Cryo-EM structures of the Spo11 core complex bound to DNA.

DNA double-strand breaks that initiate meiotic recombination are formed by the topoisomerase-relative enzyme Spo11, supported by conserved auxiliary factors. Because high-resolution structural data have not been available, many questions remain about the architecture of Spo11 and its partners and how they engage with DNA. We report cryo-electron microscopy structures at up to 3.3-Å resolution of DNA-bound core complexes of Saccharomyces cerevisiae Spo11 with Rec102, Rec104 and Ski8. In these structures, monomeric core complexes make extensive contacts with the DNA backbone and with the recessed 3'-OH and first 5' overhanging nucleotide, establishing the molecular determinants of DNA end-binding specificity and providing insight into DNA cleavage preferences in vivo. The structures of individual subunits and their interfaces, supported by functional data in yeast, provide insight into the role of metal ions in DNA binding and uncover unexpected structural variation in homologs of the Top6BL component of the core complex.

Cryo-EM structure of the Spo11 core complex bound to DNA.

The DNA double-strand breaks that initiate meiotic recombination are formed by topoisomerase relative Spo11, supported by conserved auxiliary factors. Because high-resolution structural data are lacking, many questions remain about the architecture of Spo11 and its partners and how they engage with DNA. We report cryo-EM structures at up to 3.3 Å resolution of DNA-bound core complexes of Saccharomyces cerevisiae Spo11 with Rec102, Rec104, and Ski8. In these structures, monomeric core complexes make extensive contacts with the DNA backbone and with the recessed 3'-OH and first 5' overhanging nucleotide, definitively establishing the molecular determinants of DNA end-binding specificity and providing insight into DNA cleavage preferences in vivo. The structures of individual subunits and their interfaces, supported by functional data in yeast, provide insight into the role of metal ions in DNA binding and uncover unexpected structural variation in homologs of the Top6BL component of the core complex.

Structure and DNA-bridging activity of the essential Rec114-Mei4 trimer interface.

The DNA double-strand breaks (DSBs) that initiate meiotic recombination are formed by an evolutionarily conserved suite of factors that includes Rec114 and Mei4 (RM), which regulate DSB formation both spatially and temporally. In vivo, these proteins form large immunostaining foci that are integrated with higher-order chromosome structures. In vitro, they form a 2:1 heterotrimeric complex that binds cooperatively to DNA to form large, dynamic condensates. However, understanding of the atomic structures and dynamic DNA binding properties of RM complexes is lacking. Here, we report a structural model of a heterotrimeric complex of the C terminus of Rec114 with the N terminus of Mei4, supported by nuclear magnetic resonance experiments. This minimal complex, which lacks the predicted intrinsically disordered region of Rec114, is sufficient to bind DNA and form condensates. Single-molecule experiments reveal that the minimal complex can bridge two or more DNA duplexes and can generate force to condense DNA through long-range interactions. AlphaFold2 predicts similar structural models for RM orthologs across diverse taxa despite their low degree of sequence similarity. These findings provide insight into the conserved networks of protein-protein and protein-DNA interactions that enable condensate formation and promote formation of meiotic DSBs.

Deep background-mismodeling-learned reconstruction for high-accuracy fluorescence diffuse optical tomography.

We present a deep background-mismodeling-learned reconstruction framework for high-accuracy fluorescence diffuse optical tomography (FDOT). A learnable regularizer incorporating background mismodeling is formulated in the form of certain mathematical constraints. The regularizer is then learned to obtain the background mismodeling automatically using a physics-informed deep network implicitly. Here, a deep-unrolled FIST-Net for optimizing L1-FDOT is specially designed to obtain fewer learning parameters. Experiments show that the accuracy of FDOT is significantly improved via implicitly learning the background mismodeling, which proves the validity of the deep background-mismodeling-learned reconstruction. The proposed framework can also be used as a general method to improve a class of image modalities based on linear inverse problems with unknown background modeling errors.

Real-time and accurate estimation ex vivo of four basic optical properties from thin tissue based on a cascade forward neural network.

Double integrating sphere measurements obtained from thin ex vivo tissues provides more spectral information and hence allows full estimation of all basic optical properties (OPs) theoretically. However, the ill-conditioned nature of the OP determination increases excessively with the reduction in tissue thickness. Therefore, it is crucial to develop a model for thin ex vivo tissues that is robust to noise. Herein, we present a deep learning solution to precisely extract four basic OPs in real-time from thin ex vivo tissues, leveraging a dedicated cascade forward neural network (CFNN) for each OP with an additional introduced input of the refractive index of the cuvette holder. The results show that the CFNN-based model enables accurate and fast evaluation of OPs, as well as robustness to noise. Our proposed method overcomes the highly ill-conditioned restriction of OP evaluation and can distinguish the effects of slight changes in measurable quantities without any a priori knowledge.

