduce Property Enhancer (PropEn), a data-efficient framework for low-data, heterogeneous regimes that can simultaneously optimize multiple antibody properties. PropEn proposes a matching-based augmentation that expands the training data with sequence pairs differing by only a few mutations; within each pair the second sequence improves the target value, providing an implicit optimization signal. Extensive in silico and in vitro tests show 10-39x affinity gains across four targets and nine leads, and enable joint multi-property optimization, positioning PropEn as a scalable, general solution.

molecular biology↗

bioRxiv2025

Lab-in-the-loop therapeutic antibody design with deep learning

Therapeutic antibody design is a complex multi-property optimization problem with substantial promise for improvement with the application of machine-learning methods. Towards realizing that promise, we introduce "Lab-in-the-loop," a new approach that orchestrates state-of-the-art repertoire mining methods, generative machine learning models, multi-task property predictors, active learning ranking and selection, and in vitro experimentation in a semi-autonomous, iterative optimization loop. By automating the design of antibody variants, property prediction, ranking and selection of designs to assay in the lab, and ingestion of in vitro data, we enable an end-to-end approach to developing computationally-informed therapeutic antibody design pipelines. We apply lab-in-the-loop to eleven seed antibodies obtained via animal immunization with four clinically relevant antigen targets: EGFR, IL-6, HER2, and OSM. Over 1,800 unique antibody variants are tested throughout four rounds of iterative optimization identifying 3-100x better binding variants for all targets and 10/11 seeds, with the best binders exceeding 100 pM affinity, demonstrating a process by which end-to-end machine learning can be developed for therapeutic antibody development.

bioengineering↗

bioRxiv2024

Transfer and Capture of Envelope Protein Receptor-Binding Domains in the Retroviruses

The surface subunit (SU) of the envelope protein (Env) or retroviruses is highly variable due to adaptation to different hosts during their long evolutionary history. Several exogenous and endogenous retroviral gamma-like Env have a receptor-binding domain (RBD) in the amino-terminal region of SU that folds independently from the carboxyterminal SU C-domain. Two structurally distinct RBD classes have been described, one that adopts a modified immunoglobulin (Ig)-like domain fold and a second, in the Env of RDR interference group retroviruses, with a distinct {beta}-sheet fold. Here, the distribution of different RBD classes among exogenous and endogenous gammatype Env was determined by phylogenetic analyses of Env and structural modeling of retroviral SU with AlphaFold2. The patterns of RBD distribution indicate multiple RBD transfer events in the retroviruses. In addition, SU structural modeling identified an endogenous alpharetroviral-like Env in mammalian species with an amino-terminal RBD. This RDB has a typical IgV domain fold closely related both structurally and in sequence to the mammalian signal regulatory proteins (SIRP) , {beta} and {gamma}, indicating, for the first time, acquisition of an RBD from a mammalian gene by the retroviruses. The results described herein indicate intragenic env recombination between retroviruses and between retroviruses and their hosts as a major factor in the evolution of retroviral Env.

evolutionary biology↗

bioRxiv2023

Structural Evolution of Trimeric Gamma-Type Retroviral Envelope Proteins

The envelope glycoproteins (Env) of retroviruses are highly diverse due to their relatively rapid evolution and the ancient origin of retroviruses. Sequence similarity among the Env of retroviruses is observed in the transmembrane subunit (TM) of Env mediating membrane fusion, whereas this similarity is more limited or absent in the surface subunit (SU) mediating receptor binding. Recent structural modeling of the SU of major retroviral groups identified a structurally conserved {beta}-sheet domain in SU and the SU-equivalent of filoviruses, GP1, that is differentially expanded to generate the structural diversity observed in this protein family. Here, trimeric Env structures from three major exogenous and endogenous gamma-type retroviral Env classes were modeled with AlphaFold2 multimer to address the evolution of trimeric Env in the retroviruses. The models show a close structural relationship between the trimeric Env proteins of gammaretroviruses and alpharetroviruses and the trimeric filoviral spike protein, with an extended hydrophobic structure anchoring SU into TM in retroviral Env occluding the intersubunit disulfide bond. In addition, the murine leukemia virus (MLV) Env trimer model indicates that the type-C gammaretroviral receptor binding domains (RBD) are positioned on the side of the Env trimer pointing towards the viral envelope membrane, unlike those of the gammaretroviral RDR interference group, which are positioned closer to the apical end of the Env trimer. The trimeric MLV Env structural model, including RBD position, is consistent with previous trimeric MLV Env electron-density maps. The structural models provide insights into the structure, function and evolution of trimeric retroviral gamma-type Env.

microbiology↗

bioRxiv2022

Loss of the intracellular enzyme QPCTL limits chemokine function and reshapes myeloid infiltration to augment tumor immunity

Tumor-associated macrophages are composed of distinct populations arising from monocytes or tissue macrophages, with a poorly understood link to disease pathogenesis. Here, we demonstrate that mouse monocyte migration was supported by glutaminyl-peptide cyclotransferase-like (QPCTL), an enzyme that mediates N-terminal modification of several subtrates, including the monocyte-chemoattractants CCL2 and CCL7, protecting them from proteolytic inactivation. Knockout of Qpctl disrupted monocyte homeostasis, attenuated tumor growth and reshaped myeloid cell infiltration, with loss of monocyte-derived populations with immunosuppressive and pro-angiogenic profiles. Antibody blockade of the receptor CSF1R, which more broadly eliminates tissue macrophages, reversed tumor growth inhibition in Qpctl-/- mice, and prevented lymphocyte infiltration. Modulation of QPCTL synergized with anti-PD-L1 to expand CD8+ T cells and limit tumor growth. QPCTL inhibition constitutes an effective approach for myeloid cell-targeted cancer immunotherapy.

