Search

ImportanceWith timely collection of SARS-CoV-2 viral genome sequences, it is important to apply efficient data analytics to detect emerging variants at the earliest time.ObjectiveTo evaluate the application of a statistical learning strategy (SLS) to improve early detection of novel SARS-CoV-2 variants using viral sequence data from global surveillance.Design, Setting, and ParticipantsThis case series applied an SLS to viral genomic sequence data collected from 63 686 individuals in Africa and 531 827 individuals in the United States with SARS-CoV-2. Data were collected from January 1, 2020, to December 28, 2021.Main Outcomes and MeasuresThe outcome was an indicator of Omicron variant derived from viral sequences. Centering on a temporally collected outcome, the SLS used the generalized additive model to estimate locally averaged Omicron caseload percentages (OCPs) over time to characterize Omicron expansion and to estimate when OCP exceeded 10%, 25%, 50%, and 75% of the caseload. Additionally, an unsupervised learning technique was applied to visualize Omicron expansions, and temporal and spatial distributions of Omicron cases were investigated.ResultsIn total, there were 2698 cases of Omicron in Africa and 12 141 in the United States. The SLS found that Omicron was detectable in South Africa as early as December 31, 2020. With 10% OCP as a threshold, it may have been possible to declare Omicron a variant of concern as early as November 4, 2021, in South Africa. In the United States, the application of SLS suggested that the first case was detectable on November 21, 2021.Conclusions and RelevanceThe application of SLS demonstrates how the Omicron variant may have emerged and expanded in Africa and the United States. Earlier detection could help the global effort in disease prevention and control. To optimize early detection, efficient data analytics, such as SLS, could assist in the rapid identification of new variants as soon as they emerge, with or without lineages designated, using viral sequence data from global surveillance.

The ribose 2′-hydroxyl is the key chemical difference between RNA and DNA and primary source of their divergent structural and functional characteristics. Macromolecular X-ray diffraction experiments typically do not reveal the positions of hydrogen atoms. Thus, standard crystallography cannot determine 2′-OH orientation (H2′-C2′-O2′-HO2′ torsion angle) and its potential roles in sculpting the RNA backbone and the expansive fold space. Here, we report the first neutron crystal structure of an RNA, the Escherichia coli rRNA Sarcin-Ricin Loop (SRL). 2′-OD orientations were established for all 27 residues and revealed O-D bonds pointing toward backbone (O3′, 13 observations), nucleobase (11) or sugar (3). Most riboses in the SRL stem region show a 2′-OD backbone-orientation. GAGA-tetraloop riboses display a 2′-OD base-orientation. An atypical C2′-endo sugar pucker is strictly correlated with a 2′-OD sugar-orientation. Neutrons reveal the strong preference of the 2′-OH to donate in H-bonds and that 2′-OH orientation affects both backbone geometry and ribose pucker. We discuss 2′-OH and water molecule orientations in the SRL neutron structure and compare with results from a solution phase 10 μs MD simulation. We demonstrate that joint cryo-neutron/X-ray crystallography offers an all-in-one approach to determine the complete structural properties of RNA, i.e. geometry, conformation, protonation state and hydration structure.

ImportanceEarlier detection of emerging novel SARS-COV-2 variants is important for public health surveillance of potential viral threats and for earlier prevention research. Artificial intelligence may facilitate early detection of SARS-CoV2 emerging novel variants based on variant-specific mutation haplotypes and, in turn, be associated with enhanced implementation of risk-stratified public health prevention strategies.ObjectiveTo develop a haplotype-based artificial intelligence (HAI) model for identifying novel variants, including mixture variants (MVs) of known variants and new variants with novel mutations.Design, Setting, and ParticipantsThis cross-sectional study used serially observed viral genomic sequences globally (prior to March 14, 2022) to train and validate the HAI model and used it to identify variants arising from a prospective set of viruses from March 15 to May 18, 2022.Main Outcomes and MeasuresViral sequences, collection dates, and locations were subjected to statistical learning analysis to estimate variant-specific core mutations and haplotype frequencies, which were then used to construct an HAI model to identify novel variants.ResultsThrough training on more than 5 million viral sequences, an HAI model was built, and its identification performance was validated on an independent validation set of more than 5 million viruses. Its identification performance was assessed on a prospective set of 344 901 viruses. In addition to achieving an accuracy of 92.8% (95% CI within 0.1%), the HAI model identified 4 Omicron MVs (Omicron-Alpha, Omicron-Delta, Omicron-Epsilon, and Omicron-Zeta), 2 Delta MVs (Delta-Kappa and Delta-Zeta), and 1 Alpha-Epsilon MV, among which Omicron-Epsilon MVs were most frequent (609/657 MVs [92.7%]). Furthermore, the HAI model found that 1699 Omicron viruses had unidentifiable variants given that these variants acquired novel mutations. Lastly, 524 variant-unassigned and variant-unidentifiable viruses carried 16 novel mutations, 8 of which were increasing in prevalence percentages as of May 2022.Conclusions and RelevanceIn this cross-sectional study, an HAI model found SARS-COV-2 viruses with MV or novel mutations in the global population, which may require closer examination and monitoring. These results suggest that HAI may complement phylogenic variant assignment, providing additional insights into emerging novel variants in the population.

