Evolutionary fold space preferences

At group meeting last week I focussed, alongside some metaphysical speculation, on a project which has occupied the first half of my DPhil: namely exploring the preferences of both very old and very young protein structures. This work is currently in preparation for publication so I will give only a brief overview and hopefully update the juicy details later. Feel free to contact me for more information.

Proteins are the molecular machinery of the cell. Their evolution is one of the most fundamental processes which has delivered the diversity and complexity of life that we see around ourselves today. Despite this diversity, protein domains (independent folding units) of known structure fall into just over 1,000 unique SCOP folds.

This project has sought to identify how populations of proteins at different stages of evolution explore their possible structure space.

Superfamily ages

Structural domains are clustered at different levels of similarity within the SCOP classification. At the superfamily level this classification attempts to capture evolutionary relationships through structural and functional similarities even if sequence diversion has occurred.

Evolutionary ages for these superfamilies are then estimated from their phylogenetic profiles across the tree of life. These ages are an estimate of the structural ancestor for a superfamily.

The phylogenetic occurrence profiles are constructed using predictions of superfamilies on completely sequenced genomes using HMMs and taken from the SUPERFAMILY database. Given an occurrence profile and a phylogenetic tree (for robustness we consider several possible reconstructions of the tree of life) we use a maximum parsimony algorithm (proposed by Mirkin et. al) which estimates the simplest scenario of loss events (domain loss on a genome) and gain events (domain gain) at internal nodes on the tree which explains the occurrence profile. The age estimate is the height of the first gain event, normalised between 0 (at the leaves of the tree) and 1 (at the root).

We estimated ages for 1,962 SCOP superfamilies and compared several properties relating to their primary, secondary and tertiary structures, as well as their functions. In particular, we compared two populations of superfamilies: ancients, with an age of 1, and new-borns, with an age < 0.4. Full details of our results will hopefully be published shortly so watch this space!

Antimicrobial Drug Discovery Conference (Madrid)

I am a big fan of taking something, either a poster or a talk, to a conference, and getting something back – other than a €6 box of airport chocolates. This blog post is in that spirit.

On the plane to the “Antimicrobial Drug Discovery” conference in Madrid I was reading the Cassandra Project (a novel on smallpox, how apt) instead of the stack overflow of scientific papers I planned to read. Classic JP.

The conference had a mix of experienced, invited speakers and early stage researchers. It was very “biological” for a computational scientist, so quite removed from what I normally do – but an opportunity to learn nonetheless.

The keynote lecture was by Julian Davies, a fantastic speaker who gave a general overview of antibiotics and antibiotic resistance. Antibiotic resistance is a real concern (even those politicians in the G8 noticed a few hours ago!) and there is a fear we might return to pre-antibiotics era when you could not cure common diseases like bacterial pneumonia. Pharmaceutical companies all got out of antibiotic research years ago, and there have been no new antibiotic scaffolds for more than a decade. I found this surprising as you would think that there was a truckload of money to be made from finding the new penicillin. Apparently, there is little return in anti-infectives because of rapid mutation of the pathogen and its short-term use (curing the infection, as opposed to having to take your medication for life, such as beta-blockers for hyperventilation). Bacteria should not only be considered at an individual cell level but also as a population with complex signalling between the individuals (which may offer a way to stop bacterial infection). In order to combat infections and increasing resistance sick patients are now supplied with combinations of drugs – this is still dangerous due to the possible (toxic) drug-drug interactions.

Natural products, e.g. some toxins, are good antibiotics but it is very hard to optimize such compounds to improve their drug profile (chemical synthesis of natural products is difficult). Also a lot of people at the conference were talking of how antimicrobial peptides will save the day. The attendees with drug discovery experience raised an eyebrow about this, knowing how hard it will be to make a 30 residue peptide into a drug.

Some antibiotics work by having a hydrophilic part (e.g. carboxyl) and a hydrophobic part (e.g. an alkane chain). This hydrophobic part sits in the membrane wall disrupting it, which creates a “leak” from the bacteria which eventually kills the pathogen. There are other mechanisms of action such as blocking transporter or signalling channels.

There was a brilliant, energetic talk by Bruno Gonzalez-Zorn with the audience paying rapt attention. He showed how bacteria have these multiple, small plasmids offering antibiotic resistance. He discovered there was a common two-part theme to antibiotic resistance, where a particular gene is always present.

