Biology subjects

Ni, X.

Publications and source records attributed to Ni, X..

At least 37 records · Page 2Linked to original sources

bioRxiv2024

Mapping AAV capsid sequences to functions through function-guided in silico evolution

Artificial intelligence (AI) offers significant potential to accelerate the functional engineering of adeno-associated virus (AAV) capsids in a time- and cost-efficient manner. However, existing approaches lack the ability to systematically map capsid sequences to multifunctional properties. To address this challenge, we developed ALICE, a knowledge-driven platform for in silico AAV capsid engineering that directly maps sequence to multiple functions. Our method incorporating a heuristic algorithm with contrastive learning principles, termed function-guided evolution (FE), to iteratively optimize high-performing capsid sequences generated by a naive language model toward desired functions. We elucidate FEs evolutionary mechanism, demonstrating its ability to navigate a function-guided landscape and generate capsids with tailored properties. After validating ALICEs ability to design murine-specific AAV capsids targeting Ly6a/Ly6c1 with limited training data, we developed ALICE-X by integrating a curriculum learning strategy. This upgrade facilitated exploration of sequence space beyond existing wet-lab datasets, enabling successful targeting of the human transferrin receptor 1 (hTfR1). The resulting engineered AAV variant, AAV.ALICE-H3, demonstrated improved viability, specific receptor targeting, and [~] 251-fold enhanced CNS tropism in human TFRC knock-in (hTFRC KI) mice compared to wildtype controls. This interpretable, knowledge-driven framework advances in silico AAV capsid design, offering broad applicability for translational gene therapy.

bioengineering↗

bioRxiv2024

Exploring the trade-off between deep-learning and explainable models for brain-machine interfaces

People with brain or spinal cord-related paralysis often need to rely on others for basic tasks, limiting their independence. A potential solution is brain-machine interfaces (BMIs), which could allow them to voluntarily control external devices (e.g., robotic arm) by decoding brain activity to movement commands. In the past decade, deep-learning decoders have achieved state-of-the-art results in most BMI applications, ranging from speech production to finger control. However, the black-box nature of deep-learning decoders could lead to unexpected behaviors, resulting in major safety concerns in real-world physical control scenarios. In these applications, explainable but lower-performing decoders, such as the Kalman filter (KF), remain the norm. In this study, we designed a BMI decoder based on KalmanNet, an extension of the KF that augments its operation with recurrent neural networks to compute the Kalman gain. This results in a varying "trust" that shifts between inputs and dynamics. We used this algorithm to predict finger movements from the brain activity of two monkeys. We compared KalmanNet results offline (pre-recorded data, n = 13 days) and online (real-time predictions, n = 5 days) with a simple KF and two recent deep-learning algorithms with state-of-the-art results: tcFNN and LSTM. KalmanNet achieved comparable or better results than other deep learning models in offline and online modes, relying on the dynamical model for stopping while depending more on neural inputs for initiating movements. We further validated this mechanism by implementing a heteroscedastic KF that used the same strategy, and it also approached state-of-the-art performance while remaining in the explainable domain of standard KFs. However, we also see two downsides to KalmanNet. KalmanNet shares the limited generalization ability of existing deep-learning decoders, and its usage of the KF as an inductive bias limits its performance in the presence of unseen noise distributions. Despite this trade-off, our analysis successfully integrates traditional controls and modern deep-learning approaches to motivate high-performing yet still explainable BMI designs.

neuroscience↗

bioRxiv2024

Identifying novel chemical matter against the Chikungunya virus nsP3 macrodomain through crystallographic fragment screening

Chikungunya virus (CHIKV) causes severe fever, rash and debilitating joint pain that can last for months 1,2or even years. Millions of people have been infected with CHIKV, mostly in low and middle-income countries, and the virus continues to spread into new areas due to the geographical expansion of its mosquito hosts. Its genome encodes a macrodomain, which functions as an ADP-ribosyl hydrolase, removing ADPr from viral and host-cell proteins interfering with the innate immune response. Mutational studies have shown that the CHIKV nsP3 macrodomain is necessary for viral replication, making it a potential target for the development of antiviral therapeutics. We, therefore, performed a high-throughput crystallographic fragment screen against the CHIKV nsP3 macrodomain, yielding 109 fragment hits covering the ADPr-binding site and two adjacent subsites that are absent in the homologous macrodomain of SARS-CoV-2 but may be present in other alphaviruses, such as Venezuelan equine encephalitis virus (VEEV) and eastern equine encephalitis virus (EEEV). Finally, a subset of overlapping fragments was used to manually design three fragment merges covering the adenine and oxyanion subsites. The rich dataset of chemical matter and structural information discovered from this fragment screen is publicly available and can be used as a starting point for developing a CHIKV nsP3 macrodomain inhibitor.

