This page contains the text of the original programme proposal as presented to the Foundation for Strategic Research in December, 1994 (minus the appendices).
Note that the actual programme differs from the proposal in details (and in one third of the amount of money).
G.J. Kleywegt (scientific secretary)
2.1 Definition of the field
2.2 History and status of Structural Biology in Sweden
2.3 Analysis of strengths and weaknesses today
2.3.1 Challenges for the future
2.3.2 Protein crystallography
2.3.3 Protein NMR
2.4 Philosophy, objectives and strategy
2.4.2 Objectives of the programme
2.4.3 Suggested strategy
2.5 Summary of the programme
2.6 Expected results and spin-offs
2.7 Interactions with other programmes
3. Project plan
3.1.3 Karolinska Institute
3.1.4 Lund University
3.1.5 Stockholm University
3.1.6 Swedish NMR Centre
4.5 Budget, financing and commitments
5. Miscellaneous issues
5.1 International collaborations
5.2 Industrial involvement
5.3 Career perspectives
5.4 Evaluation and (dis-) continuation of the programme
Structural Biology research will be central in all biological sciences to explain biological function and activity in terms of structure and chemistry. The aim of research in Structural Biology is to elucidate the three-dimensional structure and dynamic properties of biological macromolecules (proteins, nucleic acids, and complexes) at atomic resolution. Therefore, Structural Biology research is essential for several other biology-oriented programmes, that are expected to be initiated by the Foundation for Strategic Research. Structural Biology has already had a major impact in many areas (including medical and pharmaceutical science, food technology, forestry and plant technology) in most of the developed world. Swedish industry, however, lags behind by ~5 years in this process, even though some of the academic groups in Sweden are competing at the highest international level.
In order to strengthen the strategic value of Structural
Biology in Sweden, we propose a four-tiered approach:
(1) To reinforce excellence; by supporting new strategic research efforts of the groups in Uppsala, Lund and Stockholm that are competing at the highest international level.
(2) To remedy current weaknesses; in particular, by supplying each of the top groups with dedicated protein expression and/or labelling facilities.
(3) To elevate graduate training to a world-class level; by initiating a Swedish Structural Biology Network, which will organise advanced courses and conferences open to all students and researchers with an interest in Structural Biology.
(4) To facilitate knowledge transfer; by stimulating contacts between academic groups, contacts with industry, and participation of industrial researchers in the Network.
This proposal outlines how this approach can be implemented
as a programme of the Foundation for Strategic Research
during a five-year period (1995-1999). The total cost
amounts to 100 million Swedish crowns. In order to
be able to respond to new developments, and to direct
efforts towards other programmes funded by the Foundation
for Strategic Research, 10 % of this budget is to be allocated
after the Network has been in operation for two years.
Capital investments are limited to only ~12.5 % of
the budget, 7.5 % will be spent on improved education,
and approximately one percent is required to manage
the Network. The rest of the budget is used for research
purposes, including salary and running costs for 13
new scientist positions, 21 graduate students, and
3 new positions at the level of engineer.
We have been asked to prepare a proposal concerning a strategic research programme in Structural Biology. The Foundation for Strategic Research has suggested that the programme should create a national Network, should include structural techniques (including modelling) as well as expression techniques, and should be confined to proteins and nucleic acids.
2.3.1 Challenges for the future
Without a doubt, Structural Biology research has become a central tool in most biological research areas. The main reason is that the techniques allow for the determination of atomic resolution structures of relevant biomolecules. These structures often reveal key features that explain the biological response in chemical terms. In addition to an increased knowledge, such understanding has the potential to generate a vast number of industrial applications, e.g. in rational drug design for the pharmaceutical industry.
Today, Structural Biology is a strong field in Sweden. However, there are a number of challenges in the next ten years to strengthen the strategic value of structural biological research in Sweden, both within the national research environment and in society:
1. Most importantly, the atomic-resolution techniques must strengthen their interfaces to the various national research programmes in biology. Typically, it takes from months to years to solve atomic-resolution structures. Therefore, all such efforts must be considered early in strategic research programmes, and relevant molecules must be produced and purified. The proposed Swedish Structural Biology Network must be utilised in part to create organised connections with other national biology research networks, institutes and centres (see section 2.7). In addition, the Structural Biology groups (both in protein crystallography and NMR spectroscopy) must build their own support groups for the production of recombinant proteins. In this way, the necessary production and purification of gene products that are targets for structural analysis can be carried out near the structural groups.
2. Another challenge will be to transfer Structural Biology technology and trained personnel to relevant applications in the industrial sector. To date, only corporations in Sweden with pharmaceutical interests are utilising Structural Biology techniques in-house (Pharmacia and Symbicom). The needs of the pharmaceutical industry will expand, and in the future, we predict, many more applications linked to biotechnology will find new use for macromolecular structures, e.g. in improving enzymes for the paper industry, in ligand design for different diagnostic purposes, and in the food industry.
3. A third future challenge will be to strengthen the technology platform in Structural Biology; to maintain and further improve the technological and methodological strongholds that exist in the field in Sweden today, to spread state-of-the-art techniques to the smaller groups, and to develop leading-edge technologies in new areas that are considered to be important for biology and industry, such as structural analysis of membrane-spanning proteins.
These challenges are addressed in the present programme in Structural Biology.
2.3.2 Protein crystallography
The major stronghold consists of the combined groups of Uppsala University and the Swedish University of Agricultural Sciences. Taken separately, the Uppsala groups are much smaller than the best American groups. This follows, in part, from the traditions and form of the Swedish granting system and the large increase in funding for Structural Biology in the USA from the Howard Hughes Foundation. Taken together, however, Uppsala is a centre of excellence within Europe, and in some areas is unique. The main weaknesses we perceive are a lack of access to the latest protein-expression systems, lack of expertise in studying integral membrane proteins, and the lack of in-house NMR expertise. The latter may be remedied in the near future since Uppsala University has committed itself to setting up a biomolecular NMR research unit. With financial support from the Wallenberg Foundation, one or two 600 MHz spectrometers will be acquired, and a position at the lecturer level will soon be announced. Therefore, this area has not yet been included for support by the committee. The first two shortcomings, as well as the issue of size, will be addressed in this application. To maintain competitiveness over the duration of the programme, some improvements in the existing equipment (data collection, computing, graphics) are also required.