Paul Finn gave a much needed talk on why drug discovery is hard (e.g. target selection, difficulty to get drugs in therapeutic area, potency, toxicity, have to optimize for different variables, etc.). Unknowingly proving this point, there was this earlier talk of a whole optimization series which got a small molecule inhibitor of a viral infection from 150uM down to 1uM (IC50) – a great result in itself, and when the investigators tested this ligand in vivo rather than in vitro it simply did not have any affect on the virus.

Cele Abad Zapatero, one of the main investigators of AltasCBS, made the point that, today, we do not know where we are in drug discovery. He argued we need to move to chemical-biology space instead of simply chemical space and recommended the use of ligand efficiency indices (e.g. BEI, SEI).

Madrid was way too much fun. Zidane (and a few thousand others) kissed this Champions League Cup in exactly the same place. Talking about microbes. (click to enlarge)

And what did I take to the conference? I took a poster, the design of which is based on Dunbar’s stylish template. Marta, Ana and myself won a “highly commendable” poster prize with the best poster going to Laura (Synthetic inhibitors of bacterial cell division targeting the GTP binding site of FtsZ, since you asked). There were 24 posters in all, and mine was the only computational study in a room otherwise filled with phages, bacteria and plasmids (literally as well as metaphorically). There is a sinister heart-warming joy in winning a bottle of wine, instead of a cheque or a certificate. James deserves a sip or two.

Cheekily asking for a corkscrew during the poster prize award

[Publication] Effect of Single Amino Acid Substitution Observed in Cancer on Pim-1 Kinase Thermodynamic Stability and Structure

In this study we selected point mutations resulting in Pim-1 variants that are expressed in cancer tissues and reported in SNP databases, such as FastSNP and COSMIC. These Pim-1 variants have been comprehensively characterized to investigate the effect of single amino acid substitution on Pim-1 thermal and thermodynamic stability and structure in solution. Our results indicate that the effects of the mutation observed in cancer tissues cause local changes of tertiary structure, but do not affect binding to type I kinase inhibitors.

This work has been pioneered by researches at the Department of Biochemical Sciences “A. Rossi Fanelli”, Sapienza University of Rome and served as an inspiration for one of my thesis chapters.

Standardize a PDB

In many applications you need to constrain PDB files to certain chains. You can do it using this program.

A. What does it do?

Given a pdb file, write out the ATOM and HETATM entries for the supplied chain(s), put them on a single chain with the name provided by the user.

PDB_standardize needs four arguments:

PDB file to constrain.
Chains from the pdb file to constrain.
Output file.
Name of the new chain

B. Requirements:

Biopython – should be installed on your machines but in case you want to use it locally, download the latest version into the PDB_constrain.py’s directory (don’t need to build).

C. Example use:

C.1 Constrain 1A2Y.pdb to chains A and B, placed on a single chain C – write results in constr.pdb

python PDB_standardize.py -f 1A2Y.pdb -c AB -o const.pdb -s C

C.2 Constrain 1ACY to chain L, with a new chain name F, write results in const.pdb – this example shows that the constrainer works well with ‘insertion’ residue numbering as in antibodies where you have 27A, 27B etc.

python PDB_standardize.py -f 1ACY.pdb -c L -o const.pdb -s F

And the Oscar goes to …

08:44

I have a bet I can write a blog post in under 20 mins. Last week I asked Leila Tamara where her blog post was (we aim to have one up every week) and she replied that “It was ready but hadn’t been proof-read yet“. I guess it goes to show the diligence of the students here at Oxford. (Some of them, anyway).

So as a Marie Curie Stars ITN Fellow, my time is coming to an end (in mid-September). Part of my D.Phil was about finding novel anti-malarial inhibitors using computational methods (virtual screening). No, we haven’t cured it yet.

During a conference in Riga we made a video about the project – I find it funny to see myself on screen, (and to hear my heavily accented commentary). Pity the Oscars have already been dished out this year!

And a “making of” still

08:52.

Warning: this post hasn’t been proof-read yet!

