Search

Highly selective C-H functionalization remains an ongoing challenge in organic synthetic methodologies. Biocatalysts are robust tools for achieving these difficult chemical transformations. Biocatalyst engineering has often required directed evolution or structure-based rational design campaigns to improve their activities. In recent years, machine learning has been integrated into these workflows to improve the discovery of beneficial enzyme variants. In this work, we combine a structure-based machine-learning algorithm with classical molecular dynamics simulations to down select mutations for rational design of a non-heme iron-dependent lysine dioxygenase, LDO. This approach consistently resulted in functional LDO mutants and circumvents the need for extensive study of mutational activity before-hand. Our rationally designed single mutants purified with up to 2-fold higher yields than WT and displayed higher total turnover numbers (TTN). Combining five such single mutations into a pentamutant variant, LPNYI LDO, leads to a 40% improvement in the TTN (218±3) as compared to WT LDO (TTN = 160±2). Overall, this work offers a low-barrier approach for those seeking to synergize machine learning algorithms with pre-existing protein engineering strategies., (© 2024 Wiley‐VCH GmbH.)

Engineering stabilized proteins is a fundamental challenge in the development of industrial and pharmaceutical biotechnologies. We present Stability Oracle: a structure-based graph-transformer framework that achieves SOTA performance on accurately identifying thermodynamically stabilizing mutations. Our framework introduces several innovations to overcome well-known challenges in data scarcity and bias, generalization, and computation time, such as: Thermodynamic Permutations for data augmentation, structural amino acid embeddings to model a mutation with a single structure, a protein structure-specific attention-bias mechanism that makes transformers a viable alternative to graph neural networks. We provide training/test splits that mitigate data leakage and ensure proper model evaluation. Furthermore, to examine our data engineering contributions, we fine-tune ESM2 representations (Prostata-IFML) and achieve SOTA for sequence-based models. Notably, Stability Oracle outperforms Prostata-IFML even though it was pretrained on 2000X less proteins and has 548X less parameters. Our framework establishes a path for fine-tuning structure-based transformers to virtually any phenotype, a necessary task for accelerating the development of protein-based biotechnologies., (© 2024. The Author(s).)

Photoenzymatic intermolecular hydroalkylations of olefins are highly enantioselective for chiral centers formed during radical termination but poorly selective for centers set in the C-C bond-forming event. Here, we report the evolution of a flavin-dependent "ene"-reductase to catalyze the coupling of α,α-dichloroamides with alkenes to afford α-chloroamides in good yield with excellent chemo- and stereoselectivity. These products can serve as linchpins in the synthesis of pharmaceutically valuable motifs. Mechanistic studies indicate that radical formation occurs by exciting a charge-transfer complex templated by the protein. Precise control over the orientation of molecules within the charge-transfer complex potentially accounts for the observed stereoselectivity. The work expands the types of motifs that can be prepared using photoenzymatic catalysis.

A major challenge to achieving industry-scale biomanufacturing of therapeutic alkaloids is the slow process of biocatalyst engineering. Amaryllidaceae alkaloids, such as the Alzheimer's medication galantamine, are complex plant secondary metabolites with recognized therapeutic value. Due to their difficult synthesis they are regularly sourced by extraction and purification from the low-yielding daffodil Narcissus pseudonarcissus. Here, we propose an efficient biosensor-machine learning technology stack for biocatalyst development, which we apply to engineer an Amaryllidaceae enzyme in Escherichia coli. Directed evolution is used to develop a highly sensitive (EC 50  = 20 μM) and specific biosensor for the key Amaryllidaceae alkaloid branchpoint 4'-O-methylnorbelladine. A structure-based residual neural network (MutComputeX) is subsequently developed and used to generate activity-enriched variants of a plant methyltransferase, which are rapidly screened with the biosensor. Functional enzyme variants are identified that yield a 60% improvement in product titer, 2-fold higher catalytic activity, and 3-fold lower off-product regioisomer formation. A solved crystal structure elucidates the mechanism behind key beneficial mutations., (© 2024. The Author(s).)

