David Eisenberg - US grants

It is well established that the amino acid sequence of a protein determines how the protein folds, and whether the protein is destined to be water soluble or membrane-bound, and to which small molecules or other macromolecules the protein binds. However, the atomic basis of these events of molecular recognition is not yet understood well enough to permit many useful predictions from the knowledge of the amino acid sequence alone. This is unfortunate, because many advances in practical problems, such as drug and protein design, await further progress in understanding these relationships. This proposal takes three directions in addressing this central problem: (1) Crystal structure studies of small amphiphilic proteins and of designed proteins, to elucidate rules of protein folding, and rules for association of proteins with lipids and other apolar surfaces: (2) Computations of energies of protein-folding and protein-lipid association, using atomic solvation parameters, semi-empirical quantities that represent an atom-based solvation (or hydrophobic) free energy; and (3) Computational analysis of amino acid sequences, using extensions of the PROFILE method which is a way of representing information about families amino acid sequences in proteins.*** //

Five investigators seek program support for structural studies of macromolecules and molecular interactions related to the etiology and control of AIDS. The program has three main objectives: 1. Determination of molecular structures (Eiserling/Eisenberg/Dickerson/Rees). To provide the fundamental structural information needed for rational drug design, structures will be determined for: the HIV envelope glycoprotein and its complexes with monoclonal antibodies and the T4 (CD4) lymphocyte receptor (Eiserling/Eisenberg/Rees), and the HIV reverse transcriptase (Dickerson). Biological materials will be supplied from the laboratories of Dr. L. Lasky (Genentech) and Dr. L. Hood (Caltech). To insure the availability of crystalline specimens for structural studies, the Core facilities for cystallization and coordinated electron microscopy and diffraction will be enhanced. To permit more rapid collection of x-ray data, diffraction equipment will be improved. 2. Molecular recognition and drug design (Dickerson/Sigman/Eisenberg/Rees). Principles of DNA-drug interactions will be developed (Dickerson). Designs for sequence-specific DNA-cutting orthophenanthrolines will be extended and applied to HIV (Sigman). Characterizations of molecular surfaces by their atomic solvation parameters and surface roughness will be extended and applied to drug binding (Rees/Eisenberg). To assure biological relevance of the work, the following scientists have agreed to consult or to collaborate as the work proceeds: Dr. James Paulson (UCLA; glycoproteins); Dr. L. Lasky (Genentech; HIV genes and proteins); Dr. L. Hood (Caltech; proteins of the immune system); Dr. O. Witte and Dr. R. Gaynor (UCLA; virology); Dr. J. Feigon (UCLA; NMR and intermolecular interactions). To permit increased emphasis on characterization of interactions and drug design, facilities for computation and graphics will be improved. 3. Training (Core and all five principal laboratories). A major effort will be undertaken to provide training in principles and practice of structural determination and molecular interactions for graduate students and postdoctorals. Three one quarter graduate courses are currently offered (by the five investigators) that provide fundamentals of biological structure and interactions.