Protein kinases, the PIM story

Last week I was presenting my DPhil work. In one of my projects I address the reasons for inhibitor selectivity in PIM protein kinase family. PIM kinases play key roles in signalling pathways and have been identified as oncogenes long time ago. Slightly unusual for protein kinases ATP-binding sites and cancer roles have prompted the investigation of potential PIM-selective inhibitors for anticancer therapy. Due to overlapping functions of the three PIM isoforms, efficacious inhibitors should bind to all three isozymes. However, most reported inhibitors show considerable selectivity for PIM1 and PIM3 over PIM2 and the mechanisms leading to this selectivity remain unclear.

Figure 1. Workflow of the sequence and structure analysis of PIM kinases

To establish the sequence determinants of inhibitor selectivity we investigated the phylogenetic relationships of PIM kinases and their structural conformations upon ligand binding (Figure 1). Together with my OPIG supervisor Charlotte Deane we predicted a set of candidates for site-directed mutagenesis as illustrated in Figure 2. The mutants were designed to convert PIM1 residues into analogous PIM2 residues at the same positions.

I then moved to the wetlab to test the hypotheses experimentally. Under guidance of Oleg Fedorov, I screened the SGC library of kinase inhibitors using differential scanning fluorimetry (DSF). After comparing melting temperature shift values across the PIM kinases and mutants, a set of potent inhibitors with different chemical scaffolds have been selected for quantitative binding analysis. I worked with Peter Drueker’s team at Novartis on PIMs enzymology, where I measured activities, Km values for ATP and IC50s using mobility shift assay. For my final set of measurements I performed isothermal titration calorimetry (ITC) experiments back at the SGC and determined binding constants and enthalpic/entropic contributions to the total free energy of ligand binding.

Figure 2. An overlay of PIM1 and PIM2 structures (P-loop and hinge regions), the mutated residues are shown as sticks

The data are yet to be published, I only briefly state the results here. The hinge mutant E124L demonstrated reduced thermal stability probably due to removal of E124-R122 salt bridge. The P-loop mutants had intermediate Km ATP values between PIM1 and PIM2, indicating that those residues could be responsible for stronger ATP binding in PIM2. As shown in Figure 2, the residues are located at the tip of the P-loop and might have involvement in the P-loop movement. Importantly, three mutants have shown reduced affinity to inhibitors validating my initial hypotheses.

Ideally having PIM1 and PIM2 co-crystal structures with the same inhibitors would allow direct comparison of the binding modes. So far I was able to solve apo-PIM2 structure in addition to the single PIM2 pdb, which will be deposited shortly.

I will update you soon about on my second project which involves more mutants, type II inhibitors, equilibrium shifts and speculations about conformational transitions. Keep visiting us!

Inside Memoir: MP-T aligns membrane proteins

Although Memoir has received a lot of air-time on this blog, we haven’t gone into a great deal of detail about how it models membrane proteins. Memoir is a pipeline involving a series of programs iMembrane -> MP-T -> Medeller -> Fread, and in this post I’ll explain the MP-T step (I’ll briefly touch on Medeller too).

Let’s first look at the big-picture. There are several ways of modelling a protein’s 3D structure. In an ideal world we could specify an extended polypeptide, teach a computer some physics, set if off simulating, and watch the exact folding pathway of a protein. This doesn’t work. A second method would be to build up a protein from lots of fragments of unrelated proteins… this is usually what is meant by ‘ab initio’ modelling. The most accurate (and least sophisticated) approach is to find a protein of known structure with similar sequence, align the sequences, and copy over the coordinates of the aligned residues to make a model for the query protein. This is the approach taken by Memoir and is called homology modelling or comparative modelling.

The diagram below shows an example of how homology modelling might work. Four membrane protein sequences are aligned (left) and the alignment specifies a structural superposition (right). Assume now that the red structure is unknown: we could make a good model for it just by copying over the aligned parts of the blue, green and yellow structures.

The greatest difficulty in the modelling described above is making an accurate alignment. As sequences become more distantly related they share less and less sequence identity, and working out the optimum alignment becomes challenging. This problem is especially acute for membrane protein modelling: there are so few structures from which to copy coordinates that a randomly chosen query protein has a good chance of having <30% sequence identity to the nearest related structure.

Although alignment is the most important facet of homology modelling it is not the only consideration. In the above diagram the centres of the proteins are structurally very conserved (so copying coordinates will lead to a good model in this region), but the top of the proteins differ (the stringy loops don’t sit on top of one another). It is the role of coordinate generation software to distinguish which coordinates to copy. It turns out that the pattern of a conserved centre and varying top/bottom is generally true for membrane proteins, and Memoir uses our Medeller coordinate generation software to take advantage of this pattern.

