Dynamics and catalysis in integral membrane pyrophosphatases

Lead Research Organisation: University of Leeds
Department Name: Astbury Centre

Abstract

60% of drug targets are integral membrane proteins - but just 3% of all solved structures. In addition, fast kinetic analysis on membrane proteins has been restricted to proteins like cytochrome c oxidase. Integral membrane pyrophosphatases (mPPases) are evolutionarily conserved ionic pumps that convert the free energy in pyrophosphate into a sodium and/or proton gradient across a membrane. They are unlike any other protein, do not occur in multicellular animals, and are essential under conditions of low-energy stress.

In addition to plants and (archae)bacteria, mPPases occur in pathogens: protozoan parasites like Leishmania (leishmaniasis), Trypanosoma species (Nagana, sleeping sickness), Toxoplasma gondii (infecting up to 90% of pigs) and Plasmodium species (malaria), as well as Bacteroides vulgatus, which is the most common cause of brain abscesses (20% mortality rate). These diseases affect human health and food security across much of the world, and the protozoan diseases, except for malaria, are classes as "neglected tropical diseases". Due to global warming, the insect vectors that spread these diseases are already spreading into Europe and will be common in the summer in Northern Europe in the next 30 years. We have shown that deleting the mPPase gene in P. falciparum makes it non-infectious. mPPases are thus a potential drug target, and our preliminary work suggests it is suitable for kinetic analysis. Developing drugs against these enzymes will have important long-term benefits for animal health, food security, and human disease, by providing new weapons against major animal and human diseases.

This work extends and deepens our ground-breaking structures of the bacterial Na+-pumping Thermotoga maritima mPPase (TmPPase) and H+-pumping Vigna radiata (mung bean) mPPase (VrPPase). With previous BBSRC funding, we developed four novel mPPase inhibitor scaffolds, three of which are active against the malaria parasite at low uM concentrations. The molecules work in unexpected ways, by blocking the exit channel in an allosteric manner. Our vision is to extend our structural studies and use single molecule functional, time-resolved crystallography and molecular dynamics simulations to determine intermediate enzymatic states.

Our multidisciplinary approach has two main strands: (1) focussing on understanding the structural correlates behind the different mPPases. There are at least five different families, which pump different ions and respond differently to changes in sodium (Na) and potassium (K) concentration; and (2) using various dynamic (single-molecule fluorescence resonance energy transfer (FRET), time-resolved serial synchrotron crystallography (SSX) and solution (Pulsed Electron-Electron Double Resonance (PELDOR)) approaches to understand the choreography of the enzyme mechanism. The two strands of work inform each other, as the static structural studies will generate hypotheses that can be tested by biophysical techniques.

Our aim is to understand what motions in the helices leading to gate opening and thus ion pumping, how these differ between sodium- and proton-pumping mPPases, and how the binding and pumping conformational changes are allosterically transmitted between the two monomers, leading to half-of-the-sites reactivity. The work will use the new allosteric inhibitors that we have developed. We expect our work to be revolutionary in the level of detail we obtain about this enzyme.

Technical Summary

Integral membrane pyrophosphatases (mPPases) are novel, conserved ion pumps that use the free energy in the pyrophosphate POP bond to generate a sodium and/or proton motive force. Although not found in multicellular animals, they occur in plants, protozoan parasites (eg malaria: P. falciparum) and (archae)bacteria and are essential under conditions of low-energy stress. We showed that mPPase is essential in P. falciparum (Totanes, unpubl.), and that the mechanism appears to be "binding change" (Li et al, Nat Comm, 2016). However, our most recent work shows that the enzyme is allosteric with half-site reactivity, and we have developed allosteric inhibitors (Vidilaseris et al, Sci Adv, 2019). Our model system is T. maritima mPPase (TmPPase).

Our main objectives are thus:
1. Solving structures of integral membrane proton-pumping pyrophosphatases with different inhibitors bound and from different families, including the Na/H-pumping organisms to identify (1) the structural correlates of K-dependence/independence and (2) the structural correlates of the pumped ion (Na/H/both).

2 We have Cys mutants that report on helix motions. We will use them with PELDOR and single molecule TIRF to observe changes +/- ligands, in lipids, nanodiscs to understand changes. smFRET will allow us to examine structural dynamics. Serial synchrotron crystallography will report on global changes upon ligand binding (TmPPase is slow at 20C). We will use steered molecular dynamics to derive models of transient states. The work will report on the allosteric changes that are part of our model of dual-pumping and will lead to new designs for inhibitors, covering all the expected timescales (ns-s).