On a smaller scale, the group at Lund also operates at an internationally competitive level. The opening of a dedicated beamline at the MAX II synchrotron in Lund will be very beneficial for maintaining this position. At the present time, there is a severe shortage of synchrotron radiation beam time in Europe. Access to much larger periods of beam time will, therefore, be a major advantage for Swedish groups. This will require the purchase of suitable area-detector equipment and suitable computers to process the data. This beamline may not be suitable for multi-wavelength anomalous dispersion (MAD) phasing (developed by Wayne Hendrickson and co-workers), but fortunately a dedicated MAD beamline (BL19) is being developed at ESRF, Grenoble. We plan to develop expertise in this new phasing method by organising workshops within the planned Network.
Structural Biology is a recent enterprise at Stockholm University, but it has had an excellent start. Both this group and the one in Gothenburg are small and they will, therefore, greatly benefit from interactions within the Network.
2.3.3 Protein NMR
Biomolecular NMR spectroscopy is a much younger technique, but nevertheless there are at least two groups in Sweden who are able to compete at the highest international level (at Lund University and the Karolinska Institute).
A major limitation to the NMR structural work has been the virtual absence of dedicated laboratories for protein expression and isotopic labelling. Isotope labelling of proteins (uniform as well as specific, with 15N and 13C) has become an indispensable technique in NMR spectroscopy, for the structure determination of larger proteins, for studies of large protein complexes, and for studies of protein dynamics. Specific labelling may require the use of specifically labelled amino acids, as well as auxotrophic strains of the producing micro-organism. Whereas several prokaryotic expression systems presently are well documented and adequate for the production of a large body of proteins, they are unsuitable for the production of glycosylated eukaryotic proteins or proteins with post-translationally modified amino acids.
Experience from a large number of biophysical laboratories involved in structure-function studies of proteins strongly indicate that expression and labelling facilities should be closely integrated with the research activities. Again, it has been the hard-learned lesson in many biophysical laboratories that samples delivered to the laboratory have been far from suitable for NMR studies. We therefore propose that funds be allocated to support expression/preparation facilities at both the Karolinska Institute and Lund University. The facility at Lund University will be dedicated to the preparation of recombinant proteins with post-translationally modified amino acids.
The committee has considered and discussed several possible approaches to setting up a strategic research programme within the present framework of Structural Biology in Sweden. There are many ways in which one can try to foster excellent research, for example:
* initiating one or two major research efforts aimed at solving a major biological problem that is typically too large to be tackled by any individual group. This is the way in which the human genome project works, for instance.
* making major investments in capital equipment in order to establish large-scale facilities (for example, a dedicated synchrotron beamline for all Swedish crystallographers at ESRF in Grenoble, or a 750 MHz spectrometer for all Swedish NMR spectroscopists).
* reinforcing groups with a proven track record by injecting extra funds without requiring a major realignment of their research efforts. This may be best described as curiosity research in areas of strategic interest. Of course, a precondition is that there are excellent groups already; if not, one would have to try and attract researchers from elsewhere.
There exist other possibilities (a "mini-Research Council", the "inverse cheese-slicer" approach, or reinforcing groups which haven't yet received world-class status), but these were dismissed unanimously from the onset. The committee decided that preference should be given to the third alternative, since this has the greatest chance of producing excellent work. A precondition is that these groups also commit themselves to making methodological contributions to their field, and that they disseminate knowledge and expertise to other research groups in the country.
Simultaneously, graduate education can and should be improved substantially in order to guarantee continuity in the future. The committee proposes to initiate a national Network which will organise courses for all Swedish graduates in particular areas of Structural Biology (with foreign as well as Swedish tutors), and an annual conference at which the students would be able to present and defend their work.
2.4.2 Objectives of the programme
The committee feels that the main objective of its proposed programme should be to guarantee a continued world-class status for Swedish Structural Biology research. The area has become exponentially more important in the past decade, both because of basic scientific interest and curiosity, and because of the many potential industrial applications.
The transfer of Structural Biology from an area of academic interest to a field with important commercial applications is now far advanced in Europe, Japan and the USA. Foreign industry has relied almost totally on academia to produce their structural biologists. Many of the new industrial laboratories are competitive at the very highest international levels and are an active part of the Structural Biology community. Swedish industry lags behind most developed countries by at least 5 years in the build-up of their structural groups. Two Swedish companies have invested in setting up their own structural groups to date. Symbicom has a fully equipped X-ray laboratory close to the Biomedical Centre in Uppsala, and there is close interaction with the academic groups. Pharmacia has more recently invested in both an X-ray and an NMR group. A large part of this group was educated in either Uppsala (Lundqvist, Sundström, Kraulis) or Lund (Kördel). We expect this trend to continue with a transfer of educated structural biologists from academia to industry. We will encourage industrial groups to take an active part in the setting up and running of the Network.
We plan to strengthen our curiosity research in the following strategic areas: membrane proteins, receptor-ligand interaction, protein design, development of computational methods to study enzyme catalysis and drug design. This should be of major interest to the pharmaceutical industry and to those involved in the industrial applications of enzymes.
2.4.3 Suggested strategy
The committee proposes to follow a four-tiered approach to reach its objective:
* reinforcement of excellent research groups. These are groups with proven track records at the highest international level, adequate infra-structure, not dependent on a single excellent person, and firmly embedded in an academic setting.