Back then to alignment. The aim of alignment is to work out which amino acids in one protein are related to amino acids in another. All alignment methods have at their heart a set of scores which encode the propensity for one amino acid to mutate to another, and for that mutation to become fixed in a population. These scores form a substitution table (here mutation + fixation = substitution). More sophisticated alignment methods augment these scores in different ways — for example by adding in scoring based on secondary structure, smoothing scores over a window, or estimating a statistical supplement to the score determined from a related set of pre-aligned sequences — but at some level a substitution table is always present. Using a substitution table, the most likely evolutionary relationship between two sequences can be detected and this is reported in the form of an alignment.

So that’s general alignment, now to apply this to membrane proteins. The cell membrane is composed of a lipid bilayer: a sandwich with a hydophobic filling and hydrophilic crusts. The part of a membrane protein that touches the filling will have different preferences for amino acids (and, more importantly, substitutions between these amino acids) than the part of a membrane protein that touches the crust. Similarly there are systematic preferences for amino acid substitutions depending on whether part of a protein is buried or exposed, and on which type of secondary structure it assumes. The figure below shows a membrane protein with different regions of the membrane and different types of secondary structure annotated.

It is possible to make separate substitution tables for each environment within a membrane protein, where an environment specifies where the protein sits in the membrane, what secondary structure it has, and whether it is accessible or buried. Below is a principal components analysis of the resulting set of tables: each table is represented by a single point and the axes show the direction of the greatest variation between the tables. The plot on the right shows a separation of the points based on whether they are buried (more hydrophobic) or accessible (more hydrophilic). The hydrophobic centre (red circles) and hydrophilic edges (green circles) of the membrane fall into this general pattern. The table on the left shows that the tables further divide by secondary structure type. In summary there are systematic substitution preferences in practice as well as theory, and for membrane proteins it is most important to consider hydrophobicity when aligning two protein sequences.

On then to modelling. The conventional approach to aligning a pair of sequences for homology modelling is to take a set of pre-aligned sequences (a sequence profile), and use them to estimate a supplement to the standard substitution score for aligning two sequences. This is termed profile-profile alignment. Memoir takes a different approach by using the MP-T program to construct a multiple sequence alignment scored with environment-specific substitution tables. The alignment includes a set of homologous sequences to the pair of interest.

Profile-profile alignment methods and MP-T are very different. It is unclear whether the substitution preferences at a position are best estimated by MP-T’s tables or the supplements derived from sequence profiles, and the answer probably depends on how well the profiles are made — garbage in, garbage out. Similarly the MP-T algorithm only determines the upper limit of alignment accuracy, and the actual accuracy depends on how the homologous sequences in the alignment are chosen.

In general we find little difference between the fraction of an alignment that MP-T and either HHsearch or Promals (profile-profile alignment methods) gets right. However we do find a difference in the fraction of the alignment that these methods get wrong (part of an alignment can be right, wrong or simply not aligned, so it’s possible to get a lower fraction wrong whilst getting the same fraction right). It turns out that on average MP-T gets less of an alignment wrong for simple reasons of combinatorics: for a pair of proteins, the number of possible multiple sequence alignments is much greater than the number of possible profile-profile alignments. This means that, just by chance, the number of incorrectly aligned positions between the two sequences of interest will be lower for MP-T than for a conventional profile-profile alignment method.

Now for a little sales-pitch. The source code for MP-T is freely available and easy to expand (if you have a passing familiarity with Haskell). Only two or three lines of code need to be changed to define a new set of protein environments, and to feed it a substitution table for each environment. I’d be happy to help anyone who wants to try it out.

[Publication] Memoir: template-based structure prediction for membrane proteins

Congratulations to all involved in the Memoir publication in Nucleic Acids Research!

Memoir is a web server which builds homology models for membrane proteins. It is a web-enabled workflow combining some of OPIG’s software; MP-T, IMembrane, Medeller & Fread. The inputs are a sequence of the membrane protein you wish to model (target) and a PDB file to use as template.

Memoir may be found here and there is also a video tutorial narrated by Jamie. There is even a funny blooper of him practising, which I kept to celebrate this moment.