SARS-CoV-2 is spreading worldwide with continuously evolving variants, some of which occur in the Spike protein and appear to increase the viral transmissibility. However, variants that cause severe COVID-19 or lead to other breakthroughs have not been well characterized. To discover such viral variants, we assembled a cohort of 683 COVID-19 patients; 388 inpatients (“cases”) and 295 outpatients (“controls”) from April to August 2020 using electronically captured COVID test request forms and sequenced their viral genomes. To improve the analytic power, we accessed 7,137 viral sequences in Washington State to filter out viral single nucleotide variants (SNVs) that did not have significant expansions over the collection period. Applying this filter led to the identification of 53 SNVs that were statistically significant, of which 13 SNVs each had 3 or more variant copies in the discovery cohort. Correlating these selected SNVs with case/control status, eight SNVs were found to significantly associate with inpatient status (q-values

The emergence and establishment of SARS-CoV-2 variants of interest (VOI) and variants of concern (VOC) highlight the importance of genomic surveillance. We propose a statistical learning strategy (SLS) for identifying and spatiotemporally tracking potentially relevant Spike protein mutations. We analyzed 167,893 Spike protein sequences from US COVID-19 cases (excluding 21,391 sequences from VOI/VOC strains) deposited at GISAID from January 19, 2020 to March 15, 2021. Alignment against the reference Spike protein sequence led to the identification of viral residue variants (VRVs), i.e., residues harboring a substitution compared to the reference strain. Next, generalized additive models were applied to model VRV temporal dynamics, to identify VRVs with significant and substantial dynamics (false discovery rate q-value 10% on at least one day).Unsupervised learning was then applied to hierarchically organize VRVs by spatiotemporal patterns and identify VRV-haplotypes. Finally, homology modelling was performed to gain insight into potential impact of VRVs on Spike protein structure. We identified 90 VRVs, 71 of which have not previously been observed in a VOI/VOC, and 35 of which have emerged recently and are durably present. Our analysis identifies 17 VRVs ∼91 days earlier than their first corresponding VOI/VOC publication. Unsupervised learning revealed eight VRV-haplotypes of 4 VRVs or more, suggesting two emerging strains (B1.1.222 and B.1.234). Structural modeling supported potential functional impact of the D1118H and L452R mutations. The SLS approach equally monitors all Spike residues over time, independently of existing phylogenic classifications, and is complementary to existing genomic surveillance methods.

Background: HLA-DR4, a common antigen of HLA-DRB1, has multiple subtypes that are strongly associated with risk of type 1 diabetes (T1D); however, some are risk neutral or resistant. The pathobiological mechanism of HLA-DR4 subtypes remains to be elucidated. Methods: We used a population-based case-control study of T1D (962 patients and 636 controls) to decipher genetic associations of HLA-DR4 subtypes and specific residues with susceptibility to T1D. Using a birth cohort of 7865 children with periodically measured islet autoantibodies (GADA, IAA or IA-2A), we proposed to validate discovered genetic associations with a totally different study design and time-to-seroconversions prior to clinical onset of T1D. A novel analytic strategy hierarchically organized the HLA-DRB1 alleles by sequence similarity and identified critical amino acid residues by minimizing local genomic architecture and higher-order interactions. Findings: Three amino acid residues of HLA-DRB1 (β71, β74, β86) were found to be predictive of T1D risk in the population-based study. The “KAG” motif, corresponding to HLA-DRB1×04:01, was most strongly associated with T1D risk ([O]dds [R]atio=3.64, p = 3.19 × 10−64). Three less frequent motifs (“EAV”, OR = 2.55, p = 0.025; “RAG”, OR = 1.93, p = 0.043; and “RAV”, OR = 1.56, p = 0.003) were associated with T1D risk, while two motifs (“REG” and “REV”) were equally protective (OR = 0.11, p = 4.23 × 10−4). In an independent birth cohort of HLA-DR3 and HLA-DR4 subjects, those having the “KAG” motif had increased risk for time-to-seroconversion (Hazard Ratio = 1.74, p = 6.51 × 10−14) after adjusting potential confounders. Interpretations: DNA sequence variation in HLA-DRB1 at positions β71, β74, and β86 are non-conservative (β74 A→E, β71 E vs K vs R and β86 G vs V). They result in substantial differences in peptide antigen anchor pocket preferences at p1, p4 and potentially neighboring regions such as pocket p7. Differential peptide antigen binding is likely to be affected. These sequence substitutions may account for most of the HLA-DR4 contribution to T1D risk as illustrated in two HLA-peptide model complexes of the T1D autoantigens preproinsulin and GAD65. Funding: National Institute of Diabetes and Digestive and Kidney Diseases and the Swedish Child Diabetes Foundation and the Swedish Research Council.

BACKGROUND: The emergence and establishment of SARS-CoV-2 variants of interest (VOI) and variants of concern (VOC) highlight the importance of genomic surveillance. We propose a statistical learning strategy (SLS) for identifying and spatiotemporally tracking Spike protein mutations potentially relevant to public health. METHODS: We analyzed 189,284 Spike protein sequences from US COVID-19 cases, deposited at GISAID from January 19, 2020 to March 15, 2021. Alignment against the reference Spike protein sequence led to the identification of viral residue variants (VRVs), i.e., residues harboring a substitution. Next, generalized additive models were applied to model VRV temporal dynamics, to identify VRVs with significant and substantial dynamics (false discovery rate q-value 10% on at least one day). Unsupervised learning was then applied to hierarchically organize VRVs by spatiotemporal patterns and identify VRV-haplotypes. Finally, homology modelling was performed to gain insight into potential impact of VRVs on Spike protein structure. FINDINGS: We identified 88 VRVs, 71 of which have not previously been observed in a VOI/VOC, and 44 of which have emerged recently and appear to be durably present. Compared to identified VOIs/VOCs, our analysis identifies 17 VRVs ~78 days earlier than their first description in a publication. Unsupervised learning revealed several VRV-haplotypes that were of reported lineages in addition to several novel VRV-haplotypes of unknown lineage. Structural modeling supported functional impact of the D1118H and L452R mutations. INTERPRETATION: The SLS approach equally monitors all Spike residues over time without dependence on existing phylogenic classifications, and is thus complementary to existing genomic surveillance methods. FUNDING: National Institute of Allergy and Infectious Diseases DECLARATION OF INTERESTS: No conflicts of interest relevant to this article were reported.

