Hendrickson Laboratory

Research in the Hendrickson Laboratory

Most research projects of the Hendrickson laboratory aim for an atomic-level understanding of the biochemical action of biological macromolecules. This work entails the production and characterization of proteins and protein complexes from systems of interest, x-ray crystallographic analysis of relevant structures, and structure-inspired biochemical and computational studies of the biochemical and biophysical properties. We also engage in relevant methodology development. Nearly all of the projects have crystal production as a central and often limiting activity.

      Current activities emphasize membrane receptors as related to transmembrane signal transduction, viral proteins especially as related to HIV infection, molecular chaperones in relation to protein folding and aggregation, membrane proteins as studied by structural genomics, diffraction methods and synchrotron radiation, and investigations on biophysical principles.

      Although our research is at a very basic level, many of the studies have biomedical relevance. Notably, our work on HIV gp120 relates quite immediately to vaccine and drug development, our work on serotonin receptors addresses the neurobiology of mood disorders, our studies on the FSH receptor system relate directly to reproductive biology, and our work on Hsp70 chaperones is relevant to cancer and to neurodegenerative diseases. In another direction, our recent work on anion channels that control stomatal opening in plant leaves has relevance in environmental research.

Membrane Receptors and Cellular Signaling   

An important emphasis of our research concerns the initial phases of cellular signal transduction, including the biochemical and biophysical aspects of signal transduction across membranes by major signaling systems (Hendrickson, 2005). In most cases, the signal-initiating stimulus from the environment is chemical; it may be a small compound, a macromolecular hormone or growth factor, or even another cell. Nearly always, receptors embedded in the cellular membrane mediate transmission of signaling into the cell. Our interest lies in the mechanisms by which biochemical signals are transduced across the membrane. We concentrate on the integral membrane receptor proteins but relevant extra-membranous portions are also studied. We have current efforts on several receptor classes:

G-protein coupled receptors (GPCRs). We commit a substantial effort toward understanding the conformational states of the seven-transmembrane (7TM) receptors that function through stimulation of guanine nucleotide exchange in the Gα subunit of a heterotrimeric G protein. The major focus is on serotonin receptors and on glycoprotein hormone receptors, but we also work on other GPCRs including the chemokine receptors involved in HIV infection. Much of the effort to date has focused on producing intact molecules for crystallization, including the characterization of dimers (Mancia et al., 2008). This effort entails complexes with antibodies, which we can prepare (Mancia et al., 2007) and produce as recombinant Fab particles (Assur et al., 2007). We determined the crystal structure of the complex between human follicle stimulating hormone (FSH) and the hormone-binding portion of the human FSH receptor (FSHR_HB) (Fan & Hendrickson, 2005), and we are now producing full-length LH and FSH receptor in complexes with their respective glycoprotein hormones.

Histidine kinase receptors. Typical histidine kinase receptors have a sensor domain separated by two transmembrane helices and followed by a cytoplasmic portion that contains the kinase domain and a histidine autophosphorylation site. The sensor domains are highly variable, specific to different ligands; whereas the cytoplasmic portions are more conserved, especially so for the kinase domains. Using a structural genomics approach, we are systematically pursuing representative structures for sensor domain families that we have characterized. Several sensor structures have been solved (PhoQ: Cheung et al., 2008; DcuS and DctB: Cheung & Hendrickson, 2008; NarX: Cheung & Hendrickson, 2009; PhoQ: HK29: Cheung et al., 2009; TorS: Moore & Hendrickson, 2009; HK1: Zhang & Hendrickson, 2010), and several more are in progress. We have also completed structures on cytoplasmic domains, including one of the entire cytoplasmic dimer (Marina et al., 2005), comprising both its coiled-coil dimerization domain and its kinase domains. Taken together, these structural results are inspiring testable hypotheses about mechanisms for signal transduction in this system. Besides continued work on the system of sensor domains, our current focus is on producing active, intact histidine kinase receptors for crystallization.

Ion channel receptors. In parallel with our work on GPCRs, we also study two types of ligand-gated ion channels. The first of these are Cys-loop ion channels where we have recently completed a study of four ligands bound to the extracellular domain of the chicken &alpha9 acetylcholine receptor. These structures provide important insights into conformational changes that mediate ligand gating of the ion channel opening. We are also engaged in efforts to produce and crystallize intact ion channels, focusing on heteropentameric acetylcholine receptors and the 5HT3 cys-loop serotonin receptors from the mammalian nervous system. The second system of interest concerns the ryanodine-sensitive calcium-release channel known as the ryanodine receptor. Working in collaboration with the laboratory of Andrew Marks, we have obtained nicely diffracting crystals of the intact 2.3MDa receptor and are working to improve the diffraction and to perfect methods for phase determination.