Eisenberg 9420769 Once proteins can be designed to carry out specific functions, we will be on the way to the pharmaceuticals of the future. The fundamental methods of this research are aimed at making protein design a practical science. Genetic engineering has opened the way to designing new proteins with diverse uses: catalysts, drugs, specific toxins, scavengers, signals, and biomaterials. But crucial to these designs is control of 3-D structure. Protein function depends on structure. We must be able to design animo acid sequences that fold into predictable three dimensional structures. This project aims to define the problem of the connection of amino acid sequence to 3D structure for two classes of small, -helical proteins. These are bundles of -helices, and -helical coiled- coils, an important special class of -helical bundles. For these two classes of proteins, amino acid sequences have been designed to fold and oligomerize in predetermined ways. The main principle of folding and assembly used in these designs is the removal of apolar sidechains from contact with water, and the exposure of charged and polar side chains. This principle of segregation according to polarity may be termed amphiphilicity. The main analytical method used is x-ray structure determination of the designed proteins. Designed proteins are synthesized (or expressed); purified, crystallized, and structures determined as the basis for the next turn of the design cycle. One of the main design tools used is semi-empirical computational energetics. A major extension of such methods is termed Profile Refinement. In Profile Refinement, the conformation of a protein is computed by permitting the protein structure to change in such a way that each amino acid residue, or atomic group, seeks an environment within the protein that is most like its average environment in the database of known protein structures. Profile Refinement is based on principles of amphiphilicity, rather than of c lassical energetics. Nevertheless, Profile Refinement can be combined with classical energetics, perhaps to create a more powerful method for assessing protein designs, and for improving partially correct protein models, derived from low-resolution structure determinations or homology models. Among designed proteins whose structures will be determined is a coiled-coil that is a domain swapped dimer. That is, two tripled coiled-coils from a dimer by exchanging a helix. The ability to design domain-swapped dimers will simplify methods for designing larger oligomeric proteins. %%% This project is aimed at making protein design a practical science. Protein function depends on structure. this research is focussed on designing amino acid sequences for proteins that fold into predictable three-dimensional structures. Two classes of protein will be designed: bundles of - helices, and -helical coiled-coils. The proteins will be synthesized (or genetically expressed), crystallized and the structures determined. The results will then be used for a second turn of the design cycle. The ultimate objective is the design of proteins to carry out specific functions. ***

structural biology; membrane proteins; protein structure function; glutamate ammonia ligase; diphtheria toxin; receptor binding; protein engineering; enzyme mechanism; pheromone; drug design /synthesis /production; antineoplastics; membrane permeability; crosslink; antiinfective agents; antitubercular agents; computer program /software; enzyme inhibitors; enzyme substrate complex; intermolecular interaction; bacterial proteins; pore forming protein; crystallization; computer simulation; X ray crystallography; nuclear magnetic resonance spectroscopy;

structural biology; biomedical facility;

"Regulography"- Quantitative Reconstruction of Transcriptional Regulatory Networks
James C. Liao, Vwani Roychowdhury, Chiara Sabatti, David Eisenberg, Fuyuhiko Tamanoi, University of California, Los Angeles

Goal and Tasks: The goal of this project is to reconstruct the hidden structure and dynamics of transcriptional regulatory networks based on massive gene expression data (generated from DNA microarray) and regulatory models under the constraints of various ancillary information, such as protein interactions with DNA, other proteins, and RNA. The ultimate outcome of the project will be an integrated framework (incorporating both methodology and the required databases) for deducing transcriptional network structure and dynamic, which can be applied to define the signal transduction pathways perturbed by unknown drug effects, toxic compound challenges, and mutations. We are using Saccharomyces cerevisiae as the model eukaryote for verification of our paradigm. Toward this end, we are pursuing the following specific tasks: (1) Develop an analytical and IT-based framework for quantitative and dynamic reconstructions of various pre-translational regulatory networks. This includes transcriptional regulation (as governed by protein-DNA and protein-protein interactions), and processing, transport, and stability regulation of mRNA (as governed by various protein-mRNA interactions). This work builds on a novel system-reconstruction methodology called Network Component Analysis. (2) Acquire and organize data for protein-DNA, protein-protein interaction, protein-RNA interactions, mRNA stability, and gene-expression noise. (3) Experimental verification of the paradigm of the framework developed in (1) and (2). (4) Disseminate the results through a composite "regulographic" database.
Team Organization: The PIs of this project come from five different fields: Dr. Liao is a chemical/biochemical engineer specialized in metabolic engineering and DNA microarray analysis, Dr. Roychowdrury is an electrical engineer/computer scientist specialized in networks and systems theories, Dr. Tamamoi is a microbiologist specialized in yeast genetics, Dr. Eisenberg is a renowned biochemist/crystallographer specialized in protein structure, function, and protein-protein interactions, and Dr.Sabatti is a statistician specialized in genetic and microarray analyses.
Broader Impact: In addition to the fundamental scientific discoveries proposed here, we are pursuing the following information technology related activities: (1) As part of the interdisciplinary bioinformatics program at UCLA, we plan to develop a two-part course on intracellular regulatory networks (targeted toward upper-level undergraduate students and beginning graduate students in both engineering and biological fields) . (2) We will leverage the organizational infrastructure of a number of interdisciplinary research institutes on UCLA campus (including, the NSF Institute for Pure and Applied Mathematics, the California Nano-Science Institute, and the Institute for Cell Mimetic Space Exploration) to attract both undergraduate and graduate minority students. (3) We will integrate our experimental and analytical results into a dynamic database, and make it available to the larger community, which we believe will spur and aid research efforts on intracellular process modeling at a much larger scale, involving academic and commercial institutions nation wide.