* new efforts to try and remedy perceived current weaknesses. This means in particular, that each of the top groups should get a dedicated laboratory for protein expression and (in the case of NMR) for the labelling of proteins.
* to improve graduate student training in Structural Biology to bring it to a world-class level. This can be implemented through a Swedish Structural Biology Network. All research groups would be encouraged to join this Network, so that as many students as possible may benefit from it. The programme must be flexible enough to respond to changes in the quality landscape of Swedish Structural Biology, and to reward new initiatives which are perceived to be (or become) excellent. We will strive to ensure a close interaction with other Networks supported by the Foundation for Strategic Research that will result in new structural projects.
* to stimulate contacts with industry to facilitate the knowledge transfer that is needed to keep Swedish companies on an equal footing with their foreign competitors.
* Reinforcing excellence. Since the top groups in Sweden are quite well equipped with measuring instruments, and have easy access to national and international facilities, it seems unnecessary to make major capital investments. Therefore, the total investment in new equipment constitutes less than 13 % of the proposed budget (see Appendix I). Instead, the committee would prefer to invest predominantly in "human capital", plus defrayment of the costs of small laboratory equipment, chemicals and other running costs. Rather than applying the "inverse cheese slicer" principle (i.e., dividing resources equally, regardless of track records), the committee prefers to invest mainly in those groups that have already demonstrated that they are capable of operating at the highest international level, both with respect to skills and choice of research subjects. Naturally, some flexibility to adapt to unforeseen developments has been built into the programme. Support to the groups should come in the form of graduate students, research assistant positions ("Forskar Assistent" or "Forskare"), laboratory technicians ("Laboratorie Assistent"), and research engineers ("Forskningsingenjör"). It is important to note that all research groups in Sweden will benefit from this investment, even those that do not receive direct support, due to the transfer of competence and knowledge within an active and open Network. Naturally, the "excellent" groups are expected to make significant contributions to the methodology of their respective areas of expertise, and to disseminate their expertise nationally and internationally. Substantial new funds will be used to stimulate research in areas which can be foreseen to become increasingly important in the next few years (such as membrane-bound proteins, receptors, protein dynamics and thermophilic enzymes). In order to address relevant biological questions, the Structural Biology Network should initiate very close interactions with other national biological programmes that are initiated within the Strategic Research programme (see section 2.7).
* Remedying weaknesses. The committee does not feel that any of the important areas of Structural Biology are absent in Sweden, which implies that no major new research efforts need to be initiated at the professorial level. On the other hand, dedicated protein expression and labelling facilities are virtually absent in academia. While this may be understandable to some extent (since this largely involves service-type activities, which give little scientific credit), it also means that the choice of subjects that can be studied may be somewhat ad hoc, depending on the availability of material, and personal networks. In particular with respect to the most prestigious projects, structural biologists are critically dependent upon usable quantities of sufficiently pure material. In such cases, the laboratories that have dedicated expression (and labelling) laboratories usually have the edge. The proposal therefore includes the establishment of new protein expression and/or labelling laboratories at each of the three major sites (Uppsala, Lund, KI).
* Guaranteeing continuity. In order to maintain an internationally competitive role, training of graduate students should be at the highest level. At present, there is little cooperation and coordination between the individual universities. The committee proposes to initiate a graduate Network which would organise advanced courses (with international as well as local tutors), and an annual conference at which the students will present and defend their research. In addition, each student who is funded through the programme will have two mentors from other institutes who will annually discuss and assess the progress made by the student. Participation in the Network is open to all Structural Biology groups in Sweden, so that even the groups that don't receive direct financial support will benefit.
The areas of research are grouped according to their geographical location. Each project has a principal investigator/coordinator (PI), who is responsible for scientific leadership within the project. There already exists some overlap of interest and techniques between these projects, which will not be discouraged in the future. More details on the scientific background and current research interests of the PIs, a list of recent scientific publications of each group, and information about other issues (international collaborations, organisations, etc.) can be found in Appendix II. In this section, only short and general descriptions of the proposed research projects are given.
* Receptors of Strategic Importance.
Allocation: 1 student, 2 research assistants.
Molecular recognition is a fundamental process in biology. The existing active programme in this field will be further expanded. To date, most of the structural information produced by X-ray crystallography and NMR, has been obtained on soluble globular proteins. Integral membrane proteins are also of great interest both scientifically and industrially; many receptors that are of interest to the pharmaceutical industry fall into this class. A serious research programme will be started to elucidate the structures of integral membrane proteins and receptors that are of interest to Swedish companies, for instance ion channels and/or pumps, G-protein coupled receptors, substrate transporters, etc.
The second project is to investigate molecules engaged in cellular signalling systems that are involved in normal and cancer cell growth. Platelet-derived growth factor (PDGF) is a major mitogen for connective tissue. The PDGF isoforms exert their cellular effects by binding with different specificities to two related tyrosine kinase receptors. Ligand binding induces receptor dimerisation, which leads to receptor autophosphorylation and initiation of intracellular signalling. The most important goal for future studies will be to obtain structural information about the PDGF/PDGF-receptor complex. This would allow the design of specific PDGF antagonists. Three-dimensional information on the intracellular domain of the receptor will be important for the design of specific inhibitors of the kinase activity and the interaction with down-stream signal transduction molecules. This project is a collaboration with Dr. C. Heldin (Ludwig Institute, Uppsala). Initially, it is planned to overproduce the extra-cellular domain of the receptor for crystallisation purposes.
The third project will focus on the retinoic-acid receptor, a continuation of our work on understanding the molecular basis of the profound effects of retinoids on cell growth and differentiation. The RAR/RXR receptors are members of the steroid-hormone-receptor superfamily and are built up from small domains (transactivation, DNA-binding, ligand-binding). They modulate the transcriptional activity of specific genes in a ligand-dependent fashion. Various constructs are being expressed to produce material for crystallisation trials.
* Structures of Interest as Targets for Drug Design.
Allocation: 3 students, 1 research assistant.
