FBDD in Academia 1

I’ve now gone back to being a tourist and will be in Australia until the end of the month before heading north to Singapore and Malaysia for most of June. Feel free to get in touch if you’re based in either of those countries and would like to discuss fragment stuff or Drug Discovery in general.

Some time ago, I promised to post on FBDD in academia and really can’t keep putting this off. I’ve realised that it’s not going to be possible to squeeze everything into a single post so there should be at least one more post after this one. You should be warned that my academic career ended some years before people started to talk about FBDD so if I appear to be out of touch, it may well be because I am out of touch. Hopefully some of what I’m going to say may be of interest to some of you and please remember that this blog does allow its readers to comment.

I’ll start by making two points, both of which will be obvious to many of you. First fragment based approaches provide a means for drug discovery researchers (both in academia and start ups) to counter the advantages that Big Pharma derives from having massive screening libraries and automated compound handling. Secondly measurement of weak binding and determination of binding mode of weakly bound complexes remains a frontier area in physical biochemistry and biophysics. Remember that the power of a binding assay is defined by the weakness of the binding that can be measured reliably.

An academic group with strengths in protein structure determination and biophysical measurement of binding is well placed to contribute. I see the output of protein structural studies moving away from only determining the structure of a protein to providing a more integrated view of the protein’s ‘interaction potential’. One point worth making in this context is that measured thermodynamic parameters for fragment binding are particularly useful for developing and validating theoretical models because there are fewer protein-ligand contacts and it is easier to quantify conformational strain. Fragment based approaches also provide a means to validate and explore bioisosteric relationships without the need for a lot of synthesis and I’ve created a graphic showing how this might work.

Assembling and maintaining a usable screening library is likely to be a challenge or at very least an issue for most academic groups. However, a group that has established expertise in fragment screening does have some advantages in negotiating with suppliers of compounds who may value experimental characterisation of how well their compounds have behave under assay conditions. Vendors of specialist fragment libraries really should value this type of feedback and if they don’t they shouldn’t be in the business of marketing fragment libraries. I sometimes wonder if synthesis of fragments might form the basis for final year undergraduate synthesis projects which could be quite self-contained and include a molecular design component. In passing I’ll pose the question to readers from academia as to whether they think they’ve got molecular design adequately covered in courses at their universities although I’ll have to leave this topic for another post.

As we all know there is more to FBDD than fragment screening. Once you’ve found fragments that bind, tested analogues of these and determined crystal structures, you’ll need to do some synthesis. For a group whose main expertise is characterising binding and protein structure determination this may a good point to bail out and prepare the results for publication. A group with some access to synthesis may wish to take the project a bit further and publish once they’ve observed some SAR. One of the attractions of FBDD for academic researchers is that there are a number of points at which they can choose to write up the project for publication. It is also worth pointing out that FBDD provides an excellent framework to gain understanding of molecular properties and interactions between molecules. This understanding is essential if you’re planning to do molecular design the basis of which is manipulation of these properties with predictable results.

What if academic researchers want to take things further and generate lead series that will be of interest to Pharma? Synthesis will be necessary and life will get more complicated. I’ll pick this up in the next post (Kakadu salties permitting) since there’s quite a bit to say and I’m actually still thinking about this.

Melbourne, FBDD & Facebook

I now have less than a month left in Melbourne and since we’ve just switched over to winter time perhaps the hint should be taken. It has certainly been fun and the project is nicely under control although past experience suggests that it’s usually not a good idea to say that. I’ve not lived in a city since the mid-80s when I was a post-doc in Minneapolis and have really enjoyed the ease of getting around and ready access some of the wide range of music that Melbourne offers. I enjoyed an excellent performance by the ACO Soloists at Hamer Hall and am hoping to return there for a strong dose of Bach in a week’s time. University College, where I’m currently staying, is running a concert series and the second of these promises to be as enjoyable as the first. One of the music tutors at UC plays flute in the VYSO and I got to see them in action last weekend with their truly awesome guest soloist Kana Ohashi. This was also an excellent opportunity to watch the violinists since I was in the third row. The soloist was truly kinetic (difficult to be otherwise with the Tchaikovsky) and the first violin nearest to me appeared to have been given a special 'first violin bob' by her hairdresser. Maybe they will patent it.

On my first trip to the Paris Cat Jazz Club, I found it closed due to flooding (it was the day of The Hailstorm but at least there were back-to-back episodes of Hogan’s Heroes on TV). On returning the following week I was lucky enough to catch Monique diMattina who is an extremely warm, engaging and talented performer. By the way she also writes and composes and, as luck would have it, will be returning there the week AFTER I leave Melbourne. While wandering round town one Saturday afternoon, I caught The Wishing Well on Bourke Street and their next local gig is also the week after I leave. Good reasons to come back, I guess.

