Structure of a unique PSII-Pcb tetrameric megacomplex in a chlorophyll d-containing cyanobacterium.

Acaryochloris marina is a unique cyanobacterium using chlorophyll d (Chl d) as its major pigment and thus can use far-red light for photosynthesis. Photosystem II (PSII) of A. marina associates with a number of prochlorophyte Chl-binding (Pcb) proteins to act as the light-harvesting system. We report here the cryo-electron microscopic structure of a PSII-Pcb megacomplex from A. marina at a 3.6-angstrom overall resolution and a 3.3-angstrom local resolution. The megacomplex is organized as a tetramer consisting of two PSII core dimers flanked by sixteen symmetrically related Pcb proteins, with a total molecular weight of 1.9 megadaltons. The structure reveals the detailed organization of PSII core consisting of 15 known protein subunits and an unknown subunit, the assembly of 4 Pcb antennas within each PSII monomer, and possible pathways of energy transfer within the megacomplex, providing deep insights into energy transfer and dissipation mechanisms within the PSII-Pcb megacomplex involved in far-red light utilization.

Identification of a highly conserved neutralizing epitope within the RBD region of diverse SARS-CoV-2 variants.

The constant emergence of SARS-CoV-2 variants continues to impair the efficacy of existing neutralizing antibodies, especially XBB.1.5 and EG.5, which showed exceptional immune evasion properties. Here, we identify a highly conserved neutralizing epitope targeted by a broad-spectrum neutralizing antibody BA7535, which demonstrates high neutralization potency against not only previous variants, such as Alpha, Beta, Gamma, Delta and Omicron BA.1-BA.5, but also more recently emerged Omicron subvariants, including BF.7, CH.1.1, XBB.1, XBB.1.5, XBB.1.9.1, EG.5. Structural analysis of the Omicron Spike trimer with BA7535-Fab using cryo-EM indicates that BA7535 recognizes a highly conserved cryptic receptor-binding domain (RBD) epitope, avoiding most of the mutational hot spots in RBD. Furthermore, structural simulation based on the interaction of BA7535-Fab/RBD complexes dissects the broadly neutralizing effect of BA7535 against latest variants. Therapeutic and prophylactic treatment with BA7535 alone or in combination with BA7208 protected female mice from the circulating Omicron BA.5 and XBB.1 variant infection, suggesting the highly conserved neutralizing epitope serves as a potential target for developing highly potent therapeutic antibodies and vaccines.

AmberTools.

AmberTools is a free and open-source collection of programs used to set up, run, and analyze molecular simulations. The newer features contained within AmberTools23 are briefly described in this Application note.

Arrhythmogenic mechanism of a novel ryanodine receptor mutation underlying sudden cardiac death.

AIMS: The ryanodine receptor 2 (RyR2) is essential for cardiac muscle excitation-contraction coupling; dysfunctional RyR2 participates in the development of inherited arrhythmogenic cardiac disease. In this study, a novel RyR2 mutation A690E is identified from a patient with family inheritance of sudden cardiac death, and we aimed to investigate the pathogenic basis of the mutation. METHODS AND RESULTS: We generated a mouse model that carried the A690E mutation. Mice were characterized by adrenergic-induced ventricular arrhythmias similar to clinical manifestation of the patient. Optical mapping studies revealed that isolated A690E hearts were prone to arrhythmogenesis and displayed frequency-dependence calcium transient alternans. Upon ß-adrenoceptor challenge, the concordant alternans was shifted towards discordant alternans that favour triggering ectopic beats and Ca2+ re-entry; similar phenomenon was also found in the A690E cardiomyocytes. In addition, we found that A690E cardiomyocytes manifested abnormal Ca2+ release and electrophysiological disorders, including an increased sensitivity to cytosolic Ca2+, an elevated diastolic RyR2-mediated Ca2+ leak, and an imbalance between Ca2+ leak and reuptake. Structural analyses reveal that the mutation directly impacts RyR2-FK506 binding protein interaction. CONCLUSION: In this study, we have identified a novel mutation in RyR2 that is associated with sudden cardiac death. By characterizing the function defects of mutant RyR2 in animal, whole heat, and cardiomyocytes, we demonstrated the pathogenic basis of the disease-causing mutation and provided a deeper mechanistic understanding of a life-threatening cardiac arrhythmia.