This work will have important benefits for understanding allostery in mPPases and so in integral membrane proteins in general. By helping our design of mPPase inhibitors against protozoan parasites (in parallel work), the work will have benefits for animal health, food security and human disease.

Planned Impact

The immediate beneficiaries of this research will be other academics, as outlined under academic beneficiaries. Non-academic beneficiaries fall into the following five main classes:
1) The public, through training of the next generation of scientists.
2) Society and the economy, through improved global health and food security
3) The private sector, through novel product development and commercial revenue
4) Public stakeholders
5) Society, through public engagement and discussion of science.

We focus on understanding the dynamics and allosteric mechanism of a novel membrane protein, integral membrane pyrophosphatase (mPPase). By doing so, we will (1) understand how our allosteric inhibitors work and (2) in the long-term improve them into potential lead molecules with low nM affinities. mPPases, which do not occur in mammals nor in most bacteria, are essential under conditions of low-energy stress in e.g. protozoan parasites. Protozoan parasites are major causes of both animal and human morbidity and mortality, through diseases like malaria (Plasmodium spp: 214M cases in 2015) and Toxoplasma gondii (infection rates as high as 90% in pigs). Many are on the WHO list of neglected tropical diseases.

1. Developing highly skilled people. A major transferrable benefit will be the people trained during the project. A PDRA trained on my previous BBSRC project is now employed at Peak Proteins. These include the PDRA, who will acquire multiple specialist scientific skills to use in research-based biotechnological industry and academia, the graduate and undergraduate students who will be involved in the projects, and the technician. The University of Leeds has staff development programmes to provide transferrable skills. These trained people, as they move to other institutions in academia, in government and in industry, will affect larger society positively.

2. Global health and food security. Novel drugs to treat protistal diseases could significantly improve global human and animal health by providing more treatment options. Reducing the human disease burden allows individuals to remain economically active and reducing the animal disease burden improves food security though reduced losses of animals.

3. Industrial involvement. SMEs and big pharma will benefit from this research. AG collaborates with Novartis and is part of two separate EU Innovative Training Networks including Novartis, AstraZeneca, Biomerieux and more than 10 SMEs. Infectious and parasitic diseases are a growing burden, so fundamental research on developing new targets and potential inhibitors will be exciting for companies for use both in animal and human health. This is relevant not only for protozoan diseases but also in treating Bacteroides brain abscesses, with an associated mortality of 20%. The timeframe for development is about 10 years.

4. Public health stakeholders. Leeds has an exemplary record in disseminating research and contributing to the public understanding of science in England and in Europe. New approaches to these diseases is important for national and international stakeholders, ranging from the Department of Overseas Development to international health charities and WHO.

5. Society. Work with potential to lead to superior outcomes will be disseminated widely (TV, radio, YouTube, press releases, Blogs, Twitter). MD simulations are particularly useful for science communication and will be used to explain findings to a general audience. Our focus is also on enthusing and training the next generation of scientists being STEM (Science Technology Engineering and Maths) ambassadors. We engage with students in secondary education; we aim to enthuse school children to study science in annual Discovery Zone workshops; and we inspire undergraduates completing 3rd-year projects.
 
Description As of 1st of January 2021, Dr Christos Pliotas has taken over as the acting-PI of this grant. Christos was previously a co-PI on this grant, but after Adrian Goldman's relocation to Helsinki, due to Brexit implications, Christos agreed to act as the PI. After starting work on this project on 18th February 2020, Dr James Hillier (PDRA on this project) was only able to work in the lab for four weeks before the university labs were fully closed due to Covid-19. He was subsequently furloughed from April 1st until 3rd August 2020. Our lab was eventually allowed to open on 19th August. Therefore, progress towards the grant milestones has been affected considerably.

Despite this, good progress has been made on many of the objectives. Regarding objective 1.1, a crystal structure of PaPPase has been determined and is currently being refined, with the aim of publishing the structure along with some functional assays this year.

Regarding objective 1.2, the purification protocol for CpPPase is currently being optimised, and we anticipate that suitable samples for crystallisation experiments can be produced within the next month. A problem has arisen whereby the protein precipitates when concentrated but it is hoped by screening different conditions a buffer can be produced in which the protein is stable. This problem may also be alleviated by subjecting samples to gel filtration prior to final concentration.