Prediction of cardiovascular disease risk based on major contributing features.

The risk of cardiovascular disease (CVD) is a serious health threat to human society worldwide. The use of machine learning methods to predict the risk of CVD is of great relevance to identify high-risk patients and take timely interventions. In this study, we propose the XGBH machine learning model, which is a CVD risk prediction model based on key contributing features. In this paper, the generalisation of the model was enhanced by adding retrospective data of 14,832 Chinese Shanxi CVD patients to the kaggle dataset. The XGBH risk prediction model proposed in this paper was validated to be highly accurate (AUC = 0.81) compared to the baseline risk score (AUC = 0.65), and the accuracy of the model for CVD risk prediction was improved with the inclusion of the conventional biometric BMI variable. To increase the clinical application of the model, a simpler diagnostic model was designed in this paper, which requires only three characteristics from the patient (age, value of systolic blood pressure and whether cholesterol is normal or not) to enable early intervention in the treatment of high-risk patients with a slight reduction in accuracy (AUC = 0.79). Ultimately, a CVD risk score model with few features and high accuracy will be established based on the main contributing features. Of course, further prospective studies, as well as studies with other populations, are needed to assess the actual clinical effectiveness of the XGBH risk prediction model.

High-fidelity mesoscopic fluorescence molecular tomography based on SSB-Net.

The imaging fidelity of mesoscopic fluorescence molecular tomography (MFMT) in reflective geometry suffers from spatial nonuniformity of measurement sensitivity and ill-posed reconstruction. In this study, we present a spatially adaptive split Bregman network (SSB-Net) to simultaneously overcome the spatial nonuniformity of measurement sensitivity and promote reconstruction sparsity. The SSB-Net is derived by unfolding the split Bregman algorithm. In each layer of the SSB-Net, residual block and 3D convolution neural networks (3D-CNNs) can adaptively learn spatially nonuniform error compensation, the spatially dependent proximal operator, and sparsity transformation. Simulations and experiments show that the proposed SSB-Net enables high-fidelity MFMT reconstruction of multifluorophores at different positions within a depth of a few millimeters. Our method paves the way for a practical reflection-mode diffuse optical imaging technique.

Structure and DNA bridging activity of the essential Rec114â€"Mei4 trimer interface.

The DNA double-strand breaks (DSBs) that initiate meiotic recombination are formed by an evolutionarily conserved suite of factors that includes Rec114 and Mei4 (RM), which regulate DSB formation both spatially and temporally. In vivo , these proteins form large immunostaining foci that are integrated with higher order chromosome structures. In vitro , they form a 2:1 heterotrimeric complex that binds cooperatively to DNA to form large, dynamic condensates. However, understanding of the atomic structures and dynamic DNA binding properties of RM complexes is lacking. Here, we report a structural model of a heterotrimeric complex of the C-terminus of Rec114 with the N-terminus of Mei4, supported by nuclear magnetic resonance experiments. This minimal complex, which lacks the predicted intrinsically disordered region of Rec114, is sufficient to bind DNA and form condensates. Single-molecule experiments reveal that the minimal complex can bridge two or more DNA duplexes and can generate force to condense DNA through long-range interactions. AlphaFold2 predicts similar structural models for RM orthologs across diverse taxa despite their low degree of sequence similarity. These findings provide insight into the conserved networks of protein-protein and protein-DNA interactions that enable condensate formation and promote formation of meiotic DSBs.

Mechanical regulation of the helicase activity of Zika virus NS3.

Zika virus (ZIKV) is a positive-sense single-stranded RNA virus that infects humans and can cause birth defects and neurological disorders. Its non-structural protein 3 (NS3) contains a protease domain and a helicase domain, both of which play essential roles during the viral life cycle. However, it has been shown that ZIKV NS3 has an inherently weak helicase activity, making it unable to unwind long RNA duplexes alone. How this activity is stimulated to process the viral genome and whether the two domains of NS3 are functionally coupled remain unclear. Here, we used optical tweezers to characterize the RNA-unwinding properties of ZIKV NS3-including its processivity, velocity, and step size-at the single-molecule level. We found that external forces that weaken the stability of the duplex RNA substrate significantly enhance the helicase activity of ZIKV NS3. On the other hand, we showed that the protease domain increases the binding affinity of NS3 to RNA but has only a minor effect on unwinding per se. Our findings suggest that the ZIKV NS3 helicase is activated on demand in the context of viral replication, a paradigm that may be generalizable to other flaviviruses.