Biparatopic antibody BA7208/7125 effectively neutralizes SARS-CoV-2 variants including Omicron BA.1-BA.5.

SARS-CoV-2 Omicron subvariants have demonstrated extensive evasion from monoclonal antibodies (mAbs) developed for clinical use, which raises an urgent need to develop new broad-spectrum mAbs. Here, we report the isolation and analysis of two anti-RBD neutralizing antibodies BA7208 and BA7125 from mice engineered to produce human antibodies. While BA7125 showed broadly neutralizing activity against all variants except the Omicron sublineages, BA7208 was potently neutralizing against all tested SARS-CoV-2 variants (including Omicron BA.1-BA.5) except Mu. By combining BA7208 and BA7125 through the knobs-into-holes technology, we generated a biparatopic antibody BA7208/7125 that was able to neutralize all tested circulating SARS-CoV-2 variants. Cryo-electron microscopy structure of these broad-spectrum antibodies in complex with trimeric Delta and Omicron spike indicated that the contact residues are highly conserved and had minimal interactions with mutational residues in RBD of current variants. In addition, we showed that administration of BA7208/7125 via the intraperitoneal, intranasal, or aerosol inhalation route showed potent therapeutic efficacy against Omicron BA.1 and BA.2 in hACE2-transgenic and wild-type mice and, separately, effective prophylaxis. BA7208/7125 thus has the potential to be an effective candidate as an intervention against COVID-19.

Enantioselective semisynthesis of novel cephalotaxine esters with potent antineoplastic activities against leukemia.

Cephalotaxine-type alkaloids (CTAs), represented by homoharringtonine (HHT, 1), display potent efficacy against different types of leukemia cells. In this study, a method for hydrogenation of ß-substituted itaconic acid monoesters with chiral Ru[DTBM-SegPhos](OAc)2 was developed. This metal-catalyzed asymmetric hydrogenation enabled the convenient semisynthesis of novel cephalotaxine derivatives with chiral 2'-substituted-succinic acid 4-mono-methyl esters as side chains. The preliminary structure-activity relationship (SAR) of the compounds' antineoplastic activities was studied. Eventually, we discovered compound 10b with potent antineoplastic activities against leukemia and broadly anticancer activities against a panel of cancer cells. Our study provided a highly enantioselective process enabling the semisynthesis of cephalotaxine derivatives, which are interesting for further study on a scientific basis.

Discovery of BRAF/HDAC Dual Inhibitors Suppressing Proliferation of Human Colorectal Cancer Cells.

The combination of histone deacetylase inhibitor and BRAF inhibitor (BRAFi) has been shown to enhance the antineoplastic effect and reduce the progress of BRAFi resistance. In this study, a series of (thiazol-5-yl)pyrimidin-2-yl)amino)-N-hydroxyalkanamide derivatives were designed and synthesized as novel dual inhibitors of BRAF and HDACs using a pharmacophore hybrid strategy. In particular, compound 14b possessed potent activities against BRAF, HDAC1, and HDAC6 enzymes. It potently suppressed the proliferation of HT-29 cells harboring BRAFV600E mutation as well as HCT116 cells with wild-type BRAF. The dual inhibition against BRAF and HDAC downstream proteins was validated in both cells. Collectively, the results support 14b as a promising lead molecule for further development and a useful tool for studying the effects of BRAF/HDAC dual inhibitors.

Structure of infective Getah virus at 2.8 Å resolution determined by cryo-electron microscopy.

Getah virus (GETV), a member of the genus alphavirus, is a mosquito-borne pathogen that can cause pyrexia and reproductive losses in animals. Although antibodies to GETV have been found in over 10% of healthy people, there are no reports of clinical symptoms associated with GETV. The biological and pathological properties of GETV are largely unknown and antiviral or vaccine treatments against GETV are still unavailable due to a lack of knowledge of the structure of the GETV virion. Here, we present the structure of infective GETV at a resolution of 2.8 Å with the atomic models of the capsid protein and the envelope glycoproteins E1 and E2. We have identified numerous glycosylation and S-acylation sites in E1 and E2. The surface-exposed glycans indicate a possible impact on viral immune evasion and host cell invasion. The S-acylation sites might be involved in stabilizing the transmembrane assembly of E1 and E2. In addition, a cholesterol and a phospholipid molecule are observed in a transmembrane hydrophobic pocket, together with two more cholesterols surrounding the pocket. The cholesterol and phospholipid stabilize the hydrophobic pocket in the viral envelope membrane. The structural information will assist structure-based antiviral and vaccine screening, design, and optimization.