Objective 1.2 also involves using pulsed electron-electron double resonance (PELDOR) experiments to investigate asymmetry in mPPases. So far, samples of TmPPase, with three residues mutated to cysteine to facilitate labelling, have been produced and spin-labelled. These are currently being analysed at the electron paramagnetic spectroscopy (EPR) facility at the University of Manchester, with results expected in the coming weeks. A total of four conditions for each Cys mutant will be tested, including apo TmPPase, as well as TmPPase with a range of ligands. It is anticipated that by comparing results from the different sample conditions, it will be possible to derive insights into asymmetry in mPPase function.

1.2. Combining information with work in the Goldman group in Helsinki has given a comparison of structures between the Na+ and H+ pumps. Work on acquiring the crystal structure of a dual-pump has been temporarily set aside as more intensive work progresses on achieving PELDOR data (objective 2.1). Once the PELDOR workflow is satisfactorily functional the optimised crystallisation protocol can be applied to TmPPase homologues.

2.1. PELDOR mutants of TmPPase have been made and are being expressed and labelled. Comparative data of three mutations of the protein in detergent are likely to be forthcoming in the next few months. Nanodiscs are not, at present, a main objective, as TmPPase is stable in detergent.

2.2. Crystallisation conditions have been optimised for the TmPPase mutants used in PELDOR. Crystals have been obtained for one mutant, and expression of the others is ongoing. Furthermore, CpPPase crystals have been generated and sent to Diamond for diffraction trials. This will enable to investigate both catalytic site loop closure and helix 12 motion in CpPPase. This will address both objectives 1.2 and 2.1.

2.3 has also been addressed with the creation of a lipid screen to identify lipids that can stabilise a particular protein. Using this screen, lipids have been identified that stabilise TmPPase. These results are currently being prepared for publication. Thus far, molecular dynamic simulations have not been undertaken, although forthcoming PELDOR data to provide distance restrictions will enable these to be performed with a significantly shorter processing time.

New updates as of 14/03/2023:
Objective 4.1 (1) - has been accomplished and we have currently a manuscript in revision for eLife containing these data. In this manuscript we solved an x-ray structure of PaPPase that explains how K+ activation works.

Objective 4.1 (2-3) - Our time-resolved structures of TmPPase in various states (0-3600s) (manuscript in revision for eLife) combined with recently obtained Nanion surfer data have provided a model for half of the sites reactivity, ion-selectivity (Na, H, or both) and how this is coupled to motions at the gate. The time points selected were very long as TmPPase is a very slow enzyme at room temperature. Key is that the TmPPase time-resolved structures have two novel features (1): they show asymmetric binding of PPi - only to one active site, not both, between 300 and 600s, and the PPi first binds at the "distal" phosphate binding site (only to the leaving-group phosphate site) and then at the canonical site (with one P in the nucleophilic site and the other in the leaving group phosphate site). Overall, most of the objectives of 4.1 have been achieved, except for structures of CpPPase and BvPPase.

Objective 4.2 (1) Initial PELDOR/DEER distance measurement experiments have been conducted for detergent solubilised proteins on the 211C and 546C variants of TmPPase and under three different conditions. (2) PELDOR experiments indicate that PPi, IDP, etidronate stabilise distinct conformations, and this has been supported by x-ray structures of TmPPase with IDP (published) etidronate (new) and zolendronate (new) bound. The PPi structures solved by time-resolved crystallography also reveal unexpected binding modes that will be further investigated at faster time scales by time resolved PELDOR measurements planned facilitating the new flash freezing apparatus we developed and built for this purpose (see below). (3) Role of lipids: Molecular dynamics simulations have shown how and where lipids bind to TmPPase and other mPPases. This work has been published (Holmes, A. O. M., Goldman, A. & Kalli, A. C., 2022) mPPases create a conserved anionic membrane fingerprint as identified via multi-scale simulations (PLoS Comput. Biol. 18, e1010578) and binding data has been used to show that anionic lipids stabilise TmPPase (Cecchetti, C., Strauss, J., Stohrer, C., Naylor, C., Pryor, E., Hobbs, J., Tanley, S., Goldman, A. & Byrne, B., 2021). A novel high-throughput screen for identifying lipids that stabilise membrane proteins in detergent-based solution (PLoS One 16, e0254118). We have spin labelled 211C with MTSSL (211R1) and prepared four equivalent detergent conditions to the ones stated above (plus the apo state). We have currently requested access to the pulsed EPR spectrometer in Manchester (e.g. there was a large sample backlog created after the facilities were shut down for a long period of time due to covid) and are expected to be performed within the next couple of months. These PELDOR experiments will enable direct comparison between different substrates and thus conclude their effect on the structural mechanism and gating, including asymmetry, of TmPPase. Further, initial MD simulations using the experimental PELDOR distance restraints have been set up and should be completed when the new PDs are recruited to conclude the TmPPase states in detergent and compare with the new x-ray structures we obtained under different conditions.