GNRA (N = A, C, G, or U; R = A or G) tetraloops are common RNA secondary structural motifs and feature a phosphate stacked atop a nucleobase. The rRNA sarcin/ricin loop (SRL) is capped by G ApGA, and the phosphate p stacks on G . We recently found that regiospecific incorporation of a single dithiophosphate (PS2) but not a monothiophosphate (PSO) instead of phosphate in the backbone of RNA aptamers dramatically increases the binding affinity for their targets. In the RNA:thrombin complex, the key contribution to the 1000-fold tighter binding stems from an edge-on contact between PS2 and a phenylalanine ring. Here we investigated the consequences of replacing the SRL phosphate engaged in a face-on interaction with guanine with either PS2 or PSO for stability. We found that PS2···G and Rp-PSO···G contacts stabilize modified SRLs compared to the parent loop to unexpected levels: up to 6.3 °C in melting temperature Tm and -4.7 kcal/mol in ΔΔG°. Crystal structures demonstrate that the vertical distance to guanine for the closest sulfur is just 0.05 A longer on average compared to that of oxygen despite the larger van der Waals radius of the former (1.80 A for S vs 1.52 A for O). The higher stability is enthalpy-based, and the negative charge as assessed by a neutral methylphosphonate modification plays only a minor role. Quantum mechanical/molecular mechanical calculations are supportive of favorable dispersion attraction interactions by sulfur making the dominant contribution. A stacking interaction between phosphate and guanine (SRL) or uracil (U-turn) is also found in newly classified RNA tetraloop families besides GNRA.

Regiospecific replacement of a single phosphate (PO2) by a dithiophosphate (PS2) group in an RNA can dramatically increase its binding affinity for a target protein. Thus, complexes between antithrombin and anti-VEGF RNA aptamers with single dithiophosphate moieties and thrombin and VEGF, respectively, display equilibrium dissociation constants KD of ca. 1 pM, 1000-fold tighter than the native RNA complexes (ca. 1 nM). Inspection of crystal structures of the native and PS2-RNA aptamer:thrombin complexes reveals an RNA-induced fit in the latter. This leads to a close approach between PS2 and the phenyl ring edge of Phe-232 that is surrounded by pairs of lysines and arginines. To better understand the origins of the tighter binding and individual contributions to the interaction energy, we carried out QM calculations with phosphate- and dithiophosphate-benzene and dimethyl phosphate- and dimethyl dithiophosphate-benzene model systems. These calculations demonstrate that the dithiophosphate-benzene interaction is much stronger than the corresponding interaction with phosphate. QM/MM calculations with the full complexes confirmed this finding and support the hypothesis that the electric field generated by basic residues surrounding Phe-232 is key to the polarization of the PS2 moiety. Thus, disparate polarization and dispersion energies between the PO2 and PS2 complexes contribute critically to the difference in binding affinity. By comparison, easier desolvation of the dithiophosphate group compared to phosphate does not contribute decisively to the observed difference in binding affinity. Favorable polarization and dispersion energies may be a general feature of the dramatic affinity gains seen for complexes between RNAs carrying dithiophosphate groups and their binding proteins.

The (4&prime, &rarr, 6&prime, )-linked DNA homolog 2&prime, 3&prime, dideoxy-&beta, D-glucopyranosyl nucleic acid (dideoxy-glucose nucleic acid or homo-DNA) exhibits stable self-pairing of the Watson&ndash, Crick and reverse-Hoogsteen types, but does not cross-pair with DNA. Molecular modeling and NMR solution studies of homo-DNA duplexes pointed to a conformation that was nearly devoid of a twist and a stacking distance in excess of 4.5 &Aring, By contrast, the crystal structure of the homo-DNA octamer dd(CGAATTCG) revealed a right-handed duplex with average values for helical twist and rise of ca. 15&deg, and 3.8 &Aring, respectively. Other key features of the structure were strongly inclined base-pair and backbone axes in the duplex with concomitant base-pair slide and cross-strand stacking, and the formation of a dimer across a crystallographic dyad with inter-duplex base swapping. To investigate the conformational flexibility of the homo-DNA duplex and a potential influence of lattice interactions on its geometry, we used molecular dynamics (MD) simulations of the crystallographically observed dimer of duplexes and an isolated duplex in the solution state. The dimer of duplexes showed limited conformational flexibility, and key parameters such as helical rise, twist, and base-pair slide exhibited only minor fluctuations. The single duplex was clearly more flexible by comparison and underwent partial unwinding, albeit without significant lengthening. Thus, base stacking was preserved in the isolated duplex and two adenosines extruded from the stack in the dimer of duplexes were reinserted into the duplex and pair with Ts in a Hoogsteen mode. Our results confirmed that efficient stacking in homo-DNA seen in the crystal structure of a dimer of duplexes was maintained in the separate duplex. Therefore, lattice interactions did not account for the different geometries of the homo-DNA duplex in the crystal and earlier models that resembled inclined ladders with large base-pair separations that precluded efficient stacking.

RNA aptamers are synthetic oligonucleotide-based affinity molecules that utilize unique three-dimensional structures for their affinity and specificity to a target such as a protein. They hold the promise of numerous advantages over biologically produced antibodies; however, the binding affinity and specificity of RNA aptamers are often insufficient for successful implementation in diagnostic assays or as therapeutic agents. Strong binding affinity is important to improve the downstream applications. We report here the use of the phosphorodithioate (PS2) substitution on a single nucleotide of RNA aptamers to dramatically improve target binding affinity by ∼1000-fold (from nanomolar to picomolar). An X-ray co-crystal structure of the α-thrombin:PS2-aptamer complex reveals a localized induced-fit rearrangement of the PS2-containing nucleotide which leads to enhanced target interaction. High-level quantum mechanical calculations for model systems that mimic the PS2 moiety and phenylalanine demonstrate that an edge-on interaction between sulfur and the aromatic ring is quite favorable, and also confirm that the sulfur analogs are much more polarizable than the corresponding phosphates. This favorable interaction involving the sulfur atom is likely even more significant in the full aptamer-protein complexes than in the model systems.

We report computer simulation results using the Targeted Molecular Dynamics technique to explore possible transport mechanisms in the multidrug efflux pump AcrB for two substrates, ethidium bromide and a tetrahydropyridine derivative. These studies revealed structural elements, including specific α-helices, β-strands and flexible loops that define a physically plausible pathway for substrates to the extracellular environment. These calculation results can be used to plan future biophysical experiments and may suggest interesting drug design possibilities to address drug resistance due to AcrB function.ImportanceAddressing the issue of antimicrobial resistance mediated by efflux, this study presents possible binding sites and structures in the AcrB MDR pump that could be molecular targets for drugs. Targeted molecular dynamics simulations suggested that these sites and structures seem vital for a successful efflux. The AcrB is proposed to be divided into three distinct zones, with loops, sheets and helices mediating the passage of molecules from one zone to another. We also described possible capture sites on the outer part of the protein and access ways to its interior. Finally, we proposed that ligand competition for same pathways could be thought as an efflux inhibitory mechanism, thus assisting to conceive new ways of designing efflux pump inhibitors.