Interpretable model-driven projected gradient descent network for high-quality fDOT reconstruction.

In fluorescence diffuse optical tomography (fDOT), the quality of reconstruction is severely limited by mismodeling and ill-posedness of inverse problems. Although data-driven deep learning methods improve the quality of image reconstruction, the network architecture lacks interpretability and requires a lot of data for training. We propose an interpretable model-driven projected gradient descent network (MPGD-Net) to improve the quality of fDOT reconstruction using only a few training samples. MPGD-Net unfolds projected gradient descent into a novel deep network architecture that is naturally interpretable. Simulation and in vivo experiments show that MPGD-Net greatly improves the fDOT reconstruction quality with superior generalization ability.

Computed structures of core eukaryotic protein complexes.

Protein-protein interactions play critical roles in biology, but the structures of many eukaryotic protein complexes are unknown, and there are likely many interactions not yet identified. We take advantage of advances in proteome-wide amino acid coevolution analysis and deep-learning­based structure modeling to systematically identify and build accurate models of core eukaryotic protein complexes within the Saccharomyces cerevisiae proteome. We use a combination of RoseTTAFold and AlphaFold to screen through paired multiple sequence alignments for 8.3 million pairs of yeast proteins, identify 1505 likely to interact, and build structure models for 106 previously unidentified assemblies and 806 that have not been structurally characterized. These complexes, which have as many as five subunits, play roles in almost all key processes in eukaryotic cells and provide broad insights into biological function.

Synthesis runs counter to directional folding of a nascent protein domain.

Folding of individual domains in large proteins during translation helps to avoid otherwise prevalent inter-domain misfolding. How folding intermediates observed in vitro for the majority of proteins relate to co-translational folding remains unclear. Combining in vivo and single-molecule experiments, we followed the co-translational folding of the G-domain, encompassing the first 293 amino acids of elongation factor G. Surprisingly, the domain remains unfolded until it is fully synthesized, without collapsing into molten globule-like states or forming stable intermediates. Upon fully emerging from the ribosome, the G-domain transitions to its stable native structure via folding intermediates. Our results suggest a strictly sequential folding pathway initiating from the C-terminus. Folding and synthesis thus proceed in opposite directions. The folding mechanism is likely imposed by the final structure and might have evolved to ensure efficient, timely folding of a highly abundant and essential protein.

Energetic dependencies dictate folding mechanism in a complex protein.

Large proteins with multiple domains are thought to fold cotranslationally to minimize interdomain misfolding. Once folded, domains interact with each other through the formation of extensive interfaces that are important for protein stability and function. However, multidomain protein folding and the energetics of domain interactions remain poorly understood. In elongation factor G (EF-G), a highly conserved protein composed of 5 domains, the 2 N-terminal domains form a stably structured unit cotranslationally. Using single-molecule optical tweezers, we have defined the steps leading to fully folded EF-G. We find that the central domain III of EF-G is highly dynamic and does not fold upon emerging from the ribosome. Surprisingly, a large interface with the N-terminal domains does not contribute to the stability of domain III. Instead, it requires interactions with its folded C-terminal neighbors to be stably structured. Because of the directionality of protein synthesis, this energetic dependency of domain III on its C-terminal neighbors disrupts cotranslational folding and imposes a posttranslational mechanism on the folding of the C-terminal part of EF-G. As a consequence, unfolded domains accumulate during synthesis, leading to the extensive population of misfolded species that interfere with productive folding. Domain III flexibility enables large-scale conformational transitions that are part of the EF-G functional cycle during ribosome translocation. Our results suggest that energetic tuning of domain stabilities, which is likely crucial for EF-G function, complicates the folding of this large multidomain protein.

First-order approximation of fluorescence excitation-transmission to accelerate fluorescence molecular tomography image reconstruction.

We present a method to accelerate image reconstruction in fluorescence molecular tomography based on the historical path fluorescence Monte Carlo model. The method exploits a first-order approximation expression during the fluorescence excitation-transmission process to merge the path and state information of the photon in a voxel. The experiments show that our method not only greatly reduces the amount of data required for storage in the hard disk and accelerates image reconstruction, but also maintains the quantitative and positioning accuracy of the conventional method.

The Ribosome Cooperates with a Chaperone to Guide Multi-domain Protein Folding.