immunology↗

bioRxiv2021

Deep time structural evolution of retroviral and filoviral surface envelope proteins

The retroviral surface envelope protein subunit (SU) mediates receptor binding and triggers membrane fusion by the transmembrane subunit (TM). SU evolves rapidly under strong selective conditions, resulting in seemingly unrelated SU structures in highly divergent retroviruses. Structural modeling of the SU of several retroviruses and related endogenous retroviral elements with AlphaFold identifies a TM-proximal SU {beta}-sandwich structure that has been conserved in the orthoretroviruses for at least 110 million years. The SU of orthoretroviruses diversified by differential expansion of the {beta} -sandwich core to form domains involved in virus-host interactions. The {beta}-sandwich domain is also conserved in the SU equivalent GP1 of Ebola virus although with a significantly different orientation in the trimeric envelope protein structure. The unified structural view of orthoretroviral SU and filoviral GP1 identifies an ancient, structurally conserved and evolvable domain underlying the structural diversity of orthoretroviral SU and filoviral GP1.

microbiology↗

bioRxiv2021

Domain Organization of Lentiviral and Betaretroviral Surface Envelope Glycoproteins Modeled with AlphaFold

The surface envelope glycoproteins of non-primate lentiviruses and betaretroviruses share sequence similarity with the inner proximal domain {beta}-sandwich of the human immunodeficiency virus type 1 (HIV-1) gp120 glycoprotein that faces the transmembrane glycoprotein as well as patterns of cysteine and glycosylation site distribution that points to a similar two-domain organization in at least some lentiviruses. Here, high reliability models of the surface glycoproteins obtained with the AlphaFold algorithm are presented for the gp135 glycoprotein of the small ruminant caprine arthritis-encephalitis (CAEV) and visna lentiviruses and the betaretroviruses jaagsiekte sheep retrovirus (JSRV), mouse mammary tumor virus (MMTV) and consensus human endogenous retrovirus type K (HERV-K). The models confirm and extend the inner domain structural conservation in these viruses and identify two outer domains with a putative receptor binding site in the CAEV and visna virus gp135. The location of that site is consistent with patterns of sequence conservation and glycosylation site distribution in gp135. In contrast, a single domain is modeled for the JSRV, MMTV and HERV-K betaretrovirus envelope proteins that is highly conserved structurally in the proximal region and structurally diverse in apical regions likely to interact with cell receptors. The models presented here identify sites in small ruminant lentivirus and betaretrovirus envelope glycoproteins likely to be critical for virus entry and virus neutralization by antibodies and will facilitate their functional and structural characterization. ImportanceStructural information on the surface envelope proteins of lentiviruses and related betaretroviruses is critical to understand mechanisms of virus-host interactions. However, experimental determination of these structures has been challenging and only the structure of the human immunodeficiency virus type 1 gp120 has been determined. The advent of the AlphaFold artificial intelligence method for structure prediction allows high-quality modeling of the structures of small ruminant lentiviral and betaretroviral surface envelope proteins. The models are consistent with much of previously described experimental data, show regions likely to interact with receptors and identify domains that may be involved in mechanisms of antibody neutralization resistance in the small ruminant lentiviruses. The models will allow more precise design of mutants to further determine mechanisms of viral entry and immune evasion in this group of viruses and constructs for structure of these surface envelope proteins.

microbiology↗

bioRxiv2019

Heavy Chain CDR 3 and Junctional Length Biases in Human Antibody Repertoires Associated with Heavy and Light Chain Germline Utilization

Antibody variable domain sequence diversity is generated by recombination of germline segments. The third complementarity-determining region of the heavy chain (CDR H3) is the region of highest sequence diversity and is formed by the joining of heavy chain VH, DH and JH germline segments combined with random nucleotide trimming and additions between these segments. We show that CDR H3 length distribution is biased in human antibody repertoires as a function of VH, VL and JH germline segment utilization. Most length biases are apparent in the naive B cell compartment, with a significant bias towards shorter CDR H3 sequences observed in association with a subset of VH and VL germlines in the antigen experienced compartment. Similar biases were not observed in nonproductive heavy chain recombination products, indicating selection of the repertoire during B cell maturation as a major driver of the length biases. Some VH-associated CDR H3 length biases are dependent on utilization of specific JH germline segments in a manner not directly linked to JH segment length in the germline, but are rather associated with selection of differentially trimmed JH segments in the naive compartment. In addition, DH segment and N-region random nucleotide insertion lengths within CDR H3 in the naive compartment were also biased by specific VH/JH germline combinations, indicating a complex set of constraints between germline segments selected during repertoire maturation. Our findings reveal biases in the antibody diversity landscape shaped by VH, VL, and JH germline features with implications for mechanisms of naive and immune repertoire selection.

immunology↗

bioRxiv2019

Massively parallel single-cell B-cell receptor sequencing enables rapid discovery of diverse antigen-reactive antibodies

Using high-throughput single-cell B-cell receptor sequencing (scBCR-seq) we obtained accurately paired full-length heavy- and light-chain variable domains from thousands of individual B cells in a massively parallel fashion. We sequenced more than 250,000 B cells from rat, mouse and human repertoires to characterize their lineages and expansion. In addition, we immunized rats with chicken ovalbumin and profiled antigen-reactive B cells from lymph nodes of immunized animals. The scBCR-seq data recovered 81% (n = 56/69) of B-cell lineages identified from hybridomas generated from the same set of B cells that were subjected to scBCR-seq. Importantly, scBCR-seq identified an additional 710 candidate lineages that were not recovered as hybridomas. We synthesized, expressed and tested 93 clones from the identified lineages and found that 99% (n = 92/93) of the clones were antigen-reactive. Our results establish scBCR-seq as a powerful tool for antibody discovery.

immunology↗