Intellectual Merit:
Structural and computational methods will be applied to understand why some proteins undergo transitions from their soluble, globular structures to form insoluble, elongated fibrils. In previous work, we have shown that an essential feature of fibril structure is that the protein contains a short self-complementary sequence of amino acids. We developed a computer algorithm that accurately identifies such segments, and we have validated the algorithm by experiments, which show that the identified segments do in fact form fibrils. Our goal now is to determine the universe of fibril-forming proteins, which we term the amylome.

Broader Impacts:
The wider implications of the work include the following: (1) Protein fibrils can form the basis of new materials. (2) A UCLA course in Structural Molecular Biology is linked to this research. In the laboratory section, students learn hands-on protein structure determination. Graduates of this course now teach and carry out research at some 30 universities, colleges, and research institutes. UCLA has perhaps the most diverse student body of any university, and our course reflects the demography of the campus. We plan to recruit students to this course, and through it, to the field of structural molecular biology, from a wide range of backgrounds, particularly those from underrepresented minority groups.

The Database of Interacting Proteins (DIP) is the original and one of the most extensive databases of information on interacting proteins. It was founded on the assumption that macromoleQular interactions are at the basis of all biomedical science, and that researchers in many branches of biology and medicine can benefit from access to an accurate repository of what is known of protein interactions. To continue its expansion, DIP will adopt a procedure of direct deposition of experimental data by primary authors. Accuracy of interaction data being essential, methods for the assessment of data quality will be extended and applied to DIP, to define a Core set of highly reliable interaction data. The Core interactions will permit construction of networks of interacting proteins, useful to cell and systems biologists. From its inception, DIP has been guided by a program of research into methods for extracting knowledge of networks from genomic and proteomic data. This research emphasis will be will be deepened by introducing new methods for extraction of information on protein networks from genomic and proteomic data, including logical triplet analysis of genomic and microarray data. LiveDIP is the extension of DIP that concentrates on the interactions among proteins in defined states. A second extension of DIP, ProLinks, catalogs functional linkages between pairs of proteins inferred from genomic and proteomic data;it identifies pathways and complexes, and will serve as a feeder to DIP, as its inferred linkages are validated as true interactions. To provide faster and more reliable access to DIP, its IT infrastructure will be upgraded.