biochemistry↗

bioRxiv2024

The phylogenetic position of the Yunxian cranium elucidates the origin of Dragon Man and the Denisovans

Diverse Middle Pleistocene forms of Homo coexisted in Africa, Europe, and Asia, and its controversial whether these fossil humans represent different species or clades. The [~]1 Ma old Yunxian 2 fossil from China is crucial for understanding the cladogenesis of Homo and the origin of Homo sapiens. Here, we restored and reconstructed the distorted Yunxian 2 cranium using new technology. The results show that this cranium displays mosaic features of plesiomorphy and apomorphy. Geometric-morphometric, phylogenetic and Bayesian tip-dating analyses including the reconstructed Yunxian 2 suggest that it is an early member of the Asian Homo longi clade, which probably includes the Denisovans, and is the sister group to the Homo sapiens clade. Both the H. sapiens and H. longi clades had deep roots extending beyond the Middle Pleistocene, and the basal position of the Yunxian fossil cranium suggests that it represents a population lying close to the last common ancestor of the two clades. One-Sentence SummaryThe newly-reconstructed Yunxian 2 cranium represents a basal member of the Homo longi and Denisovan clade, and probably lies close to the last common ancestor of that clade and the clade of H. sapiens.

evolutionary biology↗

bioRxiv2024

Crystallographic fragment screen of Enterovirus D68 3C protease and iterative design of lead-like compounds using structure-guided expansions

The development of effective broad-spectrum antivirals forms an important part of preparing for future pandemics. One cause for concern is the currently emerging pathogen Enterovirus D68 (EV-D68) which primarily spreads through respiratory routes causing mostly mild to severe respiratory illness but, in severe cases, acute flaccid myelitis. The 3C protease of EV-D68 (3Cpro) is a potential target for the development of antiviral drugs due to its essential role in the viral life cycle and high sequence conservation amongst family members. In this study, we describe the identification of fragments which bind to3Cpro using crystallographic screening and the expansion of these into more lead-like compounds. The hits revealed interesting directions for hit-to-lead progression, specifically the importance of the pocket occupied by the conserved glutamine sidechain of the substrates and the interactions formed. Additionally, two pockets could be joined by not following the backbone of the native substrates, thus circumventing the screening issues arising from the flexibility of the catalytic triad. These observations of the novel binding modes of the chemical matter found by this screen can help shape future drug design campaigns against 3C proteases.

biochemistry↗

bioRxiv2024

High-throughput crystallographic fragment screening of Zika virus NS3 Helicase

The Zika virus (ZIKV), discovered in Africa in 1947, swiftly spread across continents, causing significant concern due to its recent association with microcephaly in newborns and Guillain-Barre syndrome in adults. Despite a decrease in prevalence, the potential for a resurgence remains, necessitating urgent therapeutic interventions. Like other flaviviruses, ZIKV presents promising drug targets within its replication machinery, notably the NS3 helicase (NS3Hel) protein, which plays critical roles in viral replication. However, a lack of structural information impedes the development of specific inhibitors targeting NS3Hel. Here we applied high-throughput crystallographic fragment screening on ZIKV NS3Hel, which yielded structures that reveal 3D binding poses of 46 fragments at multiple sites of the protein, including 11 unique fragments in the RNA-cleft site. These fragment structures provide templates for direct design of hit compounds and should thus assist the development of novel direct-acting antivirals against ZIKV and related flaviviruses, thus opening a promising avenue for combating future outbreaks.

biophysics↗

bioRxiv2024

Crystallographic fragment screening delivers diverse chemical scaffolds for Zika virus NS2B-NS3 protease inhibitor development