Molecular structures have been studied that are or could be of interest as targets for future drug design. So far, there has been no direct involvement in a structure-based drug-design cycle (i.e., "rational drug design"). In this cycle, the macromolecular structure alone or in complex with a ligand or inhibitor, is used to design a new ligand with specific properties. After synthesis, the behaviour of the new compound is evaluated and, if necessary, the structure of the complex determined. This cycle may be repeated a number of times until a successful new compound with the desired characteristics is produced. This technique is now widely used by the international pharmaceutical community. Appendix V lists some of the successful compounds that have been designed in this way.
It is intended to expand studies in this area, and to strengthen our ability to use structural features of the enzymes under investigation in order to find potent selective inhibitors. Research would focus on the structure determination of suitable macromolecules and the development of related bio-computing methods. One of the macromolecules that is of interest is the catalytic subunit of ribonucleotide reductase, which is a key enzyme in DNA synthesis and a potential target for anti-cancer, anti-viral and anti-bacterial drugs. Prolyl 4-hydroxylase is studied because it is medically interesting: various fibric disorders, such as liver cirrhosis, are characterised by excessive collagen deposition. Other projects involve the study of 5-lipoxygenase; deacetoxycephalosporin C synthetase from Streptomyces clavuligerus (which catalyses the ring expansion of penicillin N and the subsequent hydroxylation step in the synthesis of cephalosporin C); pili proteins from infectious bacteria that use pili to adhere to carbohydrate receptors on target cells; DNA analogues, and peptide nucleic acid analogues (with a backbone of N-(2-aminoethyl)glycine to which nucleobases are attached) as potential gene-targeted drugs.
Research will also be focused on the development and application of modern computer simulation techniques for studying the relationships between structure, function, energetics and dynamics of biomolecules. In the area of drug design and protein-ligand interactions, new techniques for the evaluation of binding affinities have been developed by Åqvist. They have been successfully applied to various binding and molecular recognition problems, such as inhibitor binding to endothiapepsin, HIV proteinase, trypsin and the recognition of sugars by a bacterial glucose/galactose receptor.
This programme would act as a link between our structural investigators (both theoretical and experimental) and drug companies. It would also interact with some of the planned new projects, especially that on membrane receptors.
* Enzymes as Laboratory and Industrial Tools.
Allocation: 2 students, 2 research assistants.
This project includes the use of enzymes in stereo-specific organic synthesis and separation, and the development of more efficient, stable enzymes. The on-going structural and protein engineering studies of a number of enzymes (e.g., hydrolases, lipases, cellulases) would be extended with the aim to develop and tailor enzymes as catalysts for chemical processes in the laboratory and in industry. Significant new structures have recently been solved in our laboratory, including the first cellulase structure to be determined (CBH II) and, more recently, another cellobiohydrolase, CBH I. These proteins are of interest for reasons of basic research and practical applications. Increasing the catalytic efficiency of cellulases, for example, could result in significant savings in biomass-conversion processes.
Research carried out within this section of the programme will, it is believed, be important for applications in the Swedish forest industry and, more generally, for the movement towards a future with a "greener" industrial environment.
* Viruses and Viral Components.
PI: L. Liljas.
Allocation: 2 students.
The main goal of this project is to elucidate the mechanisms of the different steps in the life cycle of viruses through structure determination of virus particles and viral proteins. One part of this project involves the study of small RNA bacteriophages (MS2, Qb), as well as more complex viruses. The recent structure determination of the MS2 protein capsid with a specific stem-loop fragment of the viral RNA, for the first time revealed the conformation of a complex between a translational repressor and mRNA. This complex is also the starting point for the assembly of the virus. Studies of other complexes between the RNA fragment and mutants of the protein will give more insight into the factors that govern the specificity of such a complex. Structure determination of proteins from phage F6 will shed light on the function of the complicated RNA-replicating machinery of this virus.
The second part of this project is a collaboration with several laboratories (both industrial and academic) to develop new inhibitors of HIV-1 reverse transcriptase (RT) and proteinase. The structure of the palm sub-domains of RT has recently been solved by Unge and co-workers at high resolution. It is now planned to solve the structures of inhibitor complexes of both enzymes and to develop expression systems for a number of other viral enzymes.
* Protein Expression Laboratory.
Allocation: 1 research engineer, 1 laboratory assistant.
One of the most important problems in X-ray crystallography is the preparation of interesting proteins in sufficient quantities for crystallisation. The creation of a modern expression laboratory is viewed as the highest priority item by the Uppsala structural community. Large-scale protein production is best carried out within the crystallographic laboratories concerned. Experience shows that the resulting protein is of higher quality when the crystallographers are directly involved in all steps of the preparation. The expression laboratory would serve two functions, protein expression and purification. By grouping together workers with similar concerns (in both expression and purification), and by providing adequate equipment and expertise, we will greatly improve our present situation, and create a unit more responsive to our particular needs. The laboratory will maintain a collection of useful expression strains and plasmids. Centralisation will improve the training given to new members of the departments, and will result in more effective use of equipment and expertise. It would have two permanent staff members who would be responsible for the day-to-day running of the facility, and for specific preparative projects. They would also be involved in the training of up to four visiting students or post-doctoral fellows from our own laboratories who have specific short-term goals.
* Crystallographic Software and Methods Developments.
Allocation: 2 students.
Most of the computer programs used in structural research are either old or written in languages designed forty years ago. Jones has made major contributions to the development of methods to construct models from electron-density maps. Further improvements require the development of a new family of molecular modelling and crystallography programs that are designed using object-oriented principles and that make use of modern graphical workstations. Their ease of use would be a principal design goal, resulting in substantial long-term savings.