The Hippo pathway kinases LATS1 and LATS2 attenuate cellular responses to heavy metals through phosphorylating MTF1.

Heavy metals are both integral parts of cells and environmental toxicants, and their deregulation is associated with severe cellular dysfunction and various diseases. Here we show that the Hippo pathway plays a critical role in regulating heavy metal homeostasis. Hippo signalling deficiency promotes the transcription of heavy metal response genes and protects cells from heavy metal-induced toxicity, a process independent of its classic downstream effectors YAP and TAZ. Mechanistically, the Hippo pathway kinase LATS phosphorylates and inhibits MTF1, an essential transcription factor in the heavy metal response, resulting in the loss of heavy metal response gene transcription and cellular protection. Moreover, LATS activity is inhibited following heavy metal treatment, where accumulated zinc directly binds and inhibits LATS. Together, our study reveals an interplay between the Hippo pathway and heavy metals, providing insights into this growth-related pathway in tissue homeostasis and stress response.

Motif-dependent immune co-receptor interactome profiling by photoaffinity chemical proteomics.

Identification of the tyrosine phosphorylation (pY)-dependent interactome of immune co-receptors is crucial for understanding signal pathways involved in immunotherapy. However, identifying the motif-specific interactome for each pY commonly found on these multi-phosphorylated membrane proteins remains challenging. Here, we describe a photoaffinity-based chemical proteomic approach to dissect the motif-specific cytoplasmic interactomes of the critical immune co-receptor CD28. Various full-length CD28 cytoplasmic tails (CD28cyto) with defined pY and selectively replaced photo-methionine were synthesized and applied to explore three pY-motif-dependent CD28cyto interactomes. We identified a stand-alone interaction of phospholipase PLCG1 with the Y191 motif with enhanced affinity for the sequence neighboring the transmembrane domain. Importantly, taking advantage of native top-down mass spectrometry with a 193-nm laser, we discovered the direct association of a previously undefined pY218 motif with the kinase PKCθ through its C2 domain. This synthetic CD28cyto-based photoaffinity proteomic approach is generically applicable to the study of other immune co-receptors with multiple pY sites on their linear cytoplasmic tail.

Negatively charged phospholipids accelerate the membrane fusion activity of the plant-specific insert domain of an aspartic protease.

Various plants use antimicrobial proteins/peptides to resist phytopathogens. In the potato, Solanum tuberosum, the plant-specific insert (PSI) domain of an aspartic protease performs this role by disrupting phytopathogen plasma membranes. However, the mechanism by which PSI selects target membranes has not been elucidated. Here, we studied PSI-induced membrane fusion, focusing on the effects of lipid composition on fusion efficiency. Membrane fusion by the PSI involves an intermediate state whereby adjacent liposomes share their bilayers. We found that increasing the concentration of negatively charged phosphatidylserine (PS) phospholipids substantially accelerated PSI-mediated membrane fusion. NMR data demonstrated that PS did not affect the binding between the PSI and liposomes but had seminal effects on the dynamics of PSI interaction with liposomes. In PS-free liposomes, the PSI underwent significant motion, which was suppressed on PS-contained liposomes. Molecular dynamics simulations showed that the PSI binds to PS-containing membranes with a dominant angle ranging from -31° to 30°, with respect to the bilayer, and is closer to the membrane surfaces. In contrast, PSI is mobile and exhibits multiple topological states on the surface of PS-free membranes. Taken together, our data suggested that PS lipids limit the motion of the anchored PSI, bringing it closer to the membrane surface and efficiently bridging different liposomes to accelerate fusion. As most phytopathogens have a higher content of negatively charged lipids as compared with host cells, these results indicate that the PSI selectively targets negatively charged lipids, which likely represents a way of distinguishing the pathogen from the host.

Heparin remodels the microtubule-binding repeat R3 of Tau protein towards fibril-prone conformations.