[unreadable] DESCRIPTION (provided by applicant): The aggregation of proteins in Alzheimer's, Parkinson's, prion diseases, Huntington's, and other amyloid-like diseases will be illuminated by methods of structural and computational biology. Background research by microcrystallography has shown that the fundamental structural unit of amyloid-like fibrils is a set of beta sheets, in which the amino acid sidechains of neighboring sheets intermesh, in what is termed a steric zipper. The steric zipper interface between the faces of the sheets is completely dry. The protein segments that form the sheets are as short as 4-8 residues in length, stacking either in parallel or antiparallel to grow a fibril, but the segments can be longer and some proteins contain several such segments. To learn the structures of amyloid fibrils from disease-associated proteins, the same methods of microcrystallography will be applied to microcrystals grown from short segments of the A? and Tau proteins of Alzheimer's disease, from the PrP protein of the prion diseases, from ?-synuclein of Parkinson's disease, and from proteins involved in ALS and diabetes type 2. To learn what happens during fibril formation to the rest of the protein, structural studies will also be conducted on larger segments and entire fibril-forming proteins, using novel methods of crystal screening and microcrystallography. Preliminary work shows that computational energetics can identify which segments from proteins are those that are fibril-forming and can be grown into microcrystals, suitable for structural determination. This procedure is based on the 3D Profile method for finding sequences that fit a given fold (in this case the steric zipper), using energy functions. The procedure will be extended and applied to amyloids. Once a segment has been discovered which forms fibrils, and its structure has been determined by crystallography as belonging to the steric-zipper type of architecture, the connection between the segment and fibrils of the full protein can be assessed by whether the segment can seed the full protein into fibrils. Further proof of the connection of the segment and fibrils of the full protein can be obtained by mutating residues in the protein that correspond to residues of the segment, and looking for diminished fibrillization. These structures, derived by novel methods of microcrystallography, are the first high-resolution (up to 0.85 A resolution), fully refined atomic structures for the amyloid state. They show that there are at least 4 basic patterns for the steric zipper spines of amyloid fibrils, and perhaps up to 7 such patterns. These structures offer a solid foundation on which to devise diagnostics and therapeutics for these devastating neurodegenerative diseases. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): The Database of Interacting Proteins (DIP) is the original and one of the most extensive databases of information on interacting proteins. It was founded on the assumption that macromolecular interactions are at the basis of all biomedical science, and that researchers in many branches of biology and medicine can benefit from access to an accurate repository, documenting what is known of protein interactions. DIP has forged the International Molecular Interaction Exchange Consortium (IMEx) with the two principal European interaction databases to cooperate in curation of literature information on interactions and with the SwissProt bioinformatics group. IMEx has established formats and standards for data curation, now endorsed by HUPO (the Human Proteome Organization). To continue its expansion, DIP with its IMEx partners and journal editors, is developing a procedure for direct deposition of experimental data by primary authors. Accuracy of interaction data being essential, the DIP methods for the assessment of data quality will be extended to define a Core set of highly reliable interaction data for each organism. The Core interactions will permit construction of networks of interacting proteins, useful to cell and systems biologists. From its inception, DIP has been guided by a program of research into methods for extracting knowledge of networks from genomic and proteomic data and methods for annotation of protein function. This research has resulted in the ProLinks database of functional linkages between proteins, the ProKnow server for annotating "unknown" protein sequences and structures, the method of Generalized Functional Linkages (GFL) for supplying functional annotations to proteins based on their networks of interactions. New depositions to DIP will be used to evaluate the effectiveness of inferred annotations from ProKnow and ProLinks. Thus, the challenge of curating DIP stimulates discovery of new methods which in turn enhance the quality and usefulness of DIP data, which in turn provide validation for the new methods. In the new grant period, the inferred linkages and annotations will be combined with DIP to offer researchers "one stop information" about the interactions and functional context of their proteins of interest. Every interaction, both experimental and inferred linkages, will be accompanied by its uncertainty value. A new Web interface is currently being implemented to ease user access to all this information. This groundwork should allow DIP to supply information on hundreds of thousands of interactions in the coming years. PUBLIC HEALTH RELEVANCE: The Database of Interacting Proteins (DIP) documents the discoveries of tens of thousands of scientists, described in thousands of research articles, about the interactions of proteins that form the basis of all biomedical science. DIP is a free and open community resource for research physicians and scientists. DIP takes special care in documenting the accuracy of the data it contains.