The Zika viral protease NS2B-NS3 is essential for the cleavage of viral polyprotein precursor into individual structural and non-structural (NS) proteins and is therefore an attractive drug target. Generation of a robust crystal system of co-expressed NS2B-NS3 protease has enabled us to perform a crystallographic fragment screening campaign with 1076 fragments. 47 fragments with diverse scaffolds were identified to bind in the active site of the protease, with another 6 fragments observed in a potential allosteric site. To identify binding sites that are intolerant to mutation and thus suppress the outgrowth of viruses resistant to inhibitors developed from bound fragments, we performed deep mutational scanning of NS2B-NS3 protease. Merging fragment hits yields an extensive set of mergers, defined as synthetically accessible compounds that recapitulate constellations of observed fragment-protein interactions. In addition, the highly sociable fragment hits enable rapid exploration of chemical space via algorithmic calculation and thus yield diverse possible starting points that maximally explore the binding opportunities to NS2B-NS3 protease, facilitating its resistance-resilient antiviral development.

biochemistry↗

bioRxiv2024

Crystallographic Fragment Screen of Coxsackievirus A16 2A Protease identifies new opportunities for the development of broad-spectrum anti-enterovirals

Enteroviruses are the causative agents of paediatric hand-foot-and-mouth disease, and a target for pandemic preparedness due to the risk of higher order complications in a large-scale outbreak. The 2A protease of these viruses is responsible for the self-cleavage of the poly protein, allowing for correct folding and assembly of capsid proteins in the final stages of viral replication. These 2A proteases are highly conserved between Enterovirus species, such as Enterovirus A71 and Coxsackievirus A16. Inhibition of the 2A protease deranges capsid folding and assembly, preventing formation of mature virions in host cells and making the protease a valuable target for antiviral activity. Herein, we describe a crystallographic fragment screening campaign that identified 75 fragments which bind to the 2A protease including 38 unique compounds shown to bind within the active site. These fragments reveal a path for the development of non-peptidomimetic inhibitors of the 2A protease with broad-spectrum anti-enteroviral activity.

biochemistry↗

bioRxiv2023

An end-to-end approach for protein folding by integrating Cryo-EM maps and sequence evolution

Protein structure modeling is an important but challenging task. Recent breakthroughs in Cryo-EM technology have led to rapid accumulation of Cryo-EM density maps, which facilitate scientists to determine protein structures but it remains time-consuming. Fortunately, artificial intelligence has great potential in automating this process. In this study, we present SMARTFold, a deep learning protein structure prediction model combining sequence alignment features and Cryo-EM density map features. First, using density map, we sample representative points along the predicted high confidence areas of protein backbone. Then we extract geometric features of these points and integrate these features with sequence alignment features in our proposed protein folding model. Extensive experiments confirm that our model performs best on both single-chain and multi-chain benchmark dataset compared with state-of-the-art methods, which makes it a reliable tool for protein atomic structure determination from Cryo-EM maps.

bioinformatics↗

bioRxiv2023

A parabrachial-hypothalamic parallel circuit governs cold defense in mice

Thermal homeostasis is vital for mammals and is controlled by brain neurocircuits. Remarkable advances have been made in understanding how neurocircuits centered in the hypothalamic preoptic area (POA), the brains thermoregulation center, control warm defense, whereas mechanisms by which the POA regulates cold defense remain unclear. Here, we confirmed that the pathway from the lateral parabrachial nucleus (LPB) to the POA, is critical for cold defense. Parallel to this pathway, we uncovered that a pathway from the LPB to the dorsomedial hypothalamus (DMH), namely the LPB[->]DMH pathway, is also essential for cold defense. Projection-specific blockings revealed that both pathways provide an equivalent and cumulative contribution to cold defense, forming a parallel circuit. Specifically, activation of the LPB[->]DMH pathway induced strong cold-defense responses, including increases in thermogenesis of brown adipose tissue (BAT), muscle shivering, heart rate, and physical activity. Further, we identified a subpopulation of somatostatin+ neurons in the LPB that target the DMH to promote BAT thermogenesis. Therefore, we reveal a parabrachial-hypothalamic parallel circuit in governing cold defense in mice. This not only enables resilience to hypothermia but also provides a scalable and robust network in heat production, reshaping our understanding of how neural circuits regulate essential homeostatic behaviors.

neuroscience↗

bioRxiv2022

Cryo-EM complex structure of active GPR75 with a nanobody

Although there has been enormous progress in the last half-century in the drug discovery targeting obesity and associated co-morbidities, the clinical treatment of obesity remains tremendously challenging. GPR75 is an orphan receptor and is suggested to be a potential novel target for the control of obesity and related metabolic disorders. Inhibition of the GPR75 signaling pathway by small molecules, antibodies, or genetic manipulations may provide a therapeutic strategy for obesity. Here, we report the active-like Cryo-EM structure of human GPR75 with an intracellular nanobody, which reveals the receptor activation mechanism. The extensive interaction network required to achieve the active structure helps explain the allosteric coupling between the orthosteric pocket and the G-protein coupling domain. The well-defined orthosteric ligand binding pocket of human GPR75 provides a structural basis for anti-obesity drug discovery.