Abnormal aggregation of proteins into pathological amyloid fibrils is implicated in a wide range of devastating human neurodegenerative diseases. Intracellular fibrillary inclusions formed by Tau protein are characterized as the hallmark of tauopathies, including Alzheimer's disease and frontotemporal dementia. Heparin has been often used to trigger Tau aggregation in in vitro studies. However, the conformational changes induced by heparin and the underlying mechanism of promotion of Tau aggregation by heparin are not well understood. Structural characterization of Tau oligomers in the early stage of fibrillation is of great importance but remains challenging due to their dynamic and heterogeneous nature. R3, the third microtubule-binding repeat of Tau, contains the fibril-nucleating core (PHF6) and is crucial for Tau aggregation. In this study, utilizing extensive all-atom replica-exchange molecular dynamic simulations, we explored the conformational ensembles of R3 monomer/dimer in the absence and presence of heparin. Our results show that without heparin, both monomeric and dimeric R3 preferentially adopt collapsed ß-sheet-containing conformations and PHF6 plays an important role in the formation of interchain ß-sheet structures, while in the presence of heparin, R3 can populate relatively extended disordered states where chain dimension is similar to that of R3 in Tau filaments. Through electrostatic, hydrogen-bonding and hydrophobic interactions, heparin has a preference for interacting with residues V306/Q307/K317/K321/H329/H330/K331 which distribute throughout the entire sequence of R3, in turn acting as a template to extend R3 conformations. More importantly, heparin alters intramolecular/intermolecular interaction patterns of R3 and increases the intermolecular contact regions. Our results suggest that heparin remodels the conformations of R3 towards fibril-prone structures by increasing chain dimension and intermolecular contact regions, which may shed light on the atomic mechanism of heparin-induced amyloid fibrillization of Tau protein.

Estimating the Roles of Protonation and Electronic Polarization in Absolute Binding Affinity Simulations.

Accurate prediction of binding free energies is critical to streamlining the drug development and protein design process. With the advent of GPU acceleration, absolute alchemical methods, which simulate the removal of ligand electrostatics and van der Waals interactions with the protein, have become routinely accessible and provide a physically rigorous approach that enables full consideration of flexibility and solvent interaction. However, standard explicit solvent simulations are unable to model protonation or electronic polarization changes upon ligand transfer from water to the protein interior, leading to inaccurate prediction of binding affinities for charged molecules. Here, we perform extensive simulation totaling ∼540 µs to benchmark the impact of modeling conditions on predictive accuracy for absolute alchemical simulations. Binding to urokinase plasminogen activator (UPA), a protein frequently overexpressed in metastatic tumors, is evaluated for a set of 10 inhibitors with extended flexibility, highly charged character, and titratable properties. We demonstrate that the alchemical simulations can be adapted to utilize the MBAR/PBSA method to improve the accuracy upon incorporating electronic polarization, highlighting the importance of polarization in alchemical simulations of binding affinities. Comparison of binding energy prediction at various protonation states indicates that proper electrostatic setup is also crucial in binding affinity prediction of charged systems, prompting us to propose an alternative binding mode with protonated ligand phenol and Hid-46 at the binding site, a testable hypothesis for future experimental validation.

Efficient formulation of polarizable Gaussian multipole electrostatics for biomolecular simulations.

Molecular dynamics simulations of biomolecules have been widely adopted in biomedical studies. As classical point-charge models continue to be used in routine biomolecular applications, there have been growing demands on developing polarizable force fields for handling more complicated biomolecular processes. Here, we focus on a recently proposed polarizable Gaussian Multipole (pGM) model for biomolecular simulations. A key benefit of pGM is its screening of all short-range electrostatic interactions in a physically consistent manner, which is critical for stable charge-fitting and is needed to reproduce molecular anisotropy. Another advantage of pGM is that each atom's multipoles are represented by a single Gaussian function or its derivatives, allowing for more efficient electrostatics than other Gaussian-based models. In this study, we present an efficient formulation for the pGM model defined with respect to a local frame formed with a set of covalent basis vectors. The covalent basis vectors are chosen to be along each atom's covalent bonding directions. The new local frame can better accommodate the fact that permanent dipoles are primarily aligned along covalent bonds due to the differences in electronegativity of bonded atoms. It also allows molecular flexibility during molecular simulations and facilitates an efficient formulation of analytical electrostatic forces without explicit torque computation. Subsequent numerical tests show that analytical atomic forces agree excellently with numerical finite-difference forces for the tested system. Finally, the new pGM electrostatics algorithm is interfaced with the particle mesh Ewald (PME) implementation in Amber for molecular simulations under the periodic boundary conditions. To validate the overall pGM/PME electrostatics, we conducted an NVE simulation for a small water box of 512 water molecules. Our results show that to achieve energy conservation in the polarizable model, it is important to ensure enough accuracy on both PME and induction iteration. It is hoped that the reformulated pGM model will facilitate the development of future force fields based on the pGM electrostatics for applications in biomolecular systems and processes where polarization plays crucial roles.