PROJECT SUMMARY (See instructions): Structural and biochemical methods will be used to characterize amyloid-Uke fibrils of SODl and its segments. The overall goals are to understand the process of fibrillation of SODl in vitro and in cells, and to determine atomic structures for the fibril-defining segments of SODl. This informatin will be used in the process of structure-based design, to create inhibitors of fibrillation of SODl and its mutants. These inhibitors can be lead compounds for drugs against SODl fibrillation and possibly fALS. A high-risk goal is to characterize the structure of SODl aggregates that form in human cells [see Project 3). In particular, micro-X-ray diffraction will be used to assess the possibility that SODl is in the amyloid state in cells.

DESCRIPTION (provided by applicant): Treatments for systemic amyloid diseases have been held back by lack of information on the structures and causes of aggregation of the disease agents. These agents are the elongated, unbranched amyloid fibers formed by proteins having propensity for aggregation. The fibers accumulate in organs which eventually fail. In contrast, the increasingly successful attack on cancer, infectious, and metabolic diseases is rooted in part in the availability of information on the structures of the disease targets, permitting design of effective chemical interventions. In previous work we have developed a procedure for inhibiting the formation of amyloid fibers. This first step is application of our computer algorithm which identifies the short Velcro-like sequence segments that drive formation of amyloid fibers. We have applied this algorithm to find over 100 such segments in disease- related proteins, and have verified that such segments themselves form amyloid fibers, and closely related microcrystals. The second step is X-ray structure determination of these microcrystals, which reveal the atomic basis of fiber formation. The third step is to use the resulting atomic structure as a platform for the design of inhibitors to stop fiber formation. This overall procedure is robus and ready to produce inhibitors of fibers found that cause light-chain (AL) and transthyretin (TTR) systemic amyloidosis. In our research, we will validate particular segments of immunoglobulin light chains and transthyretin as the causes of aggregation, and based on their atomic structures, design inhibitors of aggregation. These inhibitors will be tested for their abilty to halt fiber formation in vitro, and in animal models. In principle, the same methods can be used to develop inhibitors for other systemic amyloid diseases.

Our hypothesis is that the lack of drugs to halt Alzheimer?s disease stems in large part from ignorance of the structures of the most pertinent drug targets: the aggregated forms of the proteins tau and beta-amyloid. Here we propose to extend our studies of the structures of amyloid fibrils and oligomers to enable structure-based design of inhibiting compounds. For each of our proposed projects, structure determination will be followed by structure-based design of one or more inhibitors. Then each inhibitor will be assayed for effectiveness in inhibiting aggregate formation in vitro and inhibiting toxicity in cell models. The most effective inhibitors will then be assessed in animal models of our collaborators. Among our projects are the following: (1) Structure determination of oligomers of tau that can seed spreading of tau from cell to cell, and subsequent design of an inhibitor of oligomerization; (2) Inhibitor design of tau aggregation based on our newly determined structure of the amyloid-forming segment of tau with sequence VQIINK; (3) Evolution by ribosome display of inhibiting single-domain antibodies against tau aggregates, with the possibility that these may penetrate the blood-brain- barrier; (4) Optimization of existing crystals of the 20 residue segment of beta-amyloid with sequence GKLVFFGENVGSNKGAIIGL, which seems to form an oligomer. Improved crystals will be followed by structure determination and inhibitor design; (5) Crystallization of beta-amyloid or its segments in a lipid environment to gain possible insight into its toxic function; (6) Structure determination of a segment of beta-amyloid bound to its putative cell-surface receptor, followed by inhibitor design; (7) Exploration of the action of our newly discovered segment of the protein transthyretin which breaks up oligomers of beta-amyloid and inhibits toxicity. Each of these projects, if successful, opens a path to a possible therapeutic agent against Alzheimer?s disease. These paths have not been previously available because the pertinent structures have been unknown. We find the principal barrier to determination of amyloid structures is the miniscule size of the crystals. We propose to surmount this barrier by further exploitation of advanced methods of electron diffraction.