Happy modelling!

Free food!

Yesterday I walked into Group Meeting not having read Bernhard‘s paper (shameful, I know), and I was immediately asked “Where is the Daleks post on the blog?”. To which I mumbled something unconstrued, because I am not sure what a Dalek is and because I didn’t know we were doing post requests.

Anyway, at every group meeting one of us is responsible to organise the talk and another to supply food. The only current rule (since the well-received demise of the “No alcohol” one) is: “No tomatoes“. We’ve had a number of original and tasty contributions: Dominos pizza, Ben’s and Millies cookies, truckloads of Haribos, Krispy Kreme Doughnuts, Sushi, Nutella baguettes and home-baked delights.

But Eoin‘s contribution takes the prize this round (a small trophy in Lab Room #1).

Eoin’s Dr. Who Daleks sugar rush inducing cakes (click for the juicy detail).

So, a small pointer to OPIG prospective students – “baking” and “creative thinking” skills are really well appreciated and look good on your CV!

Journalclub: Molecular Dynamics simulations of TCRpMHC

Introduction

T cells recognize fragments of pathogen (peptides) presented by the Major Histocompatibility Complex (MHC) via their T-cell receptor (TCR). This interaction process is commonly considered as one of the most important events taken place in the adaptive immune reaction.

Molecular Dynamics simulations are a computational technique to simulate the movement of atoms over time. For this purpose the interaction energies (bond and non-bond) between the single atoms are calculated and the spatial position are adjusted during each iteration. Such simulations are very resource and time consuming but provide insights into interaction processes which can not be obtained by any currently available experimental technique.

In this journal club we discussed 3 different papers dealing with MD simulations of the TCRpMHC complex:

A typical story

Epitope Flexibility and Dynamic Footprint Revealed by Molecular Dynamics of a pMHC-TCR Complex
Reboul et al., Plos Comp. Biol. 2012

Like similarly done by many other authors before Reboul et al. performed MD simulations of two different (however very similar MHCs) in complex with the same viral peptide. While no immune reaction is caused if the peptide is presented by HLA-B*3501 there is an reaction induced if presented in the context of HLA-B*3508.

In their MD simulations the authors find minor differences in the RMSF and claim this to be systematic and the cause for the different behaviour.

An innovative story

Toward an atomistic understanding of the immune synapse: Large-scale molecular dynamics simulation of a membrane embedded TCR–pMHC–CD4 complex
Wan et al., Molecular Immunology 2008

While several PDB structures of parts of the core of the immunological synapse are available (see image below). On overall structure was not published before this paper. This is addressed by the authors by means of superimposition, modelling of linking and trans-membrane regions, and subsequent MD simulation. The resulting structure seems to be in good agreement with experimental electron microscopy data.

My story

Early relaxation dynamics in the LC 13 T cell receptor in reaction to 172 altered peptide ligands: A molecular dynamics simulation study
Knapp et al., Plos One 2013

In most studies authors compare the same MHC but with two or three different peptides or the same peptide bound to 2 MHCs. In some cases also the same peptide and MHC are simulated in interaction with 2 different TCRs. Given the fact that the TCRpMHC consists of roughly 800 AAs one will almost certainly find some differences between those two or three simulations (multiple testing). Differences would also be present if one simulates the same complex twice with different starting velocities or more extreme even if one parametrizes the same velocities but different hardware is used. Yes, also in this case this may lead to slightly different results. On this basis such studies (if published without further experimental data to undermine the findings) are at best anecdotal stories.

Therefore we indented to address this challenge in a more systematic way: We simulated the LC 13 TCR / HLA-B*08:01 system in complex with all possible single point mutations in the EBV peptide FLRGRAYGL. This leads to a total of 172 highly related MD simulations where for each of them the experimental immunogenicity is known. Based on their immunogencity we assigned each simulations to either the more immunogenic (moreI) or less immunogenic (lessI) group. This was repeated for several thresholds. Further analysis on the basis of RMSD maps and permutation tests showed that moreI and lessI groups were significantly different in their initial relaxation dynamics from the (perturbed) x-ray structure.

They were not only significantly different but they also showed a quite interesting pattern in their most frequently different regions (highlighted in green):

Oxford Protein Informatics Group

or "OPIG" to friends