Improved Poisson-Boltzmann Methods for High-Performance Computing.

Implicit solvent models based on the Poisson-Boltzmann equation (PBE) have been widely used to study electrostatic interactions in biophysical processes. These models often treat the solvent and solute regions as high and low dielectric continua, leading to a large jump in dielectrics across the molecular surface which is difficult to handle. Higher order interface schemes are often needed to seek higher accuracy for PBE applications. However, these methods are usually very liberal in the use of grid points nearby the molecular surface, making them difficult to use on high-performance computing platforms. Alternatively, the harmonic average (HA) method has been used to approximate dielectric interface conditions near the molecular surface with surprisingly good convergence and is well suited for high-performance computing. By adopting a 7-point stencil, the HA method is advantageous in generating simple 7-banded coefficient matrices, which greatly facilitate linear system solution with dense data parallelism, on high-performance computing platforms such as a graphics processing unit (GPU). However, the HA method is limited due to its lower accuracy. Therefore, it would be of great interest for high-performance applications to develop more accurate methods while retaining the simplicity and effectiveness of the 7-point stencil discretization scheme. In this study, we have developed two new algorithms based on the spirit of the HA method by introducing more physical interface relations and imposing the discretized Poisson's equation to the second order, respectively. Our testing shows that, for typical biomolecules, the new methods significantly improve the numerical accuracy to that comparable to the second-order solvers and with ∼65% overall efficiency gain on widely available high-performance GPU platforms.

Mechanistic insight into E22Q-mutation-induced antiparallel-to-parallel ß-sheet transition of Aß16-22 fibrils: an all-atom simulation study.

Alzheimer's disease is associated with the abnormal self-assembly of amyloid-ß (Aß) peptide into toxic oligomers and fibrils. Recent experiments reported that Aß16-22, containing the central hydrophobic core (CHC) of Aß, formed antiparallel ß-sheet fibrils, while its E22Q mutant self-assembled into parallel ß-sheet fibrils. However, the molecular mechanisms underlying E22Q-mutation-induced parallel ß-sheet fibril formation are not well understood. Herein, we performed molecular dynamics (MD) simulations to study the dimerization processes of Aß16-22 and Aß16-22E22Q peptides. ß-Sheet dimers with diverse hydrogen bond arrangements were observed and they exhibited highly dynamic and interconverting properties. An antiparallel-to-parallel ß-sheet transition occurred in the assembly process of the E22Q mutant, but not in that of Aß16-22. During this conformational transformation process, the inter-molecular Q22-Q22 hydrogen bonds were first formed and acted as a binder to facilitate the two chains forming a parallel orientation, then the hydrophobic interactions between residues in the CHC region consolidated this arrangement and drove the main-chain H-bond formation, hence resulting in parallel ß-sheet formation. However, parallel ß-sheets were less populated than antiparallel ß-sheets of Aß16-22E22Q dimers. In order to explore whether parallel ß-sheets became dominant in larger size oligomers, we investigated the conformational ensembles of Aß16-22 and Aß16-22E22Q octamers by conducting replica exchange molecular dynamics (REMD) simulations. The REMD simulations revealed that the population of parallel ß-strand alignment increased with an increase of the size of ordered Aß16-22E22Q ß-sheet oligomers, implying that the formation of full parallel ß-sheets requires larger sized oligomers. Our findings provide a mechanistic explanation for the E22Q-mutation-induced formation of parallel ß-sheet fibrils observed experimentally.

Heterogeneous Dielectric Implicit Membrane Model for the Calculation of MMPBSA Binding Free Energies.

Membrane-bound protein receptors are a primary biological drug target, but the computational analysis of membrane proteins has been limited. In order to improve molecular mechanics Poisson-Boltzmann surface area (MMPBSA) binding free energy calculations for membrane protein-ligand systems, we have optimized a new heterogeneous dielectric implicit membrane model, with respect to free energy simulations in explicit membrane and explicit water, and implemented it into the Amber software suite. This new model supersedes our previous uniform, single dielectric implicit membrane model by allowing the dielectric constant to vary with depth within the membrane. We calculated MMPBSA binding free energies for the human purinergic platelet receptor (P2Y12R) and two of the muscarinic acetylcholine receptors (M2R and M3R) bound to various antagonist ligands using both membrane models, and we found that the heterogeneous dielectric membrane model has a stronger correlation with experimental binding affinities compared to the older model under otherwise identical conditions. This improved membrane model increases the utility of MMPBSA calculations for the rational design and improvement of future drug candidates.