Viral Proteins and HIV Infection   

HIV envelope interactions. The foundation of our work on interactions of the HIV envelope proteins with cellular receptors lies in structures of complexes between HIV gp120 and both its the cellular receptor CD4 and a neutralizing antibody bound to the co-receptor binding site. These were determined both for a laboratory adapted R4 strain, HxBc2 (Kwong et al., 1998), and for a primary R5 isolate, Yu2 (Kwong et al., 2000), and in each case CD4 was represented by the D1D2 binding fragment and the antibody component was the human 17b Fab fragment. We subsequently carried out studies on the thermodynamics of these interactions (Myszka et al., 2000; Kwong et al., 2002), and we have determined a number of additional structures in ongoing collaboration with Peter Kwong, including complexes with CD4 mimetics (Huang et al., 2005) recent work focuses on the development of antagonists of the gp120-CD4 interaction. Toward this end, we devised a chemical design for derivatives of F43C CD4 (D1D2) in which cysteine adducts bind into the Phe43 interfacial cavity (Xie et al., 2007). We have determined four structures of such complexes, all in the HxBc2 lattice. These structures provide details of interactions that are being used to design and synthesize new compounds with the ultimate aim of obtaining useful HIV entry inhibitors. More recently, we have obtained crystals of a small molecule compound, NBD556, and variants of this compound in complex with gp120 and CD4-induced antibody Fab fragment.

Other viral proteins. We also have programs directed at other viral proteins, including other HIV envelope receptors. In particular, we have analyzed structures of the dendritic-cell receptor DC-SIGN, a tetrameric C-type lectin that attaches to sites of glycosylation on HIV gp120 and also on Dengue virus. In collaboration with Michael Rossmann, we earlier showed how the lectin domain binds to Dengue virus (Pokidysheva et al., 2006) and now with the tetrameric stalk model in hand we provide further definition. Finally, in collaboration with Steve Goff, we have determined two crystal structures of the capsid protein from Moloney murine leukemia virus (MMLV), a retrovirus analogous with HIV, that correspond respectively to the hexagonal lattices of the immature and the mature viral capsid.

Molecular Chaperones and Protein Folding   

Hsp70 chaperones. The 70kD family of heat shock protein (Hsp70) chaperones is ubiquitous, having involvement in diverse activities in all organisms. Others had previously characterized the ATPase domain of Hsp70s and we previously analyzed the structure of the substrate-binding portion of the bacterial Hsp70 DnaK in association with a high-affinity peptide (Zhu et al., 1996). The nature of the allosteric interaction between the ATPase substrate-binding units in the chaperone cycle of bindings and releases has long remained elusive, however. We recently obtained a first structure of an Hsp70 relative showing such interactions (Liu & Hendrickson, 2007). This is the structure of yeast Sse1, an Hsp110 family member and, from the structure, a clear relative of Hsp70s. The structure shows a remarkable change in conformation relative to that in domains from Hsp70s. A battery of interface mutations in Sse1 and its DnaK homolog, tested in yeast and bacteria, respectively, informed us about general modes of conformational change and ATPase action. In vitro biochemical tests of several of the DnaK mutants have inspired a new model for the chaperone cycle and the generation of mutant-stabilized ATP states that have succumbed to crystallization. A rich battery of additional crystallographic projects is in prospect, including complexes with co-chaperones.

Other molecular chaperones. Molecular chaperones play vital roles in protein folding, in the suppression of aggregation in states of cellular stress, and specialized roles in protein trafficking. We have recently made substantial progress in the biochemical and structural characterization of four classes of molecular chaperones. These include the important bacterial chaperone trigger factor and an FKBP-based chaperone from Methanococcus. In both cases the crystallography is persistently frustrated by limited diffraction even though crystals are large, a feature that seems to be associated with intrinsic flexibility in these multidomain proteins. The best of our trigger factor structures is stabilized in a complex with ribosomal protein S7 (Martinez-Hackert & Hendrickson, 2009). Our third class of interest is in Boca/MESD, a specialized chaperones that mediates the folding of a of LDL receptors and related proteins. We have analyzied crystal structures of the Boca/MESD chaperones from three species.

Membrane Proteins and Structural Genomics   

We are among the principal participants in the New York Consortium on Membrane Protein Structure (NYCOMPS), which was a Specialized Center in phase 2 of the NIGMS-supported Protein Structure Initiative (PSI-2) and a Membrane Protein Center in the current PSI-Biology phase. We aim to develop and apply more effective methods for structural analysis of membrane proteins. The NYCOMPS pipeline has recently become highly productive (Punta et al., 2009; Love et al., 2010) and our group participates actively in this effort. We are particularly engaged in the structure determination component of the project and we have brought several NYCOMPS proteins into our laboratory for crystallization and structural analysis. We also participate with other NYCOMPS groups in other structural analyses and technology development. One of these targets, recently solved to 1.15Å resolution, proves to be an anion channel that is homologous to a channel that controls stomatal closure in plant leaves in response to darkness and to significant environmental factors including drought, high CO2 levels or high ozone levels (Chen et al., in press).