The cyclooxygenase enzymes (COX-1 and COX-2) are the therapeutic targets of nonsteroidal anti-inflammatory drugs (NSAIDs). Neutralization of the carboxylic acid moiety of the NSAID indomethacin to an ester or amide functionality confers COX-2 selectivity, but the molecular basis for this selectivity has not been completely revealed through mutagenesis studies and/or X-ray crystallographic attempts. We expressed and assayed a number of divergent secondary shell COX-2 active site mutants and found that a COX-2 to COX-1 change at position 472 (Leu in COX-2, Met in COX-1) reduced the potency of enzyme inhibition by a series of COX-2-selective indomethacin amides and esters. In contrast, the potencies of indomethacin, arylacetic acid, propionic acid, and COX-2-selective diarylheterocycle inhibitors were either unaffected or only mildly affected by this mutation. Molecular dynamics simulations revealed identical equilibrium enzyme structures around residue 472; however, calculations indicated that the L472M mutation impacted local low-frequency dynamical COX constriction site motions by stabilizing the active site entrance and slowing constriction site dynamics. Kinetic analysis of inhibitor binding is consistent with the computational findings.

We report a detailed study of a point mutation of the crucial binding site residue, D128, in the biotin-streptavidin complex. The conservative substitution, D128N, preserves the detailed structure observed for the wild-type complex but has an only minimal impact on biotin binding, even though previous experimental and computational studies suggested that a charged D128 residue was crucial for high-affinity binding. These results show clearly that the fundamental basis for streptavidin's extremely strong biotin binding affinity is more complex than assumed and illustrate some of the challenges that may arise when analyzing extremely strong ligand-protein binding interactions.

Molecular modeling studies were performed in order to gain insight into the binding mode and interaction of carborane-containing derivatives of indomethacin methyl ester with the cyclooxygenase-2 (COX-2) isoform, and to assess the predictive capability of the computational tools available for studying carboranes, a unique class of pharmacophores. Docking simulations were able to identify the correct binding mode and reproduced the experimental binding affinity trends with encouraging quality. Nevertheless, the docking results needed to be verified through extensive and resource-intensive quantum chemical calculations, and the interpretation of the theoretical results would not have been straightforward without the supporting experimental data. The inclusion of full receptor and ligand flexibility into the molecular modeling of carborane-containing drug molecules may yield more definitive results, but is currently hindered by the lack of appropriate carborane force field parameters.

Ion mobility-mass spectrometry (IM-MS) allows the separation of ionized molecules based on their charge-to-surface area (IM) and mass-to-charge ratio (MS), respectively. The IM drift time data that is obtained is used to calculate the ion-neutral collision cross section (CCS) of the ionized molecule with the neutral drift gas, which is directly related to the ion conformation and hence molecular size and shape. Studying the conformational landscape of these ionized molecules computationally provides interpretation to delineate the potential structures that these CCS values could represent, or conversely, structural motifs not consistent with the IM data. A challenge in the IM-MS community is the ability to rapidly compute conformations to interpret natural product data, a class of molecules exhibiting a broad range of biological activity. The diversity of biological activity is, in part, related to the unique structural characteristics often observed for natural products. Contemporary approaches to structurally interpret IM-MS data for peptides and proteins typically utilize molecular dynamics (MD) simulations to sample conformational space. However, MD calculations are computationally expensive, they require a force field that accurately describes the molecule of interest, and there is no simple metric that indicates when sufficient conformational sampling has been achieved. Distance geometry is a computationally inexpensive approach that creates conformations based on sampling different pairwise distances between the atoms within the molecule and therefore does not require a force field. Progressively larger distance bounds can be used in distance geometry calculations, providing in principle a strategy to assess when all plausible conformations have been sampled. Our results suggest that distance geometry is a computationally efficient and potentially superior strategy for conformational analysis of natural products to interpret gas-phase CCS data.

Atomic resolution crystallographic studies of streptavidin and its biotin complex have been carried out at 1.03 and 0.95 A, respectively. The wild-type protein crystallized with a tetramer in the asymmetric unit, while the crystals of the biotin complex contained two subunits in the asymmetric unit. Comparison of the six subunits shows the various ways in which the protein accommodates ligand binding and different crystal-packing environments. Conformational variation is found in each of the polypeptide loops connecting the eight strands in the β-­sandwich subunit, but the largest differences are found in the flexible binding loop (residues 45–52). In three of the unliganded subunits the loop is in an ‘open’ conformation, while in the two subunits binding biotin, as well as in one of the unliganded subunits, this loop ‘closes’ over the biotin–binding site. The ‘closed’ loop contributes to the protein’s high affinity for biotin. Analysis of the anisotropic displacement parameters included in the crystallographic models is consistent with the variation found in the loop structures and the view that the dynamic nature of the protein structure contributes to the ability of the protein to bind biotin so tightly.

The adaptor protein ankyrin-R interacts via its membrane binding domain with the cytoplasmic domain of the anion exchange protein (AE1) and via its spectrin binding domain with the spectrin-based membrane skeleton in human erythrocytes. This set of interactions provides a bridge between the lipid bilayer and the membrane skeleton, thereby stabilizing the membrane. Crystal structures for the dimeric cytoplasmic domain of AE1 (cdb3) and for a 12-ankyrin repeat segment (repeats 13–24) from the membrane binding domain of ankyrin-R (AnkD34) have been reported. However, structural data on how these proteins assemble to form a stable complex have not been reported. In the current studies, site-directed spin labeling, in combination with electron paramagnetic resonance (EPR) and double electron-electron resonance, has been utilized to map the binding interfaces of the two proteins in the complex and to obtain inter-protein distance constraints. These data have been utilized to construct a family of structural models that are consistent with the full range of experimental data. These models indicate that an extensive area on the peripheral domain of cdb3 binds to ankyrin repeats 18–20 on the top loop surface of AnkD34 primarily through hydrophobic interactions. This is a previously uncharacterized surface for binding of cdb3 to AnkD34. Because a second dimer of cdb3 is known to bind to ankyrin repeats 7–12 of the membrane binding domain of ankyrin-R, the current models have significant implications regarding the structural nature of a tetrameric form of AE1 that is hypothesized to be involved in binding to full-length ankyrin-R in the erythrocyte membrane.