Previously, I pointed you towards some LinkedIn groups that are particularly relevant to FBDD. There are also groups on facebook that you might want to take a look at. It’s a bit more difficult to keep discussions going using the facebook groups because you don’t get alerted by email in the same way that you do with LinkedIn. However there are a lot of folk on facebook (especially in universities) and I believe it can play a useful part in extending the FBDD web. Here is a selection of facebook groups that you may find useful:

Fragment Based Drug Discovery (This is the group that is linked to this blog. I do check it frequently and usually respond to queries.)

Crystallography Rocks (Once you’ve got fragments to bind, you’ll want to see how they bind.)

NMR (There are a number of elegant NMR techniques for detection of ligand binding and you’ll find plenty of expertise in this group.)

Chemoinformatics (Particularly relevant to screening library design)

Dan gave my round the world trip a very flattering mention at Practical Fragments which did remind me that I really need to do a post on FBDD in academia since Teddy (who used facebook to tell me where Rapamycin comes from) has also discussed this. As I’m a real sucker for peer pressure, I do promise to make sure that my next blog post focuses on this topic. The FBDD facebook group led to me giving a lecture (I normally call these harangues) in Santiago and through it I’ve also made a couple of contacts in Singapore where I’ll probably do a couple of talks. Being in a facebook group also got me a chance to look round the Australian Synchrotron during maintenance week, when you can get a better look at all the cool stuff. I’ll finish with some pics from that visit.

FBDD and Networking

Reading an account of the session at the ACS on application of computational methods to FBDD, reminded me that it would be a good time to raise awareness of networking groups in this area. Both this blog and Practical Fragments allow readers to comment on posts although this tends not to happen with the frequency that it does at In the Pipeline, probably reflecting the huge readership, frequent updating and diverse content of what I consider to be the best drug discovery blog by a long way.

People interested in FBDD may already belong to a number of relevant LinkedIn groups. The groups offer some advantages over blogs for getting discussions going in that anyone can start a discussion and group members get alerted by email whenever somebody makes a new comment. I’ll list some of these below in case there are some that you’ve not yet heard about.

Fragment Based Drug Discovery (This group is linked by both FBDD blogs)

Label Free Assay Technology Group (It is the assay that makes FBDD possible. The weaker the binding that you can measure reliably, the more powerful your assay)

Structural Biology (X-ray Crystallography, NMR Spectroscopy, Electron Microscopy) (Generally you’re going to need crystal structures to take fragment hits forward)

Job opportunities in Computational Chemistry and Biology, Xray Crystallography, Fragment Based DD

Recently, I submitted the same item for discussion at a number of LinkedIn groups. I invited group members to share their views on the most appropriate technologies for detecting fragment binding. I learned about some new ways to configure SPR experiments and the use of Tm-shift assays. Most of the discussion was in the Structural Biology group (see discussion) although there was helpful input from the relatively new Label Free Assay Technology Group (see discussion) so thank you to all the participants. It was also great to see a couple of familiar faces from my days in Big Pharma, including a co-author from an article that a number of us wrote back in 2007

Interference, PAIN and cysteine pathologies

Dan provided some useful comments on the last post and I think it’s better to respond with a post since this makes everything more visible the readers of both our blogs. I agree with Dan’s point that there are pitfalls, such as compound aggregation, in addition to interference that Adam and colleagues describe in their article. In an ideal situation one would always have the ability to measure weak affinity directly. Protein-detect NMR is one of my personal favorites but you do need labelled protein and, if you want to get full value for your money (labelled protein is not cheap), you’ll also need resonance assignments. The SPR technology is widely applicable and like the protein-detect NMR will provide a direct measurement of affinity (and a whole bunch of other stuff). Isothermal titration calorimetry (ITC) represents another option although I believe that the technique is relatively sample-hungry and more limited than the other two techniques in the weakness of binding that can be measured. Also you do need heat so to speak even though the experiment is isothermal.

Nevertheless you can get to the point of having crystal structures with bound fragments using only a biochemical assay to measure potency. Given that you may well be screening at concentrations one or two orders of magnitude above what is ‘normal’ in HTS, it does make sense to use the approach that Adam and colleagues describe even if you’re going to follow up with SPR or NMR. I do sometimes wonder if the promiscuous behaviour of some inhibitors is due to this sort of interference rather than aggregation. One intriguing question is whether aggregates can ‘inhibit’ by changing spectroscopic and fluorimetric properties of assay mixtures rather than by interacting with proteins. At least there’s usually the option of running assays with added detergent to check for aggregation.