Diffraction Methods and Synchrotron Radiation

Our laboratory has been engaged in the development of methods for diffraction analysis of biological structure for a long time. Early contributions include widely used phasing coefficients (Hendrickson & Lattman, 1970), the initiation of stereochemically restrained refinement of crystal structures (Hendrickson & Konnert, 1980; Konnert & Hendrickson, 1980), and the structural analysis of crambin based solely on anomalous scattering from sulfur atoms (Hendrickson & Teeter, 1981). The crambin structural analysis established what is now known as the single-wavelength anomalous diffraction (SAD) method and paved the way for his development of the multi-wavelength anomalous diffraction (MAD) method (Hendrickson, 1985; Hendrickson et al., 1988). Broad utility of the MAD method followed when we recognized that selenium could serve as a rich source for the required diffraction signals (Hendrickson et al., 1989) and that selenomethionine (SeMet) could be substituted readily for the natural amino acid methionine (Hendrickson et al., 1990; Yang et al., 1990). We tested MAD phasing in applications at synchrotrons around the world, and with HHMI support we developed NSLS beamline X4A at Brookhaven National Laboratory to optimize the MAD experiment (Staudenmann et al., 1989). X4A became highly productive, and it remains so this day. Subsequently, MAD beamlines were emulated around the world, and MAD and SAD methods now dominate in biological crystallography, producing many hundreds of new structures each year.

      We honed anomalous scattering methods in applications to tough problems at the forefront of biology and medicine. Among these are the following: (1) MAD phasing was used to determine the initial structure of the human CD4 (Ryu et al., 1990) and SeMet CD4 produced in mammalian cells was used to determine the full-length structure (Wu et al., 1997). CD4 is a T-cell co-receptor for the cellular immune response, but it is also used by the AIDS virus to initiate infection and later structures were determined in complexes with the HIV envelope glycoprotein gp120. These studies revealed atomic details that are providing insights for vaccine development and inhibitor design. (2) MAD phasing was also used to determine the first protein tyrosine kinase structures, those of the human insulin receptor (Hubbard et al., 1994) and of human lymphocyte kinase Lck (Yamaguchi & Hendrickson, 1996). Besides generating deep insights into diabetes and the cellular immune response, these analyses also paved the way for a new generation of cancer therapeutics. (3) Another early application was to the reproductive hormone human chorionic gonatoropin (hCG) which was determined by SeMet MAD phasing (Wu et al., 1994). Another reproductive hormone, human follicle-stimulating hormone (FSH) was later analyzed in complex with its receptor (Fan et al., 2005). FSH is used therapeutically to treat infertility, and the hormone-receptor complex provided lead concepts both for enhancing fertility and for contraception. (4) Finally, the structures of Hsp70 molecular chaperones determined by SeMet MAD phasing (Zhu et al., 1996; Liu & Hendrickson, 2007) have provided mechanistic insights of relevance for activities that protect against neurodegenerative diseases such as Alzheimer's and Parkinson's. We have further advanced MAD and SAD phasing methods in applications to dozens of other structural problems, many of which have provided significant insight into human biology.

      Synchrotron radiation is essential for effective MAD experiments and it greatly enhances SAD experiments. We developed the X4 beamlines at the National Synchrotron Light Source (NSLS), now supported by the New York Structural Biology Center (NYSBC), to test phasing methods, and this work continues with enhancements in x-ray optics (Lidestri & Hendrickson, 2007) and in applications. We are also in the midst of developing the NYSBC Microdiffraction Beamline at NSLS-II, which is under construction. This new beamline will exploit a novel design to provide exceptional energy resolution to optimize anomalous signals. We are also engaged in other approaches for optimizing MAD and SAD experiments. One direction of recent effort is focused on extracting weak anomalous signals from membrane proteins and large complexes, which oftentimes diffract only to levels corresponding to low resolution. Methods are being developed for the use of multi-crystal SAD and MAD experiments, by which the limitations from radiation damage are mitigated. Another direction relates to enhancing MAD signals through the incorporation of many additional methionine residues into Fab fragments or by the attachment of gold nanoparticles through introduced cysteine residues.

Biophysical Principles

We have longstanding interests in studies on the biophysical principles of protein structure, dynamics and evolution. Based largely on our early analyses of restrained refinements of crystallographic B factors (Konnert & Hendrickson, 1980; Hendrickson & Konnert, 1980; Hendrickson, 1989), we performed several analyses on the extractions of information about protein dynamics from atomic mobility parameters (B factors). Examples include a molecular dynamics comparison with crystal structure (Yu et al., 1985), an analysis of epitope antigenicity (Tainer et al., 1984), a study of structural heterogeneity in crystallized proteins (Smith et al., 1986), and a study of disorder by x-ray restrained molecular dynamics (Kuriyan et al., 1991). We have also performed analyses on molecular evolution (Aronson et al., 1994; Shapiro et al., 1995) and in comparative analyses of molecular structure, both routinely in our structure descriptions and more generally across families (Fan & Hendrickson, 2008; Cheung & Hendrickson, 2010). Most recently, we have developed a robust computational procedure for computing realistic pathways for large-scale conformational transitions (Korkut & Hendrickson, 2009; Korkut & Hendrickson, 2009).