biophysics↗

bioRxiv2022

Increased α-ketoglutarate as a missing link from the C3-C4 intermediate state to C4 photosynthesis in the genus Flaveria

As a complex trait, C4 photosynthesis has multiple independent origins in evolution. Phylogenetic evidence and theoretical analysis suggest that C2 photosynthesis, which is driven by glycine decarboxylation in the bundle sheath cell, may function as a bridge from C3 towards C4 photosynthesis. However, the exact molecular mechanism underlying the transition between C2 photosynthesis towards C4 photosynthesis remains elusive. Here, we provide multiple evidence suggesting a role of higher -ketoglutarate (AKG) concentration during this transition. Metabolomic data of 12 Flaveria species, including multiple photosynthetic types, show that AKG concentration initially increases in the C3-C4 intermediate with a further increase in C4 species. Petiole feeding of AKG increased the concentrations of C4 related metabolites in C3-C4 and C4 species but not the activity of C4 related enzymes. Sequence analysis shows that glutamate synthase (Fd-GOGAT), which catalyzes the generation of glutamate using AKG, was under strong positive selection during the evolution of C4 photosynthesis. Simulations with a constraint-based model for C3-C4 intermediate further show that decreasing the activity of Fd-GOGAT facilitates the transition from a C2-dominant to a C4-dominant CO2 concentrating mechanisms. All these provide an insight into the mechanistic switch from C3-C4 intermediate to C4 photosynthesis.

plant biology↗

bioRxiv2021

A real-time assay for cell-penetrating peptide-mediated delivery of molecular cargos

Cell-penetrating peptides (CPPs) are capable of transporting molecules to which they are tethered across cellular membranes. Unsurprisingly, CPPs have attracted attention for their potential drug delivery applications, but several technical hurdles remain to be overcome. Chief among them is the so-called endosomal escape problem, i.e. the propensity of CPP-cargo molecules to be endocytosed but remain entrapped in endosomes rather than reaching the cytosol. Previously, a CPP fused to calmodulin that bound calmodulin binding site-containing cargos was shown to efficiently deliver cargos to the cytoplasm, effectively overcoming the endosomal escape problem. The CPP-adaptor, "TAT-CaM," evinces delivery at nM concentrations and more rapidly than we had previously been able to measure. To better understand the kinetics and mechanism of CPP-adaptor-mediated cargo delivery, a real-time cell penetrating assay was developed in which a flow chamber containing cultured cells was installed on the stage of a confocal microscope to allow for observation ab initio. Also examined in this study was an improved CPP-adaptor that utilizes naked mole rat (Heterocephalus glaber) calmodulin in place of human and results in superior internalization, likely due to its lesser net negative charge. Adaptor-cargo complexes were delivered into the flow chamber and fluorescence intensity in the midpoint of baby hamster kidney cells was measured as a function of time. Delivery of 400 nM cargo was observed within seven minutes and fluorescence continued to increase linearly as a function of time. Cargo-only control experiments showed that the minimal uptake which occurred independently of the CPP-adaptor resulted in punctate localization consistent with endosomal entrapment. A distance analysis was performed for cell-penetration experiments in which CPP-adaptor-delivered cargo showing wider dispersions throughout cells as compared to an analogous covalently-bound CPP-cargo. Small molecule endocytosis inhibitors did not have significant effects upon delivery. The real-time assay is an improvement upon static endpoint assays and should be informative in a broad array of applications.

cell biology↗

bioRxiv2021

Bipartite binding and partial inhibition links DEPTOR and mTOR in a mutually antagonistic embrace

mTORC1 is a kinase complex regulating cell growth, proliferation and survival. Because mis-regulation of DEPTOR, an endogenous mTORC1 inhibitor, is associated with some cancers, we reconstituted mTORC1 with DEPTOR to understand its function. We find that DEPTOR is a unique partial mTORC1 inhibitor that may have evolved to preserve feedback inhibition of PI3K. Counterintuitively, mTORC1 activated by RHEB or oncogenic mutation is much more potently inhibited by DEPTOR. Although DEPTOR partially inhibits mTORC1, mTORC1 prevents this inhibition by phosphorylating DEPTOR, a mutual antagonism that requires no exogenous factors. Structural analyses of the mTORC1/DEPTOR complex showed DEPTORs PDZ domain interacting with the mTOR FAT region, and the unstructured linker preceding the PDZ binding to the mTOR FRB domain. Here we show, in contrast to previous cellular studies, that both the PDZ and linker regions are essential for inhibition, and it is likely that interaction with the FRB is crucial to the unique partial inhibition.