I won’t say much right now about the structural nasties that Jonathan Baell and Georgina Holloway have identified as PAINS since I’ll be visiting Jonathan at WEHI next Friday. I became acquainted with some of these unsavory structural types during my time in Big Pharma and do not believe that their PAINfulness is specific to the AlphaScreen technology that the WEHI researchers are using. Back in those days we had the Decrapper and a program called Flush...

Dan mentioned the Practical Fragments post on a Cruzain Screen so I thought I’d finish with a couple of papers that show how things can get unstuck when you’ve got a catalytic cysteine with a malicious streak. In the dock is none other than PTP1B, a target that is much-loved by disease area strategists and much-hated by screening groups. I’m not going to review the articles or even comment on them right now. Just read them in the correct order and perhaps we can pick up this theme later.

PTP1B: Read this first

PTP1B: Read this second

Literature cited

Baell & Holloway, New Substructure Filters for Removal of Pan Assay Interference Compounds (PAINS) from Screening Libraries and for Their Exclusion in Bioassays. J. Med. Chem. 2010, ASAP | DOI

Liljebris et al, Synthesis and biological activity of a novel class of pyridazine analogues as non-competitive reversible inhibitors of protein tyrosine phosphatase 1B (PTP1B). Bioorg. Med. Chem. 2002, 10, 3197-3122 | DOI

Tjernberg et al, Mechanism of action of pyridazine analogues on protein tyrosine phosphatase 1B (PTP1B). Bioorg. Med. Chem. Lett. 2004, 14, 891-897 | DOI

Interference correction in biochemical assays

Surface Plasmon Resonance (SPR) was in focus recently both here and across at Practical Fragments. However, now I’d like to take a look at using biochemical assays to identifying fragments that bind to targets of interest. Biochemical screens can typically be run in high throughput and are compatible with automation for high throughput screening, which makes it easy to do follow up screening with analogs. Furthermore the hits identified by biochemical assay are actually inhibiting rather than just binding. A criticism of biochemical screens is that they measure binding indirectly and are prone to interference. Sometimes they are used as a pre-screen to reduce the number of compounds that need to be evaluated in a lower throughput biophysical assay. However there are things that you can do to make your biochemical assay more reliable and meaningful. And maybe even more fun.

The article that I’ve chosen to take a look at in this post is by Adam Shapiro and some other colleagues from my days in Big Pharma. Before I met these folk, most of my fragment work had been around libraries for NMR screening and I learned from them how it is possible to correct for some of the interference from test samples in biochemical assays.

Inhibition is typically detected in a biochemical assay by quantifying changes in light absorption, fluorescence or luminescence. In high throughput applications ‘assay components are added serially to wells without any filtration or washing steps’ which means ‘that the test sample remains in the well during the optical measurement and can interfere with it’. This means that compounds that absorb in the UV or visible range and that fluoresce or quench fluorescence can all lead to changes in the readout parameter without actually binding to the target protein. Other less obvious causes of interference include insolubility of test compound (turbidity can lead to detection of highly polarised scattered light) and meniscus deepening which decreases path length. Compounds are typically assayed at relatively high concentrations in fragment screening, making it especially important to recognise and account for assay interference in these applications.

In addition to providing a useful discussion on the causes of interference, the article describes a practical approach to correcting for it by running ‘artefact assays’. These involve running additional plates in which wells contain the same test samples but no target protein. The wells in the artefact assay plate also need to contain whatever is responsible for generating the signal (e.g. reaction product) and a baseline can defined by preparing wells without test samples. The authors describe in some detail how they apply the corrections and since this is only a summary of the article, I’ll leave it to you to go and check their article out. However, I would like to conclude by noting that the authors also suggest criteria for deciding to reject data because interference is too great for meaningful correction.