Deep learning models are seeing increased use as methods to predict mutational effects or allowed mutations in proteins. The models commonly used for these purposes include large language models (LLMs) and 3D Convolutional Neural Networks (CNNs). These two model types have very different architectures and are commonly trained on different representations of proteins. LLMs make use of the transformer architecture and are trained purely on protein sequences whereas 3D CNNs are trained on voxelized representations of local protein structure. While comparable overall prediction accuracies have been reported for both types of models, it is not known to what extent these models make comparable specific predictions and/or generalize protein biochemistry in similar ways. Here, we perform a systematic comparison of two LLMs and two structure-based models (CNNs) and show that the different model types have distinct strengths and weaknesses. The overall prediction accuracies are largely uncorrelated between the sequence- and structure-based models. Overall, the two structure-based models are better at predicting buried aliphatic and hydrophobic residues whereas the two LLMs are better at predicting solvent-exposed polar and charged amino acids. Finally, we find that a combined model that takes the individual model predictions as input can leverage these individual model strengths and results in significantly improved overall prediction accuracy., (© 2023. Springer Nature Limited.)

Machine and deep learning approaches can leverage the increasingly available massive datasets of protein sequences, structures, and mutational effects to predict variants with improved fitness. Many different approaches are being developed, but systematic benchmarking studies indicate that even though the specifics of the machine learning algorithms matter, the more important constraint comes from the data availability and quality utilized during training. In cases where little experimental data are available, unsupervised and self-supervised pre-training with generic protein datasets can still perform well after subsequent refinement via hybrid or transfer learning approaches. Overall, recent progress in this field has been staggering, and machine learning approaches will likely play a major role in future breakthroughs in protein biochemistry and engineering., Competing Interests: Declaration of competing interest D.J.D. is a cofounder of Intelligent Proteins, LLC, which uses machine learning for problems of protein engineering. A.V.K., A.D.E., and C·O.W. declare that they have no conflict of interest., (Copyright © 2022 Elsevier Ltd. All rights reserved.)

The DNA polymerase I from Geobacillus stearothermophilus (also known as Bst DNAP) is widely used in isothermal amplification reactions, where its strand displacement ability is prized. More robust versions of this enzyme should be enabled for diagnostic applications, especially for carrying out higher temperature reactions that might proceed more quickly. To this end, we appended a short fusion domain from the actin-binding protein villin that improved both stability and purification of the enzyme. In parallel, we have developed a machine learning algorithm that assesses the relative fit of individual amino acids to their chemical microenvironments at any position in a protein and applied this algorithm to predict sequence substitutions in Bst DNAP. The top predicted variants had greatly improved thermotolerance (heating prior to assay), and upon combination, the mutations showed additive thermostability, with denaturation temperatures up to 2.5 °C higher than the parental enzyme. The increased thermostability of the enzyme allowed faster loop-mediated isothermal amplification assays to be carried out at 73 °C, where both Bst DNAP and its improved commercial counterpart Bst 2.0 are inactivated. Overall, this is one of the first examples of the application of machine learning approaches to the thermostabilization of an enzyme.

Genetic biosensors are integral to synthetic biology. In particular, ligand-inducible prokaryotic transcription factors are frequently used in high-throughput screening, for dynamic feedback regulation, as multilayer logic gates, and in diagnostic applications. In order to provide a curated source that users can rely on for engineering applications, we have developed GroovDB (available at https://groov.bio), a Web-accessible database of ligand-inducible transcription factors that contains all information necessary to build chemically responsive genetic circuits, including biosensor sequence, ligand, and operator data. Ligand and DNA interaction data have been verified against the literature, while an automated data curation pipeline is used to programmatically fetch metadata, structural information, and references for every entry. A custom tool to visualize the natural genetic context of biosensor entries provides potential insights into alternative ligands and systems biology.