Sensory adaptation in bacterial chemotaxis is mediated by methylation and demethylation of specific glutamyl residues in the cytoplasmic domain of chemoreceptors. Methylation is catalyzed by methyltransferase CheR. In E. coli and related organisms, methylation sufficiently rapid to be physiologically effective requires a carboxyl terminal pentapeptide sequence on the receptor being modified or, via adaptational assistance, on a neighboring homodimer in a receptor cluster. Pentapeptide-enhanced methylation is thought to be mediated by a approximately 30 residue, potentially disordered sequence that serves as a flexible arm connecting the receptor body and pentapeptide-bound methyltransferase, thus allowing diffusionally restricted enzyme to reach methyl-accepting sites. However, it was not known how many or which sites on the same or neighboring receptors were accessible to the tethered enzyme. We investigated using molecular modeling and found that, in a hexagonal array of trimers of receptor dimers, CheR tethered to a dimer of chemoreceptor Tar by its native 30-residue flexible-arm sequence could reach all methyl-accepting sites on the dimer to which it was tethered plus 48 methyl-accepting sites distributed among nine neighboring dimers, equivalent to the total sites carried by six receptors. This modeling-determined methylation neighborhood of one enzyme-binding dimer and six neighbors corresponds precisely with the experimentally identified neighborhood of seven. Thus, the experimentally observed adaptational assistance can occur by docking of pentapeptide-bound, diffusionally restricted enzyme to methyl-accepting sites on neighboring receptors. Our analysis introduces the notion that physiologically relevant adaptational assistance could occur even if only a subset of sites on a particular receptor are within reach.

We present a 250 ns simulation of the wild-type, biotin-liganded streptavidin tetramer in the solution phase and compare the trajectory to two previously published simulations of the protein in its crystal lattice. By performing both types of simulations, we are able to interpret the protein's behavior in solution in the context of its X-ray structure. We find that the rate of conformational sampling is increased in solution over the lattice environment, although the relevant conformational space in solution is also much larger, as indicated by overall fluctuations in the positions of backbone atoms. We also compare the distributions of chi1 angles sampled by side chains exposed to solvent in the lattice and in the solution phase, obtaining overall good agreement between the distributions obtained in our most rigorous lattice simulation and the crystallographic chi1 angles. We observe changes in the chi1 distributions in the solution phase, and note an apparent progression of the distributions as the environment changes from a tightly packed lattice filled with crystallization media to a bath of pure water. Finally, we examine the interaction of biotin and streptavidin in each simulation, uncovering a possible alternate conformation of the biotin carboxylate tail. We also note that a hydrogen bond observed to break transiently in previous solution-phase simulations is predominantly broken in this much longer solution-phase trajectory; in the lattice simulations, the lattice environment appears to help maintain the hydrogen bond, but more sampling will be needed to confirm whether the simulation model truly gives good agreement with the X-ray data in the lattice simulations. We expect that pairing solution-phase biomolecular simulations with crystal lattice simulations will help to validate simulation models and improve the interpretation of experimentally determined structures.

A 250 ns molecular dynamics simulation of the biotin-liganded streptavidin crystal lattice, including cryoprotectant molecules and crystallization salts, is compared to a 250 ns simulation of the lattice solvated with pure water. The simulation using detailed crystallization conditions preserves the initial X-ray structure better than the simulation using pure water, even though the protein molecules display comparable mobility in either simulation. Atomic fluctuations computed from the simulation with crystallization conditions closely reproduce fluctuations derived from experimental temperature factors (correlation coefficient of 0.88, omitting two N-terminal residues with very high experimental B-factors). In contrast, fluctuations calculated from the simulation with pure water were less accurate, particularly for two of the streptavidin loops exposed to solvent in the crystal lattice. Finally, we obtain good agreement between the water and cryoprotectant densities obtained from the simulated crystallization conditions and the electron density due to solvent molecules in the X-ray structure. Our results suggest that detailed lattice simulations with realistic crystallization conditions can be used to assess potential function parameters, validate simulation protocols, and obtain valuable insights that solution-phase simulations do not easily provide. We anticipate that this will prove to be a powerful strategy for molecular dynamics simulations of biomolecules.

We report a serious problem associated with a number of current implementations of Andersen and Langevin dynamics algorithms. When long simulations are run in many segments, it is sometimes possible to have a repeating sequence of pseudorandom numbers enter the calcuation. We show that, if the sequence repeats rapidly, the resulting artifacts can quickly denature biomolecules and are then easily detectable. However, if the sequence repeats less frequently, the artifacts become subtle and easily overlooked. We derive a formula for the underlying cause of artifacts in the case of the Langevin thermostat, and find it vanishes slowly as the inverse square root of the number of time steps per simulation segment. Numerous examples of simulation artifacts are presented, including dissociation of a tetrameric protein after 110 ns of dynamics, reductions in atomic fluctuations for a small protein in implicit solvent, altered thermodynamic properties of a box of water molecules, and changes in the transition free energies between dihedral angle conformations. Finally, in the case of strong thermocoupling, we link the observed artifacts to previous work in nonlinear dynamics and show that it is possible to drive a 20-residue, implicitly solvated protein into periodic trajectories if the thermostat is not used properly. Our findings should help other investigators re-evaluate simulations that may have been corrupted and obtain more accurate results.