Future work
Time resolved EPR/PELDOR (TR-EPR) - We have started building a flash freezing apparatus (< 1ms time resolution), which is expected to be completed next month. We will test and calibrate it using model systems and use on TmPPase to implement time-resolved PELDOR distance measurements. The TR-EPR device is almost accomplished and will be tested in Manchester, immediately after PD1 is recruited. These TR-PELDOR distance measurements will also compensate for the planned FRET experiments given the very high time resolution (<1 ms) of our new flash-freezing apparatus. In this respect, the delay in the grant will lead to better outcomes, because rather than having the device at the end, we will have it at the beginning of this second part of the project.

Plans 1: PD2 will prepare TmPPase samples for PELDOR on the 207C, 211C, 396C, 474C, 518C, 525C, 546C, and 599C, single-Cys mutants we have designed and also Cys-pair mutants we will construct, after testing the single ones for functionality. PD1 will do the spin labelling of TmPPases samples and then use time-resolved PELDOR to assess conformational TmPPases during binding. This will be done on both PPi and on IDP, etidronate and zolendronate substrates. In particular, our work suggests that etidronate is not able to cause conformational changes leading to pumping (?anion surfer data), while PPi and IDP can. We will look at these conformational changes by PELDOR, x-ray and the timescales by TR-EPR (< ms resolution). We would also use a smaller spin label than the MTSSL we currently use, to substitute for either a Glu or a Lys place a spin reporter into the ionic gate.

Plans 2: We will acquire structures of the Cp and BvPPases, which have not yet been studied (PD2). We will use mutagenesis and state of the art computational methods such as AlphaFold2 to generate stabilising mutations that should lead to better diffracting crystals.

Plans 3: The most efficiently labelled TmPPase mutants as judged in Plans 1, will be reconstituted in nanodiscs of known lipid composition and perform PELDOR distance measurements and TR-EPR/PELDOR to assess conformation and the effect of all four substrates in native lipid membrane environment (PD1). PD1 will then and use these measurements along with PD2's time resolved x-ray structures and mutational analyses to guide the MD simulations and conclude our integrated gating model for mPPases, using restraints from multiple methods (PD1 & PD2).
Exploitation Route Work on these outcomes is continuing within our research group as the grant is still active.
1. Identification of conformational transitions in membrane proteins and development of methods to study this will be of use to others studying integral membrane proteins. In particular, the development of an apparatus to provide low millisecond resolution of protein structure dynamics immediately following mixing with other molecules such as substrates, activators, or inhibitors. At present the only devices of this kind globally is in the NIH in Bethesda, MD, USA and Germany. The reproduction and refinement of this instrument and its capabilities will be beneficial to the understanding of structural and mechanistic changes in protein activity and the time scales and kinetics associated with those fundamental processes.
2. Work on these outcomes is continuing within our research group as the grant is still active and may furthermore be taken forward by the Goldman group in Helsinki to develop further inhibitory molecules. The design of structurally informed small molecule inhibitors will enable the development of small-molecule drugs targeting mPPases, ultimately leading to more effective treatment of protozoan diseases such as Leishmaniasis, Chagas Disease and African Sleeping Sickness (Trypanosomiases), and Malaria, which, in 2014, were estimated to cause a combined 1.1 million human deaths annually, with a further impact on livestock.
Sectors Agriculture, Food and Drink,Pharmaceuticals and Medical Biotechnology

 
Description Invited seminar at the Institute of Infection, Immunity and Inflammation at the University of Glasgow 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Postgraduate students
Results and Impact Christos Pliotas delivered an external seminar hosted by the Institute of Infection, Immunity and Inflammation at the University of Glasgow, presenting recent research activities of his group on bacterial mechanosensitive channels and pulsed EPR spectroscopy applications on membrane proteins.
Year(s) Of Engagement Activity 2022
URL https://www.pliotasgroup.org/invited-talk-in-glasgow/
 
Description Manchester Institute of Biotechnology Invited Seminar 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Postgraduate students
Results and Impact Christos Pliotas was invited to give a seminar at the Manchester Institute of Biotechnology, The University of Manchester as part of institute's regular bi-weekly/monthly seminar series. His talk attended postgraduates, postdocs and academics mainly from the University of Manchester but also elsewhere and involved his current research on mechanosensitive channels and PELDOR spectroscopy on membrane proteins.
Year(s) Of Engagement Activity 2022