Specifically, the problem of further automating the structure building and rebuilding process will be addressed. This will require a formalisation of our knowledge regarding protein structure. The rebuilding process seems amenable to automation, since it is basically an algorithmic problem: each residue in turn is analysed, perhaps rebuilt and regularised, and a check made to see if rebuilding improved the situation. The actual tracing problem is much fuzzier and therefore more of a challenge. Other areas to be addressed are the incorporation of (partial) structure refinement into the rebuilding process, and structure analysis and structure validation (which will lay some of the groundwork for the automatic rebuilding project).
As soon as there is in-house NMR expertise, new ways of representing and analysing protein structures solved with this technique will be explored. It is expected that many of the tools and techniques developed in twenty years of crystallographic refinement and rebuilding could also be fruitfully applied in this area. At present, NMR structures are typically represented as an ensemble of one or two dozen structures generated with molecular dynamics and distance geometry programs. Each of these individual structures is accurate to only ~1-2 Å, and has a quality which tends to be comparable to that of a ~3 Å X-ray structure. The concepts of quality control and structure rebuilding, as well as knowledge as to what a good protein structure does and does not look like, should be incorporated into the NMR-based structure determination process. Moreover, statistical means of assessing the accuracy of NMR-derived protein structures are badly needed (e.g., NMR counterparts of Luzzati or SIGMAA plots). At a later stage, it would also be possible to develop tools for joint X-ray and NMR rebuilding of protein structures, just as it is possible today to refine structures with X-ray and NMR data simultaneously. This project embodies the type of synergy to be expected when NMR expertise exists inside a largely X-ray oriented community.
* Structural Studies on Secretory Proteins, Time-resolved
Crystallography and Bio-organic Synthesis.
Allocation: 2 students.
It is likely that by the time this programme is funded, Ylva Lindqvist and Gunter Schneider will have moved their laboratory, either to the Karolinska Institute in Solna, or to Umeå University. This group plans to carry out research on three projects. One project concerns the study of proteins involved in the protein secretory and translocation machinery. In collaboration with Dr. Hans Ronne of the Ludwig Institute, Uppsala, they have begun structural studies of the proteins SSO1 and SSO2 of the secretory pathway in the yeast Saccharomyces cerevisiae. The two proteins are homologous to mammalian syntaxins, and probably function in the docking of synaptic vesicles to the plasma membrane. Truncated versions of these proteins, in which the hydrophobic membrane-anchoring tail has been removed, can be expressed in E. coli and are presently used for crystallisation experiments. They aim to crystallise and determine the structure of these proteins, and later to extend the structural studies to other proteins involved in the docking complex of transport vesicles to the plasma membrane.
The second project will involve time-resolved crystallographic studies on dethiobiotin synthetase. The structure of this enzyme has been refined to 1.6 Å resolution, making it the first structure analysis of an ATP-dependent carboxylase. In addition, the structures of eight complexes of the enzyme with substrates, co-factors, reaction intermediates, etc., have been determined, resulting in snap-shots of the catalytic process. They would extend these studies to include time-resolved crystallography in order to describe catalysis in a more dynamic way.
The third project concerns three enzymes that are of interest in the stereo-specific synthesis of carbohydrates and other organic compounds, namely transketolase, transaldolase and L-ribulose-5-phosphate 4-epimerase. Within this programme, they will attempt to redesign the substrate specificity of these enzymes, which would broaden the range of applications in the industrial synthesis of carbohydrates and drugs. Transketolase contains vitamin B1-derived thiamine diphosphate as a prosthetic group, and the crystal structures of holo-transketolase and complexes of the enzyme with substrate, analogues of reaction intermediates, etc., have been determined in order to explain catalysis in structural terms. Parallel to these crystallographic studies, site-directed mutagenesis will be used to probe residues responsible for co-enzyme binding, substrate specificity and catalysis. The transketolase gene from yeast has been cloned and an expression system has been established. A number of transketolase mutants have been prepared and characterised biochemically and structurally. Transaldolase works in the same pathways as transketolase, and its structure determination is underway. Pentose phosphate epimerases catalyse the epimerisation at an asymmetric carbon centre. An efficient expression system for L-ribulose-5-phosphate 4-epimerase has been developed, the enzyme has been crystallised, and the structure determination is in progress.
3.1.3 Karolinska Institute
* Structural and Physico-chemical Basis for Protein-Nucleic
Acid Interactions in Biology.
Allocation: 2 students, 1 research assistant.
The steroid-hormone receptors are key proteins for the understanding of the molecular basis for hormone-induced gene regulation. These receptors are also considered to be suitable targets for treatment of a large number of medical disorders. Within the scope of the present application, the on-going hormone-receptor programme would be extended to include structural studies of orphan receptors that are members of the superfamily, but have no known ligand. These proteins would also be included in on-going biophysical characterisations of the thermodynamics of sequence-specific versus non-specific DNA-protein interactions.
Ribosome assembly constitutes the first molecular event in a series that results in protein biosynthesis. The biochemistry of ribosomes and their assembly is fairly well characterised and the field is now moving towards structural studies. It is planned to investigate the structural basis for the first events in ribosome assembly using thermostable proteins from Thermus thermophilus as a model. Some of this work has already been initiated by the (on-going) structure determination of the S15 protein, which is believed to act together with a fragment of rRNA in the first steps of the assembly. These studies will be extended to include the structure determination of other small-subunit ribosomal proteins from the same organism, and complexes between S15 and a short piece of rRNA. The investigations are expected to result in a large number of spin-off projects, including studies of dynamics, kinetics, specificities and thermodynamics of protein-RNA interactions in the ribosome.
The physical basis for molecular recognition is in many cases complex, involving obscure entropic effects and interactions with solvent. However, research in this area is important for applications such as rational drug design. A programme to elucidate the molecular mechanisms for sequence recognition by DNA-binding proteins is in progress. The approach is based on extensive comparisons between experiment (structures, kinetics and thermodynamics) and simulations (molecular dynamics and free-energy simulations). It is planned to extend these studies to include the endonuclease EcoRI:DNA complex, which provides a very attractive system for the investigation of the effect of both hydration and DNA bending on binding-site specificity. This part of the project will be carried out with Dr. L. Nilsson.