A series of novel derivatives of the nonsteroidal anti-inflammatory drug (NSAID) sulindac sulfide were synthesized as potential agonists of the peroxisome proliferator-activated receptor gamma (PPARgamma). Nonpolar and aromatic substitutions on the benzylidene ring as well as retention of the carboxylic acid side chain were required for optimal activity. Compound 24 was as potent a compound as any other in the series with an EC50 of 0.1 microM for the induction of peroxisome proliferator response element (PPRE)-luciferase activity. Direct binding of compound 24 to PPARgamma was demonstrated by the displacement of [(3)H]troglitazone, a PPARgamma agonist, in a scintillation proximity assay. Compound 24 also stimulated the binding of PPARgamma to a PPRE-containing oligonucleotide and induced expression of liver fatty-acid binding protein (L-FABP) and adipocyte fatty acid-binding protein (aP2), two established PPARgamma target genes. Taken together, these compounds represent potential leads in the development of novel PPARgamma agonists.

A simulated continuous wave electron paramagnetic resonance spectrum of a nitroxide spin label can be obtained from the Fourier transform of a free induction decay. It has been previously shown that the free induction decay can be calculated by solving the time-dependent stochastic Liouville equation for a set of Brownian trajectories defining the rotational dynamics of the label. In this work, a quaternion-based Monte Carlo algorithm has been developed to generate Brownian trajectories describing the global rotational diffusion of a spin-labeled protein. Also, molecular dynamics simulations of two spin-labeled mutants of T4 lysozyme, T4L F153R1, and T4L K65R1 have been used to generate trajectories describing the internal dynamics of the protein and the local dynamics of the spin-label side chain. Trajectories from the molecular dynamics simulations combined with trajectories describing the global rotational diffusion of the protein are used to account for all of the dynamics of a spin-labeled protein. Spectra calculated from these combined trajectories correspond well to the experimental spectra for the buried site T4L F153R1 and the helix surface site T4L K65R1. This work provides a framework to further explore the modeling of the dynamics of the spin-label side chain in the wide variety of labeling environments encountered in site-directed spin labeling studies.

The genomically encoded fosfomycin resistance protein from Pseudomonas aeruginosa (FosAPA) utilizes Mn(II) and K+ to catalyze the addition of glutathione (GSH) to C1 of the antibiotic rendering it inactive. Although this protein has been structurally and kinetically characterized with respect to the substrate, fosfomycin, questions remain regarding how the enzyme binds the thiol substrate, GSH. Computational studies have revealed a potential GSH binding site in FosAPA that involves six electrostatic or hydrogen-bonding interactions with protein side-chains as well as six additional residues that contribute van der Waals interactions. A strategically placed tyrosine residue, Y39, appears to be involved in the ionization of GSH during catalysis. The Y39F mutant exhibits a 13-fold reduction of catalytic activity (kcat = 14 ± 2 s−1), suggesting a role in the ionization of GSH. Mutation of five other residues (W34, Q36, S50, K90, and R93) implicated in ionic of hydrogen-bonding interactions resulted in enzymes with reduced catalytic efficiency, affinity for GSH, or both. The mutant enzymes were also found to be less effective resistant proteins in the biological context of Escherichia coli. The more conservative W34H mutant has native-like catalytic efficiency suggesting that the imidazole NH group can replace the indole group of W34 that is important for GSH binding. In the absence of co-crystal structural data with the thiol substrate, these results provide important insights into the role of GSH in catalysis.

Equinatoxin II (EqtII) is a soluble, 20 kDa pore-forming protein toxin isolated from the sea anemone Actinia equina. Although pore formation has long been known to occur in distinct stages, including monomeric attachment to phospholipid membranes followed by detachment of the N-terminal helical domain and oligomerization into the final pore assembly, atomistic-level detail of the protein-lipid interactions underlying these events remains elusive. Using high-resolution solution state NMR of uniformly-15N-labeled EqtII at the critical micelle concentration of dodecylphosphocholine, we have mapped the lipid-binding site through chemical shift perturbations. Subsequent docking of an EqtII monomer onto a dodecylphosphocholine micelle, followed by 400 ns of all-atom molecular dynamics simulation, saw several high-occupancy lipid-binding pockets stabilized by cation-π, hydrogen bonding, and hydrophobic interactions; and stabilization of the loop housing the conserved arginine-glycine-aspartate motif. Additional simulation of EqtII with an N-acetyl sphingomyelin micelle, for which high-resolution NMR data cannot be obtained due to aggregate formation, revealed that sphingomyelin specificity might occur via hydrogen bonding to the 3-OH and 2-NH groups unique to the ceramide backbone by side chains of D109 and Y113; and main chains of P81 and W112. Furthermore, a binding pocket formed by K30, K77, and P81, proximate to the hinge region of the N-terminal helix, was identified and may be implicated in triggering pore formation.

The two cyclooxygenase enzymes, COX-1 and COX-2, are responsible for the committed step in prostaglandin biosynthesis, and are the targets of the non-steroidal anti-inflammatory drugs aspirin, ibuprofen and the COX-2 selective inhibitors, Celebrex™, Vioxx™ and Bextra™. The enzymes are remarkable in that they catalyze two dioxygenations and two cyclizations of the native substrate, arachidonic acid, with near absolute regio- and stereoselectivity. Several theories have been advanced to explain the nature of enzymatic control over this series of reactions, including suggestions of steric shielding and oxygen channeling. As proposed here, selective radical trapping and spin localization in the substrate-derived pentadienyl radical intermediate can also be envisioned. Herein we describe the results of explicit, 10 ns molecular dynamics simulations of both COX-1 and COX-2 with the substrate-derived pentadienyl radical intermediate bound in the active site. The enzymes’ influence on the conformation of the pentadienyl radical was investigated, along with the accessible space above and below the radical plane, and the width of several channels to the active site that could function as access routes for molecular oxygen. Additional simulations demonstrated the extent of molecular oxygen mobility within the active site. The results suggest that spin localization is unlikely to play a role in enzymatic control of this reaction. Instead, a combination of oxygen channeling, steric shielding and selective radical trapping appears to be responsible. This work adds a dynamic perspective to the strong foundation of static structural data available for these enzymes.