Background: Women with a cervicovaginal microbiota dominated by Lactobacillus spp. are at reduced risk of acquiring sexually transmitted infections including HIV, but the biological mechanisms involved remain poorly defined. Here, we performed metaproteomics on vaginal swab samples from young South African women (n = 113) and transcriptomics analysis of cervicovaginal epithelial cell cultures to examine the ability of lactic acid, a metabolite produced by cervicovaginal lactobacilli, to modulate genital epithelial barrier function., Results: Compared to women with Lactobacillus-depleted microbiota, women dominated by vaginal lactobacilli exhibit higher abundance of bacterial lactate dehydrogenase, a key enzyme responsible for lactic acid production, which is independently associated with an increased abundance of epithelial barrier proteins. Physiological concentrations of lactic acid enhance epithelial cell culture barrier integrity and increase intercellular junctional molecule expression., Conclusions: These findings reveal a novel ability of vaginal lactic acid to enhance genital epithelial barrier integrity that may help prevent invasion by sexually transmitted pathogens. Video abstract., (© 2022. The Author(s).)

The COVID-19 pandemic has resulted in an unprecedented global demand for in vitro diagnostic reagents. Supply shortages and hoarding have impacted testing capacity which has led to inefficient COVID-19 case identification and transmission control, predominantly in developing countries. Traditionally, RNA extraction is a prerequisite for conducting SARS-CoV-2 nucleic acid amplification tests (NAAT); however, simplified methods of sample processing have been successful at bypassing typical nucleic acid extraction steps, enabling extraction-free SARS-CoV-2 NAAT workflows. These methods involve chemical and physical approaches that are inexpensive and easily accessible alternatives to overcome extraction kit supply shortages, while offering acceptable test performance. Here we provide an overview of three main sample preparation strategies that have been shown to facilitate extraction-free SARS-CoV-2 NAATs.

Plastic waste poses an ecological challenge 1-3 and enzymatic degradation offers one, potentially green and scalable, route for polyesters waste recycling 4 . Poly(ethylene terephthalate) (PET) accounts for 12% of global solid waste 5 , and a circular carbon economy for PET is theoretically attainable through rapid enzymatic depolymerization followed by repolymerization or conversion/valorization into other products 6-10 . Application of PET hydrolases, however, has been hampered by their lack of robustness to pH and temperature ranges, slow reaction rates and inability to directly use untreated postconsumer plastics 11 . Here, we use a structure-based, machine learning algorithm to engineer a robust and active PET hydrolase. Our mutant and scaffold combination (FAST-PETase: functional, active, stable and tolerant PETase) contains five mutations compared to wild-type PETase (N233K/R224Q/S121E from prediction and D186H/R280A from scaffold) and shows superior PET-hydrolytic activity relative to both wild-type and engineered alternatives 12 between 30 and 50 °C and a range of pH levels. We demonstrate that untreated, postconsumer-PET from 51 different thermoformed products can all be almost completely degraded by FAST-PETase in 1 week. FAST-PETase can also depolymerize untreated, amorphous portions of a commercial water bottle and an entire thermally pretreated water bottle at 50 ºC. Finally, we demonstrate a closed-loop PET recycling process by using FAST-PETase and resynthesizing PET from the recovered monomers. Collectively, our results demonstrate a viable route for enzymatic plastic recycling at the industrial scale., (© 2022. The Author(s), under exclusive licence to Springer Nature Limited.)

One fundamental problem of protein biochemistry is to predict protein structure from amino acid sequence. The inverse problem, predicting either entire sequences or individual mutations that are consistent with a given protein structure, has received much less attention even though it has important applications in both protein engineering and evolutionary biology. Here, we ask whether 3D convolutional neural networks (3D CNNs) can learn the local fitness landscape of protein structure to reliably predict either the wild-type amino acid or the consensus in a multiple sequence alignment from the local structural context surrounding site of interest. We find that the network can predict wild type with good accuracy, and that network confidence is a reliable measure of whether a given prediction is likely going to be correct or not. Predictions of consensus are less accurate and are primarily driven by whether or not the consensus matches the wild type. Our work suggests that high-confidence mis-predictions of the wild type may identify sites that are primed for mutation and likely targets for protein engineering., (© 2021. The Author(s), under exclusive licence to Springer Nature B.V.)