Literature Cited

Shapiro, Walkup and Keating Correction for Interference by Test Samples in High-Throughput Assays. J. Biomol. Screen. 2009, 14, 1008-1016 | DOI

Surface Plasmon Resonance

General Reviews

Rich & Myszka, Grading the commercial optical biosensor literature – Class of 2008: ‘The Mighty Binders’ J. Mol. Recognit. 2010, 23, 1-64 Link | Review

Application to Fragment Screening

Giannetti, From experimental design to validated hits: A comprehensive walk-through of fragment lead identification using surface plasmon resonance. Methods Enzymol. 2012, 493, 169-218. DOI

Perspicace et al, Fragment-Based Screening Using Surface Plasmon Resonance Technology, J. Biomol. Screen. 2009, 14, 337-349 DOI | Review

Binding Pathologies

Giannetti et al, Surface Plasmon Resonance Based Assay for the Detection and Characterization of Promiscuous Inhibitors, J. Med. Chem. 2008, 51, 574-580 DOI | Review

Ligand protein interactions by SPR

I have now been in Melbourne for about a month and have found the city very much to my taste. I’m visiting some friends to help out with some fragment stuff and have already been wreck diving (on the HMAS Canberra) and watched the Australian Open and a rather one-sided ODI between Australia and the West Indies. On the science side of things, I was able to gatecrash Surface Plasmon Resonance (SPR) course, hosted by the Biomolecular Interaction Facility at CSIRO, Parkville, and taught by Rebecca Rich and David Myszka of the University of Utah. Not the whole course, I might add, because the participants spent the second day of the course in the lab and I’m sure there was a clause in my visa agreement that stipulated that I was not to enter a laboratory except as an observer accompanied by a responsible adult.

SPR has always represented a bit of a gap in my knowledge base so this was always going to be a great opportunity. As well as being experts in this field, Rebecca and David present their material with great clarity, enthusiasm, charm and humour. I particularly liked David’s take on the Maxwellian Demon (these molecules don’t have eyes).

When using SPR to screen ligands, the protein is typically immobilised on the surface of the sensor chip. The term ‘immobilised’ is actually a bit of a misnomer and ‘tethered’ would actually be a more appropriate term. The SPR technology can be used to look at diverse types of interaction over a wide range of affinities and kinetic parameters (e.g. on and off rates) can also be measured.

There is of course a slight catch. The experiments need to be performed carefully and this was a recurring theme in the lectures (and presumably in the practical sessions as well). Now it turns out that much of the SPR literature is perhaps based on experiments that have been performed less than perfectly and, as a public service, Rebecca and David have reviewed and graded the SPR literature of 2008. GRADED? Yes, GRADED, and there were some Fs! Of course David is just the person to do the grading since he sports whiskers of which a Victorian (historical context rather than geographical) head master would be justifiably proud and it is easy to imagine him summoning the hapless transgressors to his study.

A grading exercise like this is unlikely to win you many friends and the authors are realistic to accept that it is likely to reduce the likelihood of either being elected to the National Academy of Sciences although hopefully they will never have to employ the services of professional food tasters when they attend SPR conferences. Putting on my computational chemistry hat, I couldn’t help thinking that the QSAR and Virtual Screening fields might benefit from a similar treatment...

There are a number of articles describing the use of SPR to screen fragments against target proteins and the one I’ve chosen to take a look at is from some folk at Roche. One of the authors of this work is David Banner, whose talk at RSC BMCS 2009, I greatly enjoyed, not least because he made no reference to ligand efficiency except, if I recall correctly, to say that he would not be referring to it.

The Roche group screened a library of 2226 compounds against chymase at 200 micromolar and found 80 hits so clearly SPR technology can provide the throughput required to run a fragment screen. The compounds were screened against an inactive (zymogen) form of the protein as a check for non-specific binding. The authors also described cross-competition experiments which could be used to determine whether two fragments were binding at the same or different sites and it is worth remembering that you need to be able to measure binding very directly to get this sort of information. It would have been really interesting if the results of the cross-competition assays had been integrated with crystallography since 12 co-crystallised complexes showed fragments binding in the active site.

Both stoichiometry and kinetics of binding can be determined by SPR making it an appropriate technique with which to observe interactions between badly behaved ligands and proteins. In an excellent (A-graded by Rebecca and David) article, another Roche group exploit SPR to classify some of these binding pathologies. It is particularly good reading for anyone who has worked up results from high throughput screens but that is not a place I particularly want to go to right now since it’s getting rather late at night and I really don’t want to have nightmares about pathological fragments.

Literature cited

Rich & Myszka, Grading the commercial optical biosensor literature – Class of 2008: ‘The Mighty Binders’ J. Mol. Recognit. 2010, 23, 1-64 Link

Perspicace et al, Fragment-Based Screening Using Surface Plasmon Resonance Technology, J. Biomol. Screen. 2009, 14, 337-349 DOI

Giannetti et al, Surface Plasmon Resonance Based Assay for the Detection and Characterization of Promiscuous Inhibitors, J. Med. Chem. 2008, 51, 574-580 DOI