The constant regions of clinical monoclonal antibodies are derived from a select number of allotypes found in IgG subclasses. Despite a long-term acknowledgment that this diversity may impact both antibody function and developability, there is a lack of data on the stability of variants carrying these mutations. Here, we generated a panel of IgG1, IgG2, and IgG3 antibodies with 32 unique constant region alleles and performed a systematic comparison of stability using red edge excitation shift (REES). This technique exploits the fluorescent properties of tryptophan residues to measure antibody structural dynamics which predict flexibility and the propensity to unfold. Our REES measurements revealed broad stability differences between subclasses with IgG3 possessing the poorest overall stability. Further interrogation of differences between variants within each subclass enabled the high-resolution profiling of individual allotype stabilities. Crucially, these observed differences were not found to be linked to N297-linked glycan heterogeneity. Our work demonstrates diverse stabilities (and dynamics) for a range of naturally occurring constant domain alleles and the utility of REES as a method for rapid and sensitive antibody stability profiling, requiring only laboratory spectrophotometry equipment. This article is protected by copyright. All rights reserved.

We describe the Predicting Protein Compound Interactions (PrePCI) database which comprises over 5 billion predicted interactions between nearly 7 million chemical compounds and 19,797 human proteins. PrePCI relies on a proteome-wide database of structural models based on both traditional modeling techniques and the AlphaFold Protein Structure Database. Sequence and structural similarity-based metrics are established between template proteins, T, in the Protein Data Bank that bind compounds, C, and query proteins in the model database, Q. When these metrics pass a sequence threshold value, it is assumed that C also binds to Q with a likelihood ratio derived from machine learning. If the relationship is based on structure, this LR is based on a scoring function that measures the extent to which C is compatible with the binding site of Q as described in the LT-scanner algorithm. For every predicted complex derived in this way, chemical similarity based on the Tanimoto coefficient identifies other small molecules that may bind to Q. A likelihood ratio for the binding of C to Q is obtained from Naïve Bayesian statistics. The PrePCI database can be queried by entering a UniProt ID or gene name for a protein to obtain a list of compounds predicted to bind to it along with associated LRs. Alternatively, entering an identifier for the compound outputs a list of proteins it is predicted to bind. Specific applications of the database to lead discovery, drug mechanism of action elucidation and biological function annotation are described. This article is protected by copyright. All rights reserved.

Though many folded proteins assume one stable structure that performs one function, a small-but-increasing number remodel their secondary and tertiary structures and change their functions in response to cellular stimuli. These fold-switching proteins regulate biological processes and are associated with autoimmune dysfunction, SARS-CoV-2 infection, and more. Despite their biological importance, it is difficult to computationally predict fold switching. With the aim of advancing computational prediction and experimental characterization of fold switchers, this review discusses several features that distinguish fold-switching proteins from their single-fold and intrinsically disordered counterparts. Firstly, the isolated structures of fold switchers are less stable and more heterogeneous than single folders but more stable and less heterogeneous than intrinsically disordered proteins (IDPs). Secondly, the sequences of single-fold, fold-switching, and intrinsically disordered proteins can evolve at distinct rates. Thirdly, proteins from these three classes are best predicted using different computational techniques. Finally, late-breaking results suggest that single folders, fold switchers, and IDPs have distinct patterns of residue-residue coevolution. The review closes by discussing high and medium throughput experimental approaches that might be used to identify new fold-switching proteins. This article is protected by copyright. All rights reserved.

Herein, we describe the generation of potent NK cell engagers (NKCEs) based on single domain antibodies (sdAbs) specific for NKp46 harboring the humanized Fab version of Cetuximab for tumor targeting. After immunization of camelids, a plethora of different VHH domains were retrieved by yeast surface display. Upon reformatting into Fc effector-silenced NKCEs targeting NKp46 and EGFR in a strictly monovalent fashion, the resulting bispecific antibodies elicited potent NK cell-mediated killing of EGFR-overexpressing tumor cells with potencies (EC50 killing) in the picomolar range. This was further augmented via co-engagement of Fcγ receptor IIIa (FcγRIIIa). Importantly, NKp46-specific sdAbs enabled the construction of various NKCE formats with different geometries and valencies which displayed favorable biophysical and biochemical properties without further optimization. By this means, killing capacities were further improved significantly. Hence, NKp46-specific sdAbs are versatile building blocks for the construction of different NKCE formats. This article is protected by copyright. All rights reserved.