Under a variety of stressful conditions, cells develop small compartments that are transient. These compartments are termed dynamic intracellular bodies. Little is known about the molecular interactions that form these bodies. Many of these bodies contain RNA-protein complexes that could protect cellular RNAs during cell stress. The goal of this research is to illuminate the interactions between proteins and of proteins with RNAs that lead to the formation of these bodies. This will be achieved using a variety of analyses using structural tools such as X-Ray crystallography and electron microscopy backed up by biochemical studies. While the research is of a fundamental nature, some of the proteins that are involved in the formation of these dynamic intracellular bodies are also associated in aggregated form with Amyotrophic Lateral Sclerosis (ALS), so the structural studies may provide the additional benefit of understanding the connection of disease. A part of the research is also focused on developing novel methods for structure determination, such as micro-electron diffraction. The broader societal significance involves the recruitment into scientific careers of students and postdoctoral researchers from underrepresented groups. Additionally, through the UCLA Center for Global Mentoring, the project aims to spread the American system of scientific mentoring to countries with developing scientific infrastructures.

To fill the void of knowledge about the molecular interactions in dynamic intracellular bodies, also known as membraneless organelles, the project will discover the structural basis of a newly recognized form of protein-protein and protein-RNA interaction. In Aim 1, it has been discovered that the atomic structures of the adhesive segments of reversible amyloid-like fibrils (RALFs) formed by the low complexity sequences have both similarities and differences with disease-related amyloid. The project will explore the full variety of interactions and assembly states found in membraneless organelles. In other words, the human reversible amylome will be mapped. In Aim 2, it is hypothesized that RNA binding can strongly influence membraneless organelle and reversible amyloid formation through several parallel mechanisms. These mechanisms will be tested using biochemical and cellular assays. The contributions of RALFs and protein-RNA interactions will also be compared. Both Aims will deepen understanding of cellular organization and function, particularly mechanisms for sequestration of proteins and RNAs without the aid of membrane encapsulation. Part of the research involves determination of structures from miniscule amyloid-like crystals. The frontier method of micro-electron diffraction must be further developed.

DESCRIPTION (provided by applicant): Treatments for systemic amyloid diseases have been held back by lack of information on the structures and causes of aggregation of the disease agents. These agents are the elongated, unbranched amyloid fibers formed by proteins having propensity for aggregation. The fibers accumulate in organs which eventually fail. In contrast, the increasingly successful attack on cancer, infectious, and metabolic diseases is rooted in part in the availability of information on the structures of the disease targets, permitting design of effective chemical interventions. In previous work we have developed a procedure for inhibiting the formation of amyloid fibers. This first step is application of our computer algorithm which identifies the short Velcro-like sequence segments that drive formation of amyloid fibers. We have applied this algorithm to find over 100 such segments in disease- related proteins, and have verified that such segments themselves form amyloid fibers, and closely related microcrystals. The second step is X-ray structure determination of these microcrystals, which reveal the atomic basis of fiber formation. The third step is to use the resulting atomic structure as a platform for the design of inhibitors to stop fiber formation. This overall procedure is robus and ready to produce inhibitors of fibers found that cause light-chain (AL) and transthyretin (TTR) systemic amyloidosis. In our research, we will validate particular segments of immunoglobulin light chains and transthyretin as the causes of aggregation, and based on their atomic structures, design inhibitors of aggregation. These inhibitors will be tested for their abilty to halt fiber formation in vitro, and in animal models. In principle, the same methods can be used to develop inhibitors for other systemic amyloid diseases.

