Fragment screen: Have you identified the right hits?


One of the problems faced by drug discovery scientists running a screening programme is the correlation of hits (or the lack of correlation) from one assay technique compared to another, run against the same drug target. Most often this occurs as part of screening cascade, when you move from one primary screening method to a secondary confirmation method and a vast majority of hits don’t repeat or have different ranked orders of potency. This can be for good reasons, and why the secondary technique is being used as a counter screen to remove false positives.

This process can also be used to validate a new screening methodology by comparing the correlation of actives from screening file when it is run with two differing techniques against the same target. This latter example was described in a recent publication

This paper has also been described on the Practical Fragments blog. Please read for a far more detailed description and many other interesting articles.

In this publication the authors from two separate groups ran a NMR and SPR fragment screen on the same target (HIV-1 integrase core domain) with the same fragment file (about 500 compounds). Once primary actives had been identified, the authors used X-ray crystallography as further evidence of binding. The results from this experiment showed that the NMR technique identified 62 primary hits of which 15 generated protein/fragment complexes. The SPR generated 16 primary hits which led to six clear protein/fragment complexes. When the fragments/crystal complexes were compared to one another there was no correlation between the different assay methods.

There were some differences between the two techniques which might account for the differences, for example the total amount of DMSO present varied in each technique, which could affect solubility of compounds, and the final screening concentration of the SPR (500µM) was less than that of the NMR (1mM) which would suggest some of the weaker compounds would be more difficult to find in the SPR. In addition to this, the conditions used for crystallisation are bespoke for each protein. This would favour a certain set of physiochemical properties which might assist the formation of a crystal complex for a certain sub-population of fragments. All these factors could affect the amount of hits identified for each method.

Another critical difference it seems between the two techniques is that the SPR method used a reference protein (AMA1) as control. Hits were identified from SPR, only if the binding response was twice the level with HIV-1 integrase compared to the AMA1. When comparing the NMR actives, 8 compounds which gave crystal structures did not show this fold binding difference in the SPR (even though they bound to the HIV-1 integrase and therefore were eliminated as potential hits from the SPR hit list.

When the 5 SPR hits which gave crystal structures were run as singletons with NMR, 4 of them gave a weak response and it was suggested that the pooling of the fragments (10 compounds per sample) made the identification of these compounds indistinguishable from the background signal.

SPR vs NMR

So overall both methods did find active fragments, however importantly each method did not find all the active fragments from the collection.

The different methods of analysis and screening had an impact on the population of the hits that were found for each of the techniques. Maybe once a reasonable subset of the active fragments has been identified, the use of structural similarities searches could be of assistance to locating other fragments from the file, to ensure that as few as possible active fragments are not missed.

From this review it does not look like one method has any greater performance than another in identifying active fragments, however it seems care needs to be taken with the analysis.

Advertisements

A Staple Diet for Peptide Mimetics


Among other things, the PPI-net conference at the Royal Society in London earlier this year gave some interesting insight into different techniques that researchers are using to synthesize stable proteomimetics. One such technique makes use of hydrocarbon ‘staples’ between amino acids along one side of an α-helical polypeptide in order to fix its conformation as well as provide some protection against proteases when in vivo.

Intramolecular bridging of α-helical peptides has been attempted in many different ways, such as lactams or disulfide bridges. In 1998, Grubbs developed the use of ring closing metathesis to form peptide ‘staples’, which were both stable and relatively easy to access synthetically. Since then, this technique has been refined so that hydrocarbon staples can be incorporated into known α-helices, allowing for relatively stable competitors for therapeutically targeted PPI’s.

TM01Scheme 1: RCM by Grubbs catalyst, to form an i, i+4 stapled peptide

 This technique was pioneered by Korsmeyer, who in 2004 published a paper detailing the use of stapled peptides to target the BH3 domain on BCL-2 family proteins. Korsmeyer was able to create peptides with nanomolar binding constants (KD = 38.8 nM), which proved to be both protease-resistant as well as exhibiting good bioavailability when tested in vivo. Other PPI’s have since been targeted by this technique, such as another ‘holy grail’ of cancer therapeutics, the p53-HDM2/HDMX interaction.

TM02

Figure 1: Binding affinity of stapled peptides to Bcl-x

More recently, some creative use of this chemistry has allowed the synthesis of stapled helices with some interesting physical properties. One such example is the work done by Rudolf Allemann at the University of Cardiff. Allemann takes advantage of the physical properties of azobenzenes, whereby they are able to undergo light induced trans-cis isomerization. By incorporating the azobenzene functionality into an intramolecular peptide bridge, this property can be transferred onto the helix. By attaching the staple in an i, i+7 conformation (or i, i+4 for shorter sequences), the helix will be stabilized when in the cis conformation, and therefore activated light energy. Alternatively, by using an i, i+11 it is the trans conformation has increased stability, and the protein is therefore deactivated by light.

TM03Scheme 2: Trans-cis isomerization of azobenzene peptide bridges

Experimentally, this technique has been shown to work using in vitro protein assays on a range of proteins. The most notable of which is probably our old friend the BH3 domain, this time using Bcl-x as the target protein. Three different photocontrollable peptides were synthesized by modification of a known Bcl-x inhibitor, using an azobenzene bridge in the i, i+7, i, i+11, and i, i+4 positions. Each three of these were shown to have activity dependent on being in either the trans or cis conformation. The i, i+7 and i, i+4 bridged peptides both showed around 20-fold increase in activity after irradiation with light, with the i, i+4 shifting from a KD [nM] of 1275 ± 139 to 55 ± 4 after isomerization.