* Protein Expression and Purification.
Allocation: 2 research assistants.
The major bottleneck in almost all structural studies of proteins using NMR methods is the limited availability of adequate amounts of pure protein material as well as isotopically enriched material. NMR studies, in general, require more protein than crystallographic studies, and isotope labelling is expensive and requires high expression levels in bacteria to be economically feasible. Both NMR groups at the Karolinska Institute experience frustrating limitations in their ability to obtain sufficient amounts of (isotope-labelled) protein samples from their collaborators, and time is often wasted when NMR spectroscopists, with little or no experience in protein chemistry, try to express and label proteins.
Within the present application it is proposed that the two NMR groups at the Karolinska Institute each be provided with one expert protein chemists as well as some basic equipment needed for protein expression and purification work, and funds to purchase isotopically enriched media. The objective is to strengthen each group and the resources are therefore, for practical purposes, divided among the two laboratories. Both researchers/technicians should be knowledgeable in gene expression and protein purification, and their primary tasks in the two laboratories would be to optimise conditions for cell growth and gene expression in various (isotopically labelled) media, and to assist in purification of protein products.
* DNA-binding Proteins from Thermophilic Organisms.
Allocation: 1 student.
The DNA of all cells must be structurally organised in a compact form and yet be readily available for transcription. The structure and assembly of eukaryotic chromatin is fairly well characterised at the structural and biochemical level, but very little is known about the structural organisation of the genome in prokaryotes. Research on DNA-binding proteins from the archaebacterium Sulfolobus solfataricus and related thermophilic organisms has been initiated. This has resulted in the first published structure of a "histone-like" protein since the HU structure was published in 1984, and the structure determination of the complex between this protein (Sso7d) and DNA is underway.
Within the framework of the present application it is planned to extend the programme to include additional DNA-binding proteins from archaebacteria. The objective is to use structure determination, in combination with biochemical and genetic methods and possibly electron microscopy, to characterise histone-like proteins and their function. This research will include biophysical characterisations of the proteins, which are all very thermostable and, hence, of biophysical as well as potential biotechnological interest.
* Electron crystallography of biomolecules.
Allocation: 1 student.
Electron crystallography is emerging as an alternative to X-ray crystallography and NMR for determining protein structures, in particular for proteins that do not crystallise readily in three dimensions and are unsuitable for NMR studies. Examples are membrane proteins and large molecular complexes. The KI group is the only group working in this field in Sweden. Recently, a 4 Å projection structure of microsomal glutathione transferase was obtained. This is one of very few membrane proteins for which structural information is available at this resolution. The continued work on this detoxification enzyme will involve determination of the three-dimensional structure, identification of a transmembrane domain from a truncated protein, and studies of conformational changes during activation. Another area of interest is P-type transport ATPases such as renal Na,K-ATPase and gastric H,K-ATPase. The aim of this project is to describe in detail the molecular processes involved in the energy transfer from ATP breakdown to translocation of ions across the membrane. The number of available membrane-protein preparations with properties suitable for forming two-dimensional crystals will increase steadily. New specimens will be tried at a level which is compatible with the resources of the research group. For instance, the group recently started to work on mitochondrial nicotinamide nucleotide transhydrogenase and yeast phosphate permease. Another area of special interest is 2D crystallisation of proteins using lipid monolayers and structure analysis of large molecular complexes. Technical developments such as the use of slow-scan CCD cameras and on-line processing for automatic focussing and collection of tilt series are important in these applications.
3.1.4 Lund University
* Proteins of the Blood Coagulation/Anticoagulation
Allocation: 1 student, 1 research assistant.
Blood coagulation provides many examples of biologically crucial protein-protein and protein-membrane interactions. In blood coagulation, several zymogens of proteolytic enzymes are sequentially activated by limited proteolysis in a precisely regulated manner. These reactions take place on biological membranes such as activated platelets. Each enzyme is virtually inactive against its physiological substrate in the absence of a suitable (phosphatidylserine-containing) membrane surface and the appropriate co-factor-binding protein, which together enhance the activity of the enzyme ~105-fold. At the same time, the structural basis for these complex interactions is largely unknown, mainly due to the occurrence of essential post-translationally modified amino acids. For example, the membrane-binding enzymes contain 9 to 12 Ca2+-binding g-carboxyglutamic acid (Gla) residues, which are a prerequisite for membrane interaction. Other modifications include b-hydroxyaspartic acid (Hya) and b-hydroxy-asparagine (Hyn). Difficulties encountered in eukaryotic expression of these proteins have prohibited production of sufficient quantities of the proteins for structural analysis.
Proteins involved in blood coagulation consist of several apparently individually folded domains, e.g. Gla, EGF, kringle and serine-protease domains. Recently, the structure of the Ca2+-free form of the Gla domain has been determined in this laboratory by NMR. The structure of the Ca2+-loaded form of the Gla domain of prothrombin had previously been determined by X-ray crystallography. By combining these two structures, insight has been gained into a dramatic Ca2+-induced conformational transition in a Gla domain. Upon Ca2+ binding, three juxtaposed hydrophobic amino acids expose their side chains to solvent in a manner that strongly implies that they mediate hydrophobic contact with the membrane. Inhibition of the carboxylation of Glu by vitamin K-antagonistic drugs leads to the synthesis of proteins that do not bind Ca2+ or interact with membranes, and which therefore are biologically inactive. Insight into this system is therefore highly desirable from many points of view. The following continuations and extensions to these studies are proposed:
1) Studies of the interaction of the N-terminal Gla-EGF domain pair of factor VII with its receptor ("tissue factor"). This interaction is crucial for the initiation of blood coagulation.
2) Studies of the mode of interaction of Gla modules with phospholipids. The binding affinity to various types of phospholipid will be determined.