An efficient second-order poisson-boltzmann method.

Immersed interface method (IIM) is a promising high-accuracy numerical scheme for the Poisson-Boltzmann model that has been widely used to study electrostatic interactions in biomolecules. However, the IIM suffers from instability and slow convergence for typical applications. In this study, we introduced both analytical interface and surface regulation into IIM to address these issues. The analytical interface setup leads to better accuracy and its convergence closely follows a quadratic manner as predicted by theory. The surface regulation further speeds up the convergence for nontrivial biomolecules. In addition, uncertainties of the numerical energies for tested systems are also reduced by about half. More interestingly, the analytical setup significantly improves the linear solver efficiency and stability by generating more precise and better-conditioned linear systems. Finally, we implemented the bottleneck linear system solver on GPUs to further improve the efficiency of the method, so it can be widely used for practical biomolecular applications. © 2019 Wiley Periodicals, Inc.

Conformational stability and dynamics of the cancer-associated isoform Δ133p53ß are modulated by p53 peptides and p53-specific DNA.

p53 is a tumor suppressor protein that maintains genome stability, but its Δ133p53ß and Δ160p53ß isoforms promote breast cancer cell invasion. The sequence truncations in the p53 core domain raise key questions related to their physicochemical properties, including structural stabilities, interaction mechanisms, and DNA-binding abilities. Herein, we investigated the conformational dynamics of Δ133p53ß and Δ160p53ß with and without binding to p53-specific DNA by using molecular dynamics simulations. We observed that the core domains of the 2 truncated isoforms are much less stable than wild-type (wt) p53ß, and the increased solvent exposure of their aggregation-triggering segment indicates their higher aggregation propensities than wt p53. We also found that Δ133p53ß stability is modulable by peptide or DNA interactions. Adding a p53 peptide (derived from truncated p53 sequence 107-129) may help stabilize Δ133p53. Most importantly, our simulations of p53 isomer-DNA complexes indicate that Δ133p53ß dimer, but not Δ160p53ß dimer, could form a stable complex with p53-specific DNA, which is consistent with recent experiments. This study provides physicochemical insight into Δ133p53ß, Δ133p53ß-DNA complexes, Δ133p53ß's pathologic mechanism, and peptide-based inhibitor design against p53-related cancers.-Lei, J., Qi, R., Tang, Y., Wang, W., Wei, G., Nussinov, R., Ma, B. Conformational stability and dynamics of the cancer-associated isoform Δ133p53ß are modulated by p53 peptides and p53-specific DNA.

Robustness and Efficiency of Poisson-Boltzmann Modeling on Graphics Processing Units.

Poisson-Boltzmann equation (PBE) based continuum electrostatics models have been widely used in modeling electrostatic interactions in biochemical processes, particularly in estimating protein-ligand binding affinities. Fast convergence of PBE solvers is crucial in binding affinity computations as numerous snapshots need to be processed. Efforts have been reported to develop PBE solvers on graphics processing units (GPUs) for efficient modeling of biomolecules, though only relatively simple successive over-relaxation and conjugate gradient methods were implemented. However, neither convergence nor scaling properties of the two methods are optimal for large biomolecules. On the other hand, geometric multigrid (MG) has been shown to be an optimal solver on CPUs, though no MG have been reported for biomolecular applications on GPUs. This is not a surprise as it is a more complex method and depends on simpler but limited iterative methods such as Gauss-Seidel in its core relaxation procedure. The robustness and efficiency of MG on GPUs are also unclear. Here we present an implementation and a thorough analysis of MG on GPUs. Our analysis shows that robustness is a more pronounced issue than efficiency for both MG and other tested solvers when the single precision is used for complex biomolecules. We further show how to balance robustness and efficiency utilizing MG's overall efficiency and conjugate gradient's robustness, pointing to a hybrid GPU solver with a good balance of efficiency and accuracy. The new PBE solver will significantly improve the computational throughput for a range of biomolecular applications on the GPU platforms.