PROJECT SUMMARY/ABSTRACT The successful treatment of conditions such as cancer, metabolic disorders, and infectious diseases has been enabled by a detailed understanding of the structures of the disease-causing agents, permitting design of effective chemical interventions. In contrast, a lack of structural information about disease targets has impeded the development of therapies for systemic amyloid diseases. The targets in these diseases are protein aggregates called amyloid fibrils; they accumulate in body organs and lead to eventual organ failure. Amyloid fibrils are difficult to study by traditional structure-determination methods, but our recent efforts have yielded significant results. We have identified and determined the amyloid structures of the aggregation-driving segments from two disease-causing proteins: transthyretin, which causes transthyretin amyloidosis (ATTR), and immunoglobulin light chain, which causes light-chain amyloidosis (AL). Our transthyretin amyloid structures have enabled us to design effective inhibitors of transthyretin aggregation; they cap the ends of the developing amyloid fibrils, preventing their growth. In the new grant period, we aim to (1) further our understanding of amyloid structure, and (2) apply this understanding to the development of new and better candidate therapeutics and diagnostics. In Aim 1, we propose to determine near-atomic resolution structures of ATTR and AL patient-derived amyloid fibrils, using cryo-electron microscopy. We will use these structures to design and further optimize amyloid- capping inhibitors for both transthyretin and immunoglobulin light chains, testing their effectiveness on isolated proteins and fibrils. In Aim 3, we will evaluate the transthyretin inhibitors in worm and mouse models of ATTR, as the next steps in moving these potential therapies toward use in humans (clinical trials). In Aim 2, we address the present difficulty of diagnosing and monitoring systemic amyloid diseases. We take advantage of the specific binding of our fibril-capping inhibitors to amyloid fibrils to develop a safe, non- radioactive diagnostic for ATTR. Our inhibitors will be coupled to injectable magnetic nanoparticles that are detectable by magnetic resonance imaging (MRI). Thus, once injected into patients, the inhibitors will bind the nanoparticles to amyloid fibrils in organs, permitting detection of pathogenic aggregates by MRI. In Aim 4, we will capitalize on the natural protective role of transthyretin against fibril formation and toxicity of ?-amyloid, linked to Alzheimer's disease. We have identified the minimum segment of transthyretin that confers such protection and we now aim to assess its potential as a treatment of Alzheimer's in a mouse model. Our discovery of effective chemical interventions for amyloid diseases could prevent suffering and extend the lives of millions of Americans afflicted with these fatal conditions. Moreover, our proposed diagnostic tool will enable doctors and researchers to evaluate the usefulness of emerging therapies. The same methods we propose here may be eventually used to develop inhibitors and diagnostics for other amyloid diseases.

TRD 1. Dedicated sample preparation for MicroED ? Eisenberg (Lead) Summary Sample preparation for microcrystal electron diffraction (MicroED) is challenging and it is a rate-limiting step. Here, we will develop methods for growing nanocrystals, optimizing them, and work out a protocol for reproducible grid preparation for MicroED. Currently, we prepare samples for MicroED using single particle cryoEM protocols. This involves pipetting the sample solution onto an EM grid, blotting the excess, and freezing the sample by plunging the grid into liquid ethane. This process is harsh on the sample, particularly that blotting can expose it to the water-air interface compromising its structure and collapsing the underlying crystal lattice. For membrane proteins the problem is exacerbated by the growth of ?soft? crystals of protein surrounded by lipids or detergents. Moreover, growing crystals in lipidic cubic phase is even more challenging as they embed into a lipid matrix that is viscous and thus are almost impossible to blot without damaging the crystals. Understanding which crystallization and sample preparation approaches are applicable to MicroED requires a consorted and systematic effort challenged by and tested on select biological projects. Here we propose to systematically test conditions for nanocrystal growth and determine which procedures yield nanocrystals of the highest quality. We will establish new strategies for nanocrystal cryo protection and develop nanocrystal gwoth and preservation kits. Finally, we will decide, through a series of experiments, what are the best practices for FIB milling to serve MicroED as a strategy for preparing nanocrystals of membrane proteins grown in lipidic cubic phase. We will achieve these goals through three aims: 1. Directed nanocrystal growth and detection; 2. Strategies for nanocrystal cryo protection; 3. Strategies for grid preparation MicroED for experiments. Overall, we will deliver reproducible and reliable procedures for sample preparation, detection, and preservation for MicroED including membrane proteins grown in LCP. The long-term goal is to enable routine and high- throughput crystallization and structure determination by MicroED.