biochemistry↗

bioRxiv2021

Dorsomedial hypothalamic BDNF neurons integrate thermal afferent signals to control energy expenditure

Mutations in the gene brain-derived neurotrophic factor (BDNF) cause obesity in humans. BDNF signaling and its expressing neurons in the hypothalamus help control feeding, energy expenditure (EE), and physical activity. However, whether the BDNF neurons interact with another EE-regulating system, the thermoregulation circuitry, remains unclear. Here, we show that BDNF neurons in the dorsomedial hypothalamus (DMH) are activated by environmental cooling and sufficient to induce body temperature increases and brown adipose tissue (BAT) thermogenesis. Conversely, blocking these neurons impairs BAT thermogenesis and cold defense, causing body weight gain and glucose intolerance. DMH BDNF neurons are therefore an important type of thermoregulatory neuron, integrating thermal afferent signals to control EE during cold defense. This reveals a critical intersection between the BDNF circuitry and the thermoregulatory system.

neuroscience↗

bioRxiv2021

Structural insights into plasticity and discovery of remdesivir metabolite GS-441524 binding in SARS-CoV-2 macrodomain

The nsP3 macrodomain is a conserved protein interaction module that plays essential regulatory roles in host immune response by recognizing and removing posttranslational ADP-ribosylation sites during SARS-CoV-2 infection. Thus, targeting this protein domain may offer a therapeutic strategy to combat the current and future virus pandemics. To assist inhibitor development efforts, we report here a comprehensive set of macrodomain crystal structures complexed with diverse naturally-occurring nucleotides, small molecules as well as nucleotide analogues including GS-441524 and its phosphorylated analogue, active metabolites of remdesivir. The presented data strengthen our understanding of the SARS-CoV-2 macrodomain structural plasticity and it provides chemical starting points for future inhibitor development.

biochemistry↗

bioRxiv2021

Structure and inhibitor binding characterization of oncogenic MLLT1 mutants

Dysfunction of YEATS-domain-containing MLLT1, an acetyl/acyl-lysine dependent epigenetic reader domain, has been implicated in the development of aggressive cancers. Mutations in the YEATS domain have been recently reported as a cause of MLLT1 aberrant reader function. However, structural basis for the reported alterations in affinity for acetyled/acylated histone has remained elusive. Here, we report the crystal structures of both insertion and substitution present in cancer, revealing significant conformational changes of the YEATS-domain loop 8. Structural comparison demonstrates that such alteration not only altered the binding interface for acetylated/acylated histones, but the sequence alterations in the T1 loop may enable dimeric assembly consistent inducing self-association behavior. Nevertheless, we show that also the MLLT1 mutants can be targeted by developed acetyllysine mimetic inhibitors with affinities similarly to wild type. Our report provides a structural basis for the altered behaviors and potential strategy for targeting oncogenic MLLT1 mutants.

biochemistry↗

bioRxiv2020

Parabrachial neuron types categorically encode thermoregulation variables during heat defense

Heat defense is crucial for survival and fitness, and its dysregulation may result in deaths due to poor management. Transmission of thermosensory signals into hypothalamic thermoregulation centers represent a key layer of regulation in heat defense. However, the mechanism by which these signals are transmitted into the hypothalamus remains poorly understood. Here, we reveal that glutamatergic prodynorphin and cholecystokinin neuron populations in the lateral parabrachial (LPB) are progressively recruited to defend elevated body temperature. These two nonoverlapping neuron types form circuitries with downstream preoptic hypothalamic neurons to inhibit BAT thermogenesis and activate tail vasodilation, respectively. Both circuitries are selectively activated by warm temperatures and are required for fever limiting. The prodynorphin circuitry is further required for regulation of energy expenditure and weight homeostasis. Thus, these findings establish that the genetic and functional specificity of heat defense neurons occurs as early as in the LPB and uncover categorical neuron types for encoding two heat defense variables, which may provide targets for treating thermoregulation disorders.

neuroscience↗