Outer membrane protein (OMP) biogenesis in gram-negative bacteria is managed by a network of periplasmic chaperones that includes SurA, Skp, and FkpA. These chaperones bind unfolded OMPs (uOMPs) in dynamic conformational ensembles to suppress aggregation, facilitate diffusion across the periplasm, and enhance folding. FkpA primarily responds to heat-shock stress, but its mechanism is comparatively understudied. To determine FkpA chaperone function in the context of OMP folding, we monitored the folding of three OMPs and found that FkpA, unlike other periplasmic chaperones, increases the folded yield but decreases the folding rate of OMPs. The results indicate that FkpA behaves as a chaperone and not as a folding catalyst to influence the OMP folding trajectory. Consistent with the folding assay results, FkpA binds all three uOMPs as determined by sedimentation velocity (SV) and photo-crosslinking experiments. We determine the binding affinity between FkpA and uOmpA171 by globally fitting sedimentation velocity titrations and find it to be intermediate between the known affinities of Skp and SurA for uOMP clients. Notably, complex formation steeply depends on the urea concentration, suggesting an extensive binding interface. Initial characterizations of the complex using photo-crosslinking indicate that the binding interface spans the entire FkpA molecule. In contrast to prior findings, folding and binding experiments performed using subdomain constructs of FkpA demonstrate that the full-length chaperone is required for full activity. Together these results support that FkpA has a distinct and direct effect on OMP folding that it achieves by utilizing an extensive chaperone-client interface to tightly bind clients. This article is protected by copyright. All rights reserved.

To advance our ability to predict impacts of the protein scaffold on catalysis, robust classification schemes to define features of proteins that will influence reactivity are needed. One of these features is a protein's metal-binding ability, as metals are critical to catalytic conversion by metalloenzymes. As a step toward realizing this goal, we used convolutional neural networks (CNN)'s to enable the classification of a metal cofactor binding pocket within a protein scaffold. Convolutional neural networks enable images to be classified based on multiple levels of detail in the image, from edges and corners to entire objects, and can provide rapid classification. First, six CNN models were fine-tuned to classify the 20 standard amino acids to choose a performant model for amino acid classification. This model was then trained in two parallel efforts: to classify a 2D image of the environment within a given radius of the central metal binding site, either an Fe ion or a [2Fe-2S] cofactor, with the metal visible (effort 1) or the metal hidden (effort 2). We further used two sub-classifications of the [2Fe-2S] cofactor: 1) a standard [2Fe-2S] cofactor and 2) a Rieske [2Fe-2S] cofactor. The accuracy for the model correctly identifying all three defined features was >95%, despite our perception of the increased challenge of the metalloenzyme identification. This demonstrates that machine learning methodology to classify and distinguish similar metal-binding sites, even in the absence of a visible cofactor, is indeed possible and offers an additional tool for metal-binding site identification in proteins. This article is protected by copyright. All rights reserved.

Photoreceptors containing the light-oxygen-voltage (LOV) domain elicit biological responses upon excitation of their flavin mononucleotide (FMN) chromophore by blue light. The mechanism and kinetics of dark-state recovery are not well understood. Here we incorporated the non-canonical amino acid p-cyanophenylalanine (CNF) by genetic code expansion technology at forty-five positions of the bacterial transcription factor EL222. Screening of light-induced changes in infrared (IR) absorption frequency, electric field and hydration of the nitrile groups identified residues CNF31 and CNF35 as reporters of monomer/oligomer and caged/decaged equilibria, respectively. Time-resolved multi-probe UV/Visible and IR spectroscopy experiments of the lit-to-dark transition revealed four dynamical events. Predominantly, rearrangements around the A'α helix interface (CNF31 and CNF35) precede FMN-cysteinyl adduct scission, folding of α-helices (amide bands), and relaxation of residue CNF151. This study illustrates the importance of characterizing all parts of a protein and suggests a key role for the N-terminal A'α extension of the LOV domain in controlling EL222 photocycle length. This article is protected by copyright. All rights reserved.