Multi-domain proteins, containing several structural units within a single polypeptide, constitute a large fraction of all proteomes. Co-translational folding is assumed to simplify the conformational search problem for large proteins, but the events leading to correctly folded, functional structures remain poorly characterized. Similarly, how the ribosome and molecular chaperones promote efficient folding remains obscure. Using optical tweezers, we have dissected early folding events of nascent elongation factor G, a multi-domain protein that requires chaperones for folding. The ribosome and the chaperone trigger factor reduce inter-domain misfolding, permitting folding of the N-terminal G-domain. Successful completion of this step is a crucial prerequisite for folding of the next domain. Unexpectedly, co-translational folding does not proceed unidirectionally; emerging unfolded polypeptide can denature an already-folded domain. Trigger factor, but not the ribosome, protects against denaturation. The chaperone thus serves a previously unappreciated function, helping multi-domain proteins overcome inherent challenges during co-translational folding.

Folding up and Moving on-Nascent Protein Folding on the Ribosome.

All cellular proteins are synthesized by the ribosome, an intricate molecular machine that translates the information of protein coding genes into the amino acid alphabet. The linear polypeptides synthesized by the ribosome must generally fold into specific three-dimensional structures to become biologically active. Folding has long been recognized to begin before synthesis is complete. Recently, biochemical and biophysical studies have shed light onto how the ribosome shapes the folding pathways of nascent proteins. Here, we discuss recent progress that is beginning to define the role of the ribosome in the folding of newly synthesized polypeptides.

Genetically tunable frustration controls allostery in an intrinsically disordered transcription factor.

Intrinsically disordered proteins (IDPs) present a functional paradox because they lack stable tertiary structure, but nonetheless play a central role in signaling, utilizing a process known as allostery. Historically, allostery in structured proteins has been interpreted in terms of propagated structural changes that are induced by effector binding. Thus, it is not clear how IDPs, lacking such well-defined structures, can allosterically affect function. Here, we show a mechanism by which an IDP can allosterically control function by simultaneously tuning transcriptional activation and repression, using a novel strategy that relies on the principle of 'energetic frustration'. We demonstrate that human glucocorticoid receptor tunes this signaling in vivo by producing translational isoforms differing only in the length of the disordered region, which modulates the degree of frustration. We expect this frustration-based model of allostery will prove to be generally important in explaining signaling in other IDPs.

The ribosome destabilizes native and non-native structures in a nascent multidomain protein.

Correct folding is a prerequisite for the biological activity of most proteins. Folding has largely been studied using in vitro refolding assays with isolated small, robustly folding proteins. A substantial fraction of all cellular proteomes is composed of multidomain proteins that are often not amenable to this approach, and their folding remains poorly understood. These large proteins likely begin to fold during their synthesis by the ribosome, a large molecular machine that translates the genetic code. The ribosome affects how folding proceeds, but the underlying mechanisms remain largely obscure. We have utilized optical tweezers to study the folding of elongation factor G, a multidomain protein composed of five domains. We find that interactions among unfolded domains interfere with productive folding in the full-length protein. The N-terminal G-domain constitutes an independently folding unit that, upon in vitro refolding, adopts two similar states that correspond to the natively folded and a non-native, possibly misfolded structure. The ribosome destabilizes both of these states, suggesting a mechanism by which terminal misfolding into highly stable, non-native structures is avoided. The ribosome may thus directly contribute to efficient folding by modulating the folding of nascent multidomain proteins.

Structure of C-terminal tandem BRCT repeats of Rtt107 protein reveals critical role in interaction with phosphorylated histone H2A during DNA damage repair.

Rtt107 (regulator of Ty1 transposition 107; Esc4) is a DNA repair protein from Saccharomyces cerevisiae that can restore stalled replication forks following DNA damage. There are six BRCT (BRCA1 C-terminal) domains in Rtt107 that act as binding sites for other recruited proteins during DNA repair. Several Rtt107 binding partners have been identified, including Slx4, Rtt101, Rad55, and the Smc5/6 (structural maintenance of chromosome) protein complex. Rtt107 can reportedly be recruited to chromatin in the presence of Rtt101 and Rtt109 upon DNA damage, but the chromatin-binding site of Rtt107 has not been identified. Here, we report our investigation of the interaction between phosphorylated histone H2A (γH2A) and the C-terminal tandem BRCT repeats (BRCT(5)-BRCT(6)) of Rtt107. The crystal structures of BRCT(5)-BRCT(6) alone and in a complex with γH2A reveal the molecular basis of the Rtt107-γH2A interaction. We used in vitro mutagenesis and a fluorescence polarization assay to confirm the location of the Rtt107 motif that is crucial for this interaction. In addition, these assays indicated that this interaction requires the phosphorylation of H2A. An in vivo phenotypic analysis in yeast demonstrated the critical role of BRCT(5)-BRCT(6) and its interaction with γH2A during the DNA damage response. Our results shed new light on the molecular mechanism by which Rtt107 is recruited to chromatin in response to stalled DNA replication forks.