Insulin-dependent diabetes mellitus (IDDM) is positively associated with HLA-DQ8, DQ2, and DQ6 (B*0604) and negatively with DQ6 (B*0602). The mechanisms by which the DQ molecules control the development of IDDM is not known. DQ6 (B*0602) and DQ6 (B*0604) molecules share the same DQalpha chain but differ in the beta chain by six residues at positions 9, 30, 57, 70, 86, and 87. The aim of the study was to sequence the peptides eluted from both DQ6 molecules and to determine the binding motifs and construct peptides for docking them into the DQ6 peptide binding groove by molecular modeling. EBV transformed B cell line homozygous for DQ6 and hybridoma cell line secreting DQ6 specific antibody were grown in large-scale culture. The DQ6 molecules were precipitated with solid-phase bound antibodies specific for DQ6. The dissociation of peptides from MHC was done with ultrafiltration and separation of peptides by reversed-phase HPLC, using Edman degradation. A special application of Edman degradation is pool sequencing. This approach allowed us to determine common characteristics of all peptides associated with a given MHC molecule. The precipitation of DQ6 molecules and the peptide elution were done successfully. The sequencing of the peptides from DQ6 (B*0602) identified three fractions: (1) IINEPTAAAIAYGLD (Bovine HSP70), (2) IINEPTAAAIAGLDR (Human HSP70), and (3) NPRDAKACVVHGSDLK (Na+/K+ ATPase). Peptide eluted from DQ6 (B*0604) had a sequence ADLFRGTLDPVEK with sequence homology to HSP70 (residues 307-319). We were able to predict the motifs for DQ6 from the ligands eluted. We used molecular modeling as a tool to identify plausible binding motifs for peptides. Our studies show that peptide ADLFRGTLDPVEK and NPRDAKACVVHGSDLK fit well in the respective DQ6 binding grooves. These predicted motifs should then be useful for screening of autoantigens associated with diabetes and identifying the epitopes that are likely to interact with T cells.

The chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2) is a G protein-coupled receptor that mediates the pro-inflammatory effects of prostaglandin D(2) (PGD(2)) generated in allergic inflammation. The CRTH2 receptor shares greatest sequence similarity with chemoattractant receptors compared with prostanoid receptors. To investigate the structural determinants of CRTH2 ligand binding, we performed site-directed mutagenesis of putative mCRTH2 ligand-binding residues, and we evaluated mutant receptor ligand binding and functional properties. Substitution of alanine at each of three residues in the transmembrane (TM) helical domains (His-106, TM III; Lys-209, TM V; and Glu-268, TM VI) and one in extracellular loop II (Arg-178) decreased PGD(2) binding affinity, suggesting that these residues play a role in binding PGD(2). In contrast, the H106A and E268A mutants bound indomethacin, a nonsteroidal anti-inflammatory drug, with an affinity similar to the wild-type receptor. HEK293 cells expressing the H106A, K209A, and E268A mutants displayed reduced inhibition of intracellular cAMP and chemotaxis in response to PGD(2), whereas the H106A and E268A mutants had functional responses to indomethacin similar to the wild-type receptor. Binding of PGE(2) by the E268A mutant was enhanced compared with the wild-type receptor, suggesting that Glu-268 plays a role in determining prostanoid ligand selectivity. Replacement of Tyr-261 with phenylalanine did not affect PGD(2) binding but decreased the binding affinity for indomethacin. These results provided the first details of the ligand binding pocket of an eicosanoid-binding chemoattractant receptor.

We have used molecular modeling studies and molecular dynamics simulations to generate three-dimensional models for cyclooxygenase-1 complexes with a series of indomethacin ethanolamide derivatives. These studies provide a plausible explanation for the stereoselective ligand binding preferences observed experimentally for these inhibitors and predict the general binding mode as well as specific structural details for the ligand-enzyme complexes. These studies provide insight into the nature of cyclooxygenase-1 interactions with a series of novel inhibitors and should help increase our understanding of key structural determinants for cyclooxygenase isozyme-selective inhibitor binding.

Fluorescence resonance energy transfer is a powerful biophysical technique used to analyze the structure of membrane proteins. Here, we used this tool to determine the distances between a distinct position within a docked agonist and a series of distinct sites within the intramembranous confluence of helices and extracellular loops of the cholecystokinin (CCK) receptor. Pseudo-wild-type CCK receptor constructs having single reactive cysteine residues inserted into each of these sites were developed. The experimental strategy included the use of the full agonist, Alexa488-CCK, bound to these receptors as donor, with Alexa568 covalently bound to the specific sites within the CCK receptor as acceptor. Site-labeling was achieved by derivatization of intact cells with a novel fluorescent methanethiosulfonate reagent. A high degree of spectral overlap was observed between receptor-bound donor and receptor-derivatized acceptors, with no transfer observed for a series of controls representing saturation of the receptor binding site with nonfluorescent ligand and use of a null-reactive CCK receptor construct. The measured distances between the fluorophore within the docked agonist and the sites within the first (residue 102) and third (residue 341) extracellular loops of the receptor were shorter than those directed to the second loop (residue 204) or to intramembranous helix two (residue 94). These distances were accommodated well within a refined molecular model of the CCK-occupied receptor that is fully consistent with all existing structure-activity and photoaffinity-labeling studies. This approach provides the initial insights into the conformation of extracellular loop regions of this receptor and establishes clear differences from analogous loops in the rhodopsin crystal structure.

Distinct spatial approximations between residues within the secretin pharmacophore and its receptor can provide important constraints for modeling this agonist-receptor complex. We previously used a series of probes incorporating photolabile residues into positions 6, 12, 13, 14, 18, 22, and 26 of the 27-residue peptide and demonstrated that each covalently labeled a site within the receptor amino terminus. Although supporting a critical role of this domain for ligand binding, it does not explain the molecular mechanism of receptor activation. Here, we developed probes having photolabile residues at the amino terminus of secretin to explore possible approximations with a different receptor domain. The first probe incorporated a photolabile p-benzoyl-l-phenylalanine into the position of His(1) of rat secretin ([Bpa(1),Tyr(10)]secretin-27). Because His(1) is critical for function, we also positioned a photolabile Bpa as an amino-terminal extension, in positions -1 (rat [Bpa(-1),Tyr(10)]secretin-27) and -2 (rat [Bpa(-2),Gly(-1),Tyr(10)]secretin-27). Each analog was shown to be a full agonist, stimulating cAMP accumulation in receptor-bearing Chinese hamster ovary-SecR cells in a concentration-dependent manner, with the position -2 probe being most potent. They bound specifically and saturably, although the position 1 analog had lowest affinity, and all were able to label the receptor efficiently. Sequential specific cleavage, purification, and sequencing demonstrated that the sites of covalent attachment for each probe were high within the sixth transmembrane segment. This suggests that secretin binding may exert tension between the receptor amino terminus and the transmembrane domain to elicit a conformational change effecting receptor activation.