Despite the promise of deep learning accelerated protein engineering, examples of such improved proteins are scarce. Here we report that a 3D convolutional neural network trained to associate amino acids with neighboring chemical microenvironments can guide identification of novel gain-of-function mutations that are not predicted by energetics-based approaches. Amalgamation of these mutations improved protein function in vivo across three diverse proteins by at least 5-fold. Furthermore, this model provides a means to interrogate the chemical space within protein microenvironments and identify specific chemical interactions that contribute to the gain-of-function phenotypes resulting from individual mutations.

Non-optimal vaginal microbiota, as observed in bacterial vaginosis (BV), is typically characterized by a depletion of beneficial lactobacilli and an abundance of numerous anaerobes. These non-optimal conditions are associated with subclinical cervicovaginal inflammation and an increased risk of HIV infection compared to women colonized with optimal vaginal microbiota dominated by lactobacilli. Lactic acid (LA) is a major organic acid metabolite produced by vaginal lactobacilli that elicits anti-inflammatory effects from cervicovaginal epithelial cells and is dramatically depleted during BV. However, it is unclear if LA retains its anti-inflammatory activity in the presence of vaginal microbiota metabolites comprising short chain fatty acids (SCFAs) and succinic acid, which are also produced by an optimal vaginal microbiota. Furthermore, the immunomodulatory effect of SCFAs and succinic acid on cervicovaginal epithelial cells at higher concentrations present during BV is unknown. Here we report that in the presence of physiologically relevant concentrations of SCFAs and succinic acid at pH 3.9 (as found in women with lactobacillus-dominated microbiota) LA induced an anti-inflammatory state in cervicovaginal epithelial cells and inhibited inflammation elicited by the toll-like receptor (TLR) agonists polyinosinic:polycytidylic acid and Pam3CSK4. When cervicovaginal epithelial cells were treated with a vaginal microbiota metabolite mixture representative of BV, containing a lower concentration of LA but higher concentrations of SCFA/succinic acid at pH 7, no anti-inflammatory was observed. Rather, the vaginal microbiota metabolite mixture representative of BV dysregulated the immune response of cervicovaginal epithelial cells during prolonged and sustained treatments. This was evidenced by increased basal and TLR-induced production of pro-inflammatory cytokines including tumor necrosis factor-α, but decreased basal production of chemokines including RANTES and IP-10. Further characterization of individual components of the BV vaginal microbiota mixture suggested that acetic acid is an important vaginal microbiota metabolite capable of eliciting diverse immunomodulatory effects on a range of cervicovaginal epithelial cell targets. These findings indicate that elevated levels of SCFAs are a potential source of cervicovaginal inflammation in women experiencing BV, and support the unique anti-inflammatory properties of LA on cervicovaginal epithelial cells as well as a role for LA or LA-producing lactobacilli to reverse genital inflammation associated with increased HIV risk., (Copyright © 2020 Delgado-Diaz, Tyssen, Hayward, Gugasyan, Hearps and Tachedjian.)

Statement of Problem: Currently, no commercially available facial shade guide exists in the United States for the fabrication of facial prostheses., Purpose: The purpose of this study was to measure facial skin and lip color in a human population sample stratified by age, gender, and race. Clustering analysis was used to determine optimal color coordinates for a proposed facial shade guide., Material and Methods: Participants (n=119) were recruited from 4 racial/ethnic groups, 5 age groups, and both genders. Reflectance measurements of participants' noses and lower lips were made by using a spectroradiometer and xenon arc lamp with a 45/0 optical configuration. Repeated measures ANOVA (α=.05), to identify skin and lip color differences, resulting from race, age, gender, and location, and a hierarchical clustering analysis, to identify clusters of skin colors) were used., Results: Significant contributors to L*a*b* facial color were race and facial location (P<.01). b* affected all factors (P<.05). Age affected only b* (P<.001), while gender affected only L* (P<.05) and b* (P<.05). Analyses identified 5 clusters of skin color., Conclusions: The study showed that skin color caused by age and gender primarily occurred within the yellow-blue axis. A significant lightness difference between gender groups was also found. Clustering analysis identified 5 distinct skin shade tabs., (Copyright © 2013 The Editorial Council of the Journal of Prosthetic Dentistry. Published by Mosby, Inc. All rights reserved.)