TM04Table 1: Binding affinities of photocontrollable Bcl-x inhibitors

Although it is not immediately obvious on face value how this work could be transferred into the therapeutic environment, it provides an interesting example of how peptide staples can provide more than just stabilization of the parent α-helix.

Assessment of Cancer Genes for Drug Discovery


The Cancer Gene Census documents a list of genes which when genetically altered are known to contribute directly to cancer.

A recent paper by Patel et al describes a systematic, computational protocol, that they have used to identify which of these genes code for proteins that would be possible candidate targets, suitable for therapeutic modulation in the treatment of cancer.  A suite of analyses were undertaken to explore the biological and chemical space of these proteins (shown below).

Frances3

Following the computational analysis, the authors prioritize these proteins for drug development.  First they identified twenty-five proteins already known to be drug targets, with compounds with full FDA approval.  They suggest that some of the compounds may be useful for repurposing in different types of cancer.  For instance Smoothened SMO is the target of Vismodegib was recently approved for the treatment of basal cell carcinoma.  By mining multi-omic data from The Cancer Genome Atlas the authors suggest that Vismodegib might also be of use in treating Multiforme Glioblastoma (GB), as SMO was over-expressed in 95% of the GB samples analysed.

A  further eight-six proteins had active chemical compounds with submicromolar activity in biochemical or binding assays reported in the Chembl database.

They also explored which proteins had a known structure and predicted potential druggable pockets. Figure 2 illustrates the three-dimensional structure of GNAS with the druggable cavity displayed as a surface.  GNAS has an activating dominant mutation in pituitary adenoma, and further activating mutations have also been identified in kidney, thyroid, adenocortical, colorectal and Leydig tumours.  The authors suggest that small-molecule inhibitors of this enzyme regulator may have potential therapeutic applications.

Frances4

Of the 488 cancer gene census proteins, the authors identify 103 with good evidence of chemical tractability and group them by  “drug development” risk. They identify 46 proteins, whose genes are known to be genetically altered in cancer, whose structures are predicted to be druggable, with few or no know active small molecule modulators, that may be potential therapeutic targets. They suggest that these targets indicate new biological areas for chemical exploration in the treatment of cancer, but they also represent a high potential drug development risk.

Frances5

Safe azide or oxymoron?


The Huisgen cycloaddition or ‘click chemistry’ could certainly take part of the blame for the resurgence of one of the bad boys of organic chemistry: azide. Coming from a large – Risk averse – pharma, I have always tried to avoid such fragments, possibly due to the implications and paperwork if (or more likely when) it goes wrong. So when an article with ‘Azide’, ‘Facile’ and ‘Safe’ is published it’s certainly worth reading.

The article from Wang’s group is looking at the diazotransfer reaction converting a primary amine to the corresponding azide.

The pre-Goddard-Borger and Stick era was using triflate azides as diazotransfer reagents which proved to be prone to explosion (Figure 1).

azide1Figure 1

The Wang article here is looking at a safe protocol to the imidazole-1-sulfonyl azide (compound 4, Figure 2) reagents developed and optimised over the last 6 years.

azide2Figure 2

Key requirements to the described ‘safe’ route were to avoid the presence of NaN3 with strong acids, minimise the excess of NaN3 and avoid the formation of explosive intermediates.  Previously reported procedures to prepare diazotransfer reagents such as those depicted in Figure 2 all seem to engage sulfonyl chloride, leading to the generation of (N3)2SO2 as a highly explosive byproduct. Wang starts from sulfuryl diimidazole which after mono methylation is treated with NaN3 to give the  imidazole sulfuryl azide reagent (Figure 3).

azide3

Figure 3

Worth noting that dimethylation of the sulfuryl diimidazole is not observed, so no highly explosive (N3)2SO2 species were observed and that sulfuryl diimidazole itself proved unreactive with NaN3 (Figure 4).

azide4Figure 4

Furthermore, Wang report that the aqueous conditions the reaction is performed in prevent the formation of the explosive (N3)2SO2 intermediate from the diazotransfer reagent itself (Figure 5).

azide5

Figure 5

Does that make the whole process safe? What about stability and storage? Wang et al. prepared the  imidazole-1-sulfonyl azide (compound 4, Figure 2) in over 100g scale but seems to have used it in-situ….

Imidazole-1-sulfonyl azide (the preferred diazotransfer reagent from the Figure 2 bunch) is usually prepared as a HCl salt. The ‘safety update’ from Goddard-Borger and Stick, published in 2011 as a follow up to their original 2007 ‘shelf-stable’ imidazole-1-sulfonyl azide, reports that ‘imidazole-1-sulfonyl azide hydrochloride is hygroscopic and reacts slowly with water to produce hydrazoic acid. Concentration of the mother liquors from which imidazole-1-sulfonyl azide hydrochloride crystallised has resulted in an explosion. This solution may contain sulfonyl diazides and/or hydrazoic acid byproducts which are both extremely sensitive, explosive substances’.

 

The Sejer group seems to work regularly with such diazotransfer reagents and reports that ‘rigorous drying of the HCl salt of imidazole-1-sulfonyl azide followed by storage at -20⁰C makes it stable for >1 year’ and that tetrafluoroborate and hydrogensulfate salts of imidazole-1-sulfonyl azide were found to be much better with respect to shelf life.

 

Back to Wang et al.’s conclusion that the protocol can be applied to large scale preparation in both academia and industry…. I will ensure it’s on my day off!