Photoaffinity labeling of receptors by bound agonists can provide important spatial constraints for molecular modeling of activated receptor complexes. Secretin is a 27-residue peptide hormone with a diffuse pharmacophoric domain that binds to the secretin receptor, a prototypic member of the Class B family of G protein-coupled receptors. In this work, we have developed, characterized, and applied two new photolabile probes for this receptor, with sites for covalent attachment in peptide positions 12 and 14, surrounding the previously most informative site of affinity labeling of this receptor. The [Tyr10,(BzBz)Lys12]rat secretin-27 probe covalently labeled receptor residue Val6, whereas the [Tyr10,(BzBz)Lys14]rat secretin-27 probe labeled receptor residue Pro38. When combined with previous photoaffinity labeling data, there are now seven independent sets of constraints distributed throughout the peptide and receptor amino-terminal domain that can be used together to generate a new molecular model of the ligand-occupied secretin receptor. The aminoterminal domain of this receptor presented a stable platform for peptide ligand interaction, with the amino terminus of the peptide hormone extended toward the transmembrane helix domain of the receptor. This provides clear insights into the molecular basis of natural ligand binding and supplies testable hypotheses regarding the molecular basis of activation of this receptor.

Numerous techniques have been used to elucidate the structural basis for interaction of cholecystokinin (CCK)-related peptides with their hormone-binding receptor, the CCK-A receptor (CCK-AR), including structure-activity relationship studies, site-directed mutagenesis, photoaffinity-labeling, and solution NMR analysis of both CCK peptide ligands and peptide fragments derived from the CCK-A receptor. Different structural models have been developed for the peptide-receptor complexes using various subsets of the available experimental data (Giragossian & Mierke 2001; Ding et al. 2002; Escrieut et al. 2002). Here, we review details of the various models and evaluate the impact of selected experimental data sets on model development.

The receptors for cholecystokinin (CCK) peptides are guanine nucleotide-binding protein-coupled receptors in the rhodopsin/beta-adrenergic receptor family. The molecular basis of natural ligand binding to the type A CCK receptor has been studied using ligand structure-activity series, receptor mutagenesis, and photoaffinity labeling studies. These have focused attention on the extracellular loop and tail domains, with the most direct insights coming from intrinsic photoaffinity labeling studies. A model of the binding of CCK to this receptor is consistent with all these studies. This model places the carboxyl terminus of CCK adjacent to the amino-terminal tail outside of transmembrane segment 1, the mid-region of the peptide adjacent to the third extracellular loop outside of transmembrane segment 7, and includes a charge-charge interaction between peptide residue tyrosine-sulfate 27 and the arginine residue in the second extracellular loop of the receptor in position 197.

The amino terminus of the secretin receptor (SecR) is known to be critical for natural agonist action, although the role it plays is still unclear. We have demonstrated that photolabile residues within both the amino-terminal (position 6) and carboxyl-terminal (positions 22 and 26) halves of secretin each covalently label receptor amino-terminal tail residues [Dong et al., J Biol Chem, 274:19161-19167 (1999), 274:903-909 (1999), and 275:26032-26039 (2000)]. Here, we extend this series of studies with an additional probe having its site of covalent attachment in a distinct region of the peptide, between amino- and carboxyl-terminal helical domains. This probe incorporated a photolabile (epsilon-p-benzoylbenzoyl)lysine in position 18 and a site for oxidative radioiodination [(tyrosine(10),(benzoyl-benzoyl)lysine(18))rat secretin-27]. This analog represented a full agonist, stimulating cAMP accumulation in Chinese hamster ovary-SecR cells in a concentration-dependent manner. It bound to the SecR specifically and saturably, and was able to efficiently label that molecule within its amino terminus. Sequential specific cleavage, purification, and sequencing demonstrated that this probe labeled receptor residue arginine(14), in the same subdomain as that labeled by previous probes. Consistent with the importance of this residue, alanine replacement mutagenesis (R14A) resulted in substantial reductions in the potency (127-fold) and binding affinity (400-fold) of secretin relative to its action at the wild-type receptor. We have been able to accommodate all four extant pairs of residue-residue approximations between divergent regions of the secretin pharmacophore and the first forty residues of the SecR into a credible molecular model of this interaction. Additional experimentally derived constraints will be necessary to determine the spatial positioning of this complex with the remainder of the SecR.

We report here specialized functions incorporated recently in the rigid-body docking software toolkit TagDock to utilize electron paramagnetic resonance derived (EPR-derived) interresidue distance measurements and spin-label accessibility data. The TagDock package extensions include a custom methanethiosulfonate spin label rotamer library to enable explicit, all-atom spin-label side-chain modeling and scripts to evaluate spin-label surface accessibility. These software enhancements enable us to better utilize the biophysical data routinely available from various spin-labeling experiments. To illustrate the power and utility of these tools, we report the refinement of an ankyrin:CDB3 complex model that exhibits much improved agreement with the EPR distance measurements, compared to model structures published previously.

The high affinity energetics in the streptavidin-biotin system provide an excellent model system for studying how proteins balance enthalpic and entropic components to generate an impressive overall free energy for ligand binding. We review here concerted site-directed mutagenesis, biophysical, and computational studies of aromatic and hydrogen bonding interaction energetics between streptavidin and biotin. These results also have provided insight into how streptavidin builds a large activation barrier to dissociation by managing the enthalpic and entropic activation components. Finally, we review recent studies of the biotin dissociation pathway that address the fundamental question of how ligands exit protein binding pockets.