3) Protein S contains four Ca2+-binding EGF domains arranged in tandem, with Hya and Hyn residues in the Ca2+-site region. It will be investigated how Ca2+-binding affects the structure of the four domains, and how an EGF domain affects the Ca2+-affinity of its neighbours.
4) Protein S interacts with C4b-binding protein of the complement system which provides a link between the coagulation and complement system. As a first step towards a study of the interaction between the proteins, the structure of a recombinant domain from C4b-binding protein that interacts with protein S will be determined.
For continued work on this project, efficient eukaryotic expression systems are indispensable.
* Experimental Studies of Biomolecular Dynamics by NMR
Allocation: 2 students, 1 research assistant.
While information about the three-dimensional structure of proteins is accumulating at an impressive pace, our knowledge about the detailed intramolecular dynamics of proteins still remains rather sketchy and incomplete. Yet, it has long been recognised that such dynamic information is necessary for a full understanding of the molecular basis of a wide range of biological phenomena. such as enzymatic catalysis, ligand binding and molecular recognition. Recent advances in molecular biology and NMR spectroscopy have made it possible to monitor the dynamics at virtually every atomic site of a protein molecule over a broad range of time scales. The present research project is a rather ambitious attempt to apply a wide variety of NMR methods, some recently developed, to study in detail the intramolecular dynamics (backbone motions and side chain motions) of a few selected proteins over as wide a time domain as possible. The influence of specific amino-acid modifications on protein dynamics and the effects of complex formation on the intramolecular dynamics will be investigated. The proteins in these studies will be chosen from those presently under study, or previously studied, in our laboratory for which efficient expression systems exist and for which site-specific mutants are available.
Characterisation of protein dynamics on the nanosecond to picosecond time scale will be carried out by measurements of different nuclear spin relaxation rates on proteins labelled with 15N as well as with 13C. Uniform labelling as well as specific labelling will be used. Studies will be carried out at different magnetic field strengths (2.3 T to 13.8 T ) in order to obtain a good description of the spectral density functions. In order to probe the dynamics of side-chain atoms we plan to use 2H, 15N and 13C triple-labelled proteins. Such labelling will also yield information arising from dipolar-quadrupolar cross-relaxation effects. Conformational dynamics on the millisecond to microsecond time scales will be measured through rotating-frame relaxation, line shape and magnetisation-transfer methods. In addition to the relaxation studies, amide-proton exchange rates will be studied on 15N labelled proteins
Close interactions are planned with internationally leading NMR experimentalists as well as with Swedish scientists interested in molecular dynamics simulations of proteins.
* Expression Laboratory for Post-translationally Modified
Allocation: 1 laboratory assistant, 1 research assistant.
The proteins of the blood coagulation/anticoagulation system contain a number of post-translationally modified amino acids such as Gla, Hya, and Hyn. These are essential for proper biological function. In order to carry out the research programme outlined above, it is essential to be able to produce proteins with native amino acids. For the NMR studies, it is necessary to produce 15N and 13C labelled material. To carry out this plan, a eukaryotic expression laboratory needs to be set up, which is dedicated to the preparation of recombinant proteins with post-translationally modified amino acids. Since eukaryotic expression systems are less well developed then prokaryotic systems, the laboratory staff will need to improve expression levels and protein purification.
* Crystallography at the MAX II Synchrotron.
PI: A. Liljas.
Allocation: area detector, two workstations.
The new synchrotron in Lund, MAX II (1.5 GeV), is in the final stages of construction and electron injection is planned at the end of 1994. With the aid of wigglers, it will produce intense X-ray radiation suitable for protein crystallography. Equipment for the first experimental station for crystallography (a multipole wiggler with 27 poles at a magnetic field of 1.8 T) was delivered in October 1994. The system will give a critical wavelength of 4.6 Å. At wavelengths around 0.9 to 1.0 Å, the beam intensity will be comparable to that of Station 9.6 at Daresbury, but a factor two less intense than a bending magnet at ESRF in Grenoble. At a wavelength of 1.5 Å, the station will be much more intense than Station 7.2 at Daresbury. The MAX II station will thus become a highly competitive facility for crystallography within Sweden, and constitutes a major advantage for Swedish crystallographic groups.
Optical elements at the station will be a focussing mirror followed by a bent single monochromator. The station will therefore be tunable for work with optimised anomalous dispersion. A cold room and equipment for sample handling will be available at the station. It is planned that the staff of the Department of Molecular Biophysics at Lund University will assist visiting scientists in the use of the station.
This project is to provide an image-plate area detector for use in macromolecular diffraction studies, and modern computer workstations to allow immediate processing of data as well as inspection of the results using computer graphics.
* Ribosomal Proteins and Factors.
PI: A. Liljas.
Allocation: 1 student, 1 research assistant.
The process of protein biosynthesis, in which the genetic information of the nucleic acids is translated into an amino-acid sequence, is of fundamental interest and is an example of a highly sophisticated form of molecular recognition. In modern biotechnology, the expression of proteins in vivo or in vitro is of vital importance. In addition, the protein-synthesis machinery acts as a multifaceted receptor system for a large number of antibiotics. One way to tackle the rapidly growing problem of antibiotic resistance is to use rational drug design starting from receptor-inhibitor interactions.
All organisms synthesise their proteins on ribosomes. Bacterial ribosomes have become the model system for this process. The functional steps are being characterised with increasing accuracy. To put the functional steps of protein biosynthesis on a firm structural basis, it is necessary to determine the three-dimensional structures of ribosomal proteins and factor proteins alone and in complex with ribosomal RNA. Some of the ribosomal proteins function as repressors of translation. In these cases, complexes of the protein with fragments of the mRNA are of great interest. The knowledge of RNA structure and its interactions with proteins is rather meagre except in the case of tRNA and a few recent complexes.
