Phenotypic screening for inflammasome modulators: A powerful example of a high content phenotypic screening approach where there are complex biological systems.

Using a phenotypic approach for HTS

HTS is a ubiquitous and well-established approach for hit identification; it is often used in conjunction with other techniques, such as virtual and fragment-based drug discovery, for a target-based screening approach. Phenotypic screening, however, is used to identify new molecular target(s) or mechanisms and/or hits that modulate complex disease mechanisms in a desirable way.

The mechanism of a hit target interaction and the target’s identity are initially unknown post-phenotypic screening. This could be because the signalling cascade or disease process is poorly understood. The advantages of using this approach are that it delivers active molecules within a disease relevant cellular background, inclusive of solubility and cell permeability, providing candidates for further medicinal chemistry optimization. Phenotypic screening approaches are often aimed at delivering first-in-class compounds with disease-modifying modalities (Heilker, Lessel and Bischoff, 2019). Usually, developing an appropriate phenotypic assay is a compromise between achieving HTS assay practicality whilst maintaining a strong assay-disease linkage.

NLRP1 mediated inflammasome formation – a case study 

NLRP1 (aka NALP1) was the first described inflammasome sensor of the Nod-like Receptors (NLR) family, (Taabazuing, Griswold and Bachovchin, 2020). NLRP1 is an innate pattern recognition receptor; it senses molecules that pose a danger to the cell (i.e., bacterial toxins, stress factors and viral proteases). DPP-8/-9 (a dipeptidyl peptidase) holds NLRP1 in an auto-inhibitory state until stimulation/release, a process that can be artificially induced by Val-boroPro (VbP), a small-molecular inhibitor of DPP-8/-9 activity.

When active, NLRP1 facilitates inflammasome assembly, enabling activation of caspase-1, which, in turn, leads to the activation and release of IL-1ß, IL-18 and Gasdermin (which undergoes oligomerization and forms pores in the membrane), culminating in pyroptic cell death. There is considerable interest in the identification of inhibitors of excessive NLRP1 activity associated with a wide range of chronic inflammatory diseases, like IBD and asthma. Whilst NLRP1 activators are being explored as an immuno-oncology approach to treat cancer.

Several pathways can trigger NLRP1 activation that are unique to this class of inflammasome: –

• Inhibition of the NLRP1-repressors DPP-8/-9 by Val-boroPro (investigated here)
• Cleavage of NLRP1 at its N-terminal domain by viral proteases
• ZAK-α-dependent NLRP1-phosphorylation in its N-terminal domain following UVB exposure.
• Sensing of long viral dsRNA through the LRR and NACHT domains.

We used a phenotypic screening approach for this target, considering the complex multimeric nature of NLPR1 and the technical difficulties in recapitulating an NLRP1 activation model using biochemical assay platforms. Binding type assays could potentially be used with purified NLRP1 or its domains (NACHT, FIIND, CARD, etc) but hits would not necessarily translate to functional inhibitors of NLRP1 inflammasome activation.

Here, we developed a High Content Screening assay where ASC, one of the components of the inflammasome, was labelled with GFP enabling facile visualization of intracellular inflammasome formation as a punctate ‘speck.’

This cell-based phenotypic assay provided an opportunity to evaluate NLRP1 activation (and inhibition) in a holistic cellular context, with the potential to identify modulators acting via multiple mechanisms. Often, when screening phenotypically, a disease relevant or patient derived cell line can be used to make the screen proximal to the disease state. In this screen, we used a cell line supplied by Invivogen. A549 is a lung carcinoma epithelial cell line, endogenously expressing proteins involved in the inflammasome signalling including ASC, caspase-1, and Gasdermin D/E.

Figure 1 Schematic of inflammasome formation, including GFP-labelled ASC, used to visualize the inflammasome as a ‘speck.’

Figure 2 Representative images of A549 cells expressing ASC::GFP ± VbP induction. Cells were stained with Hoescht to visualize nuclei.

Figure 3 Graphs showing the effect of VbP on cell nuclei count, intracellular speck count and the calculated specks/cell ratio

We have established a fixed endpoint phenotypic cell assay to identify potential modulators of NLRP1-mediated inflammasome formation.

The phenotypic screen

A 5000-compound subset of the Sygnature Discovery LeadFinder Diversity library was run in the primary screen. Compounds were screened at 10 μM with 2 assay replicates. Data was analysed in Genedata Screener. Compounds exhibiting speck/cell modulatory activity were identified, and compounds were additionally filtered for cytotoxicity. Using a high content end point allows multiple parameters to be measured and assessed, in this case the nuclei count per well for each test compound was plotted and those which were significantly lower than the mean population (by an AVG-3SD cut-off) were flagged as cytotoxic and removed from further study.

Figure 4 A: Primary Screen and B: Hit Potency phases. Compound dispensing using HighRes® automation with compound mapping tracked through Titian Mosaic.

An automated workflow was established to facilitate high-content phenotypic screening to identify NLRP1 modulators

Figure 5 Table of assay performance metrics of both instances

Figure 6 Distribution histogram of all well types of all primary screening data. Black= test compound wells, Teal= neutral control and Blue= Inhibitor control.

The primary screen identified 79 compounds that were re-tested in a concentration-response curve. The hit potency phase identified weak inhibitors and activators of NLRP1 inflammasome formation.

Phenotypic screen/Target deconvolution

There is a significant challenge in Target Deconvolution post-screen. However, it is possible to progress the medicinal chemistry optimization of phenotypic hits without target deconvolution. The FDA will approve a drug without a molecular target as long as it’s safe and effective, as exemplified by pirfenidone reaching the market without knowledge of the target (Nakazato et al., 2002). Pursuing compound optimisation without a direct read on the target can be time consuming due to poly-pharmacology and variable permeability which can confound the SAR and hamper progress.

In our case study, the next steps would be the deconvolution of the hits with initial removal of technical false positives. The screening cascade could involve screening hits in a cell line closer to the disease setting, i.e. patient-derived cells, no engineered proteins with a relevant disease stimulus. These hits, if potent enough, could then be used to deconvolute the target. Evaluating how known compounds behave in the screening cascade and establishing SAR around the hit to suggest specific target-based pharmacology and increased potency. Development of orthogonal assays and direct target engagement assays such as SPR or ASMS where possible.

There are a number of approaches that can be used for target deconvolution, and these can be broadly separated as follows:

• Chemical probe-based methods such as affinity based pull down
• Label free biophysical techniques methods as mentioned above SPR, ASMS or Thermal shift.
• Thermal stability profiling including CETSA are applicable in a cell-based environment and does not rely on target activity or reactivity, making it amenable to detecting allosteric binding interactions.

Bibliography

At Sygnature, we have the capability and expertise to run both target based and phenotypic high throughput screens. We can tailor a program to your project and requirements using our HIT SYNERGY platform, producing exciting leads to start the rational drug design for your target of interest.

If you have a hit identification project, then get in touch today.