Statement of Problem: It remains unclear which gingival shade guide is most effective in producing the best visual shade match., Purpose: The purpose of this study was to determine and compare the coverage errors (CEs) of 2 different gingival shade guides and their combination in a selected population., Material and Methods: The CEs of the following gingival shade guides were evaluated: (1) Lucitone 199, (2) IPS Gingiva, and (3) a combination of the 2 shade guide systems. The spectral reflectance values of the center of each gingival shade tab (1.5 mm in diameter) were measured (without a backing) with a spectroradiometer and an external light source at wavelengths from 380 nm to 780 nm at 2-nm intervals. All spectral reflectance measurements were made using 0-degree observer and 45-degree illumination and then converted to CIE values. The attached gingiva of 120 subjects was measured with the same protocol. The CEs of each of the 120 subjects' attached gingiva for each of the 2 shade guide systems and their combination were determined and averaged. The mean minimum CEs were analyzed with repeated measures ANOVA and the post hoc Tukey HSD for multiple comparison (alpha=.05)., Results: A significant difference (P<.001) was found among the mean minimum CEs of the 2 gingival shade guide systems and their combination. The Tukey HSD test revealed that the mean minimum CEs for Lucitone 199 (Delta E 10.8 +/-4.4) were significantly different from the IPS Gingiva (Delta E 8.6 +/-3.6) shade guide system. However, the combination of the 2 gingival shade guide systems (Delta E 7.9 +/-3.4) was not significantly different from the IPS Gingiva system, but a significant difference was found among races (P=.016). No interactions were found between the other tested groups. The mean CE for white individuals (Delta E 8.3 +/-2.5) is significantly different and less than the mean CE for African Americans (Delta E 10.0 +/-4.4). The CEs for Asians (Delta E=9.1 +/-4.3) and others (Delta E 9.0 +/-4.4) are not significantly different from CEs for whites or African Americans., Conclusions: The IPS Gingiva shade guide system resulted in the lowest CEs compared to the Lucitone 199 shade guide system. The mean CE for Lucitone 199 was significantly greater than the mean CE for IPS Gingiva and for the combination of the 2 guides. The mean CE for whites is significantly different and less than the mean CE for African Americans. The CEs for Asians and others are not significantly different from CEs for whites or African Americans.

Statement of Problem: The color of vital craniofacial structures has not been measured accurately., Purpose: The purpose of this study was to determine the color of vital craniofacial structures and evaluate the validity and test-retest reliability of a noncontacting 45/0-degree optical configuration., Material and Methods: A spectroradiometer and an external light source were configured in a noncontacting 45/0-degree (45-degree illumination and 0-degree observer) optical configuration to measure the color of subjects' vital craniofacial structures (central and lateral incisor and canine, attached gingiva, lips, and facial skin). The 120 subjects were stratified into 5 age groups with 4 racial categories and balanced for gender. For evaluation of validity, linear regressions and 95% confidence intervals were calculated for DeltaL( *), Deltaa( *), Deltab( *) [color difference of (CIE) LAB values] between the measured and certified values of the 22 color patches of the DC Color Checker. For test-retest reliability, a random sample of 12 (10%) subjects was remeasured at a second visit. Paired t tests, correlations, and Bland-Altman analyses were performed between the first and second measurements of the 12 pairs of L( *), a( *), and b( *) values for the 6 craniofacial structures., Results: For validity, the mean color difference and linear regression for Commission Internationale d'Eclairage (CIE) LAB values between measured and certified color of the 22 opaque color patches were DeltaE of 1.46 and 0.99 for all regressions, respectively. Only Deltaa( *) did not contain zero in its 95% confidence interval. For test-retest reliability, no paired t tests were significantly different from each other, and the Pearson correlation coefficient ranged from 0.9 (9 pairs) to 0.7 (3 pairs). Ten of the 18 Bland-Altman plots showed good reliability., Conclusion: The spectral reflectance of craniofacial structures can be measured with acceptable validity and test-retest reliability using a noncontacting 45/0-degree optical configuration.