The structures of three ribosomal proteins, S6, L1 and L12, have been solved by this group. Two unexpected observations were the structural homology of ribosomal proteins (not recognisable from sequence data), and their recurring surfaces with exposed aromatic and basic groups, presumably for RNA interactions. Protein interactions with mRNA and ribosomal RNA are currently studied, with a special emphasis on proteins with mutations that lead to antibiotic resistance.
Protein synthesis on ribosomes is assisted by a large number of protein factors that bind transiently to the ribosome in different functional steps. Several of these proteins are G-proteins that bind to the ribosome in complex with GTP, and dissociate after GTP hydrolysis. The structure of elongation factor G (EF-G) in two functional conformations has been determined by this group, and progress has been made in investigating the whole conformational cycle. The observation that the structure of EF-G can be superimposed on the structure of the complex between EF-Tu and tRNA has far-reaching consequences that will be further explored. Fusidic acid is an antibiotic that stops protein synthesis by interacting with EF-G. Fusidic-acid-resistant mutants have been characterised and are presently analysed. These studies may lead to a new type of antibiotic that would inhibit the EF-G-dependent interaction with the ribosome.
Through various international collaborations, a number of other factors (tetM, SEL-B, RF-3, relA, spoT, and EF-2) with different functional interactions on the ribosome will be studied. Of special interest are the tetracyclin-resistance factor, and complexes between antibiotics and protein or RNA.
Personnel in this project will spend part of their time assisting visiting crystallographers on the MAX II diffraction beamline.
3.1.5 Stockholm University
* Di-iron Proteins and Phosphatases.
Allocation: 1 student.
Di-iron proteins are found in a number of important biological systems. The structures of metal-site mutants of the R2 subunit of ribonucleotide reductase (RNR) will be determined. This will allow the altered activities and metal-binding affinities to be related to changes in the enzyme's structure. Free radicals are important intermediates in the R2-catalysed chemistry; therefore, time-resolved cryo-crystallography will be used to determine the structure of radical forms of the protein. Methane monooxygenase (MMO) catalyses the hydroxylation of methane and of a large number of other hydrocarbons at a buried di-iron site. The study of the MMO hydroxylase in functional forms will be continued to investigate the structural basis for di-oxygen activation and substrate recognition. Rubrerythrin, an iron protein from sulfate-reducing bacteria, which is structurally related to RNR and MMO, will also be studied. The structural information on di-iron proteins will be used in the de novo design of a di-iron hydroxylase with broad substrate specificity. An MMO-type catalytic di-iron core will be engineered into the four-helix bundle of the iron-storage protein bacterioferritin, an unusually stable protein. Protein from active clones will be purified, crystallised and their structure determined.
Phosphotyrosine protein phosphatases (PTPases) are key enzymes in intracellular signal-transduction pathways. Work on the recently solved structure of a low-molecular-weight PTPase will be continued to investigate the structural basis of substrate recognition and the catalytic mechanism, by studying structures of the enzyme in complex with substrates and inhibitors. Acylphosphatase catalyses the hydrolysis of both synthetic and physiologically relevant acylphosphates such as 1,3-diphosphoglycerate, acetylphosphate, succinylphosphate and aspartylphosphate. The structure of acylphosphatase in two crystal forms will be determined, as well as complexes with substrate analogues.
The Rel-transcription factors are involved in immune response, oncogenesis, Drosophila development and HIV-1 replication. A new Drosophila Rel-family member, called dorsal-related immunity factor (Dif), was recently cloned. It encodes a transcription factor which is activated via a signalling pathway, and translocated into the nucleus upon bacterial infection. In the nucleus, Dif binds to the kB-like motifs in the promoter region of the Drosophila Cec genes and thereby induces the transcription of these genes, and possibly also of other immune-response related genes. The proteins of the Dif system will be crystallised separately and in complexes with each other and with cognate DNA.
3.1.6 Swedish NMR Centre
This facility, which is presently based in Stockholm, is available to all Swedish scientists. The funding of the Centre runs out in 1996. However, the board of directors and the scientific council are planning a continuation. The new NMR Centre should be located at a university with an active NMR research group. The old instruments will be transferred to the new location and additional funds will be made available to upgrade the instrumentation to the best available technical standard. This investment will hopefully include additional high-field (600 and 750 MHz) spectrometers. The organisation of the new NMR Centre will be similar to that of the present centre, and instruments and expertise will be made available to academics free of charge.
We feel that there are several advantages associated with a National Centre for NMR spectroscopy. These include: a fair division of large research investments among various research groups; efficient use of instruments through peer-reviewing of projects and availability of expert personnel; the possibility for researchers who are not experts in NMR to obtain help with problems that they otherwise would not be able to solve, and the possibility of knowledge-transfer through direct interactions between NMR spectroscopists as well as courses, symposia etc. which are organised at the Centre. We therefore propose to co-support a continuation of the Swedish NMR Centre within the framework of this programme. This support would amount to one position for a trained NMR spectroscopist (whose task it will be to assist other users; the salary would be paid from the contingency fund), and support for travel and lodging for researchers and students who visit the Centre.
The Network should in principle be open to anyone who is working in the field of Structural Biology, or has a strong interest in it. Research funds, however, should be allocated only to those groups who have a proven track record (see chapters 2 and 3). Some dynamics are essential to allow adequate response from the Network to unforeseen future developments. Therefore, 10 % of the budget has been reserved for allocation after a review which will be undertaken after the programme has been up and running for two years. Groups that do not qualify for funding at present might develop favourably and be eligible at that time. In addition, new groups might be established that deserve support from the Network. As for the graduate courses, students and research assistants from all Swedish Structural Biology groups will be invited and encouraged to attend. In addition, structural biologists from industry will be able to participate in courses and the annual conference.
We fully expect to see interactions develop between the networks supported by the Foundation for Strategic Research. This could result in new structural projects that would receive supported from the contingency fund. Allocations from this fund would be made by the Network's Scientific Board.