Development of Fluorescent and Biotin Probes Targeting NLRP3

Extracellular signals drive the nucleation of the NLRP3 inflammasome which leads to the release of cytokines and causes inflammatory events. Hence, the inflammasome has gained enormous momentum in biomedical basic research. The detailed mechanisms of inflammasome generation and regulation remain to be elucidated. Our study was directed toward the design, convergent synthesis, and initial biochemical evaluation of activity-based probes addressing NLRP3. For this purpose, probes were assembled from a CRID3/MCC950-related NLRP3-binding unit, a linker portion and a coumarin 343 fluorophore or biotin. The affinity of our probes to NLRP3 was demonstrated through SPR measurements and their cellular activity was confirmed by reduction of the interleukin 1β release from stimulated bone marrow-derived macrophages. The initial characterizations of NLRP3-targeting probes highlighted the coumarin probe 2 as a suitable tool compound for the cellular and biochemical analysis of the NLRP3 inflammasome.

INTRODUCTION NLRP3 (NOD-, LRR-and PYD-containing protein 3) has attracted increasing attention as an important player in the (patho)physiology of inflammations. NLRP3 (or NALP3 or cryopyrin) is a tripartite protein of the family of nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs). It contains an N-terminal pyrin domain (PYD), a C-terminal leucine-rich repeat (LRR) domain, and a central NACHT domain. The NACHT domain exists in NAIP (neuronal apoptosis inhibitory protein), CIITA (MHC class II transcription activator), HET-E (incompatibility locus protein from Podospora anserine), and TP1 (telomerase-associated protein) proteins and consists itself of four subdomains, i.e., of a nucleotide-binding domain, the helical domain 1, a wingedhelix domain, and the helical domain 2 (Eldeeb et al., 2019;Vande Walle et al., 2019). The NLRP3 inflammasome represents a key mediator of the inflammatory response. Inflammasomes are intracellular supramolecular complexes. Their formation is triggered either by damage-associated (such as ATP) or pathogen-associated (such as nigericin) molecular patterns or by Toll-like receptor activation (Baldwin et al., 2016;Chauhan et al., 2020). NLRP3 activation is controlled by PYD and NACHT domain phosphorylation (Song et al., 2017;Stutz et al., 2017). In the NLRP3 inflammasome, NLRP3 acts as a sensor and forms a platform together with the adaptor protein ASC (apoptosis-associated speck-like protein containing a CARD) and the effector protein caspase-1. Crucial steps for NLRP3 inflammasome assembly are NLRP3 oligomerization and recruitment of ASC to NLRP3 oligomers. NLRP3 operates through caspase-1 activation which, in turn, results in the processing of cytokine pro-forms and the release of the maturated pro-inflammatory interleukins IL-1β and IL-18. Caspase 1-catalyzed cleavage of gasdermin D (GSDMD) generates a pore-forming N-terminal protein (GSDMD-N) initiates a lytic form of cell death, referred to as pyroptosis (Figure 1) (Dubois et al., 2019;Pfalzgraff and Weindl, 2019;Broz et al., 2020).
NLRP3 is associated with a variety of disorders such as central nervous system (CNS) diseases, rheumatoid arthritis, gout, atherosclerosis, asthma, and crystal nephropathy (Ludwig-Portugall et al., 2016;Dhana et al., 2018;Mangan et al., 2018;Chauhan et al., 2020). The cryopyrin-associated periodic syndromes (CAPS) are autoinflammatory disorders caused by various gain-of-function missense mutations of the NLRP3 gene (Mangan et al., 2018). The strong inflammatory role of NLRP3 provides the impetus for the development of drugs, acting as "NLRP3 inflammasome blockers" (Cocco et al., 2014). On the one hand, agonists could be useful to reverse the immunosuppressive conditions in tumors. On the other hand, inhibitors of the NLRP3 pathway are promising candidates for the treatment of several chronic and auto-inflammatory diseases, including those for which adequate therapies currently do not exist (Mangan et al., 2018). Accordingly, NLRP3 was long considered as a target for the development of small molecule inhibitors. A broad spectrum of NLRP3 inhibitors has been found, from natural products to small synthetic molecules. Cytokine-release-inhibitory-drug-3 FIGURE 1 | Schematic formation of the NLRP3 inflammasome and induction of pyroptosis.
(CRID3), a diarylsulfonylurea derivative, showed potent inhibition of the NLRP3 inflammasome formation (Coll et al., 2015). This compound CRID3 is also known as MCC950 or CP-456,773 (Figure 2). Due to reversible binding to the NACHT domain of wildtype NLRP3, ASC oligomerization could be blocked resulting in decreased IL-1β release. It was reported that CRID3/MCC950 binds in proximity to the Walker B motif in NLRP3, impeding the ATP hydrolysis and leading to an inactive NLRP3 conformation (Coll et al., 2019;Tapia-Abellán et al., 2019). Nevertheless, further research attempts are required to decipher NLRP3 inflammasome formation and activation. In this study, we developed new fluorescent and biotin-tagged activity-based probes which could be achieved through a convergent synthetic strategy. These probes are intended to be used, e.g., for fast competition assays, pulldown experiments or confocal microscopy.

RESULTS AND DISCUSSION
Covalent and non-covalent conjugation of proteins to fluorophores is commonly used to study their cellular localization, to discover protein-protein interactions, and to label proteins in their native environment. Fluorescence can be provided by means of a fluorescent low-molecular weight probe possessing sufficient affinity for the target protein. A widely used class of fluorophores, 7-aminocoumarins, are characterized by chemical and enzymatic stability, a small molecular size and large Stokes shifts (Breidenbach et al., 2020). Rigidization of the amino group due to cyclization resulted in increased quantum yields and restored fluorescence in aqueous media. A prominent example is coumarin 343 with an emission maximum at 480 nm in aqueous media after excitation at 440 nm, which is frequently employed to assemble activity-based probes (Terai and Nagano, 2008;Meimetis et al., 2014;Mertens et al., 2014;Kohl et al., 2015). The appendage of a biotin moiety to a bioactive probe offers the opportunity to elucidate and analyze intracellular binding partners of the probe. Biotinylation profits from the exceedingly strong interaction between biotin and either avidin or streptavidin (Trippier, 2013;Verdoes and Verhelst, 2016;Chakrabarty et al., 2019). Accordingly, in order to provide tool compounds to unravel hitherto unknown processes of inflammasome activation and dynamics, we designed NLRP3-specific probes with either coumarin 343 or biotin.
CRID3/MCC950 (Figure 2), the best-known and prevalently applied NLRP3 inflammasome inhibitor, constitutes a sulfonylurea with a hexahydro-s-indacen-4-yl group at the terminal nitrogen and a furan moiety at the sulfur atom (Vande Walle et al., 2019;Chauhan et al., 2020;Wu et al., 2020). 11 C-labeled CRID3/MCC950 was developed for non-invasive PET imaging studies (Hill et al., 2020a). CRID3/MCC950 was chosen as the template for the NLRP3 binding moiety for all of our probes. First, we had to identify an appropriate exit vector to connect either a fluorescent label or a biotin moiety to the NLRP3 ligand without losing entire binding affinity to the target. Recently published structure-activity relationships concerning diarylsulfonylurea-based NLRP3 inhibitors revealed the intact western indacene moiety of CRID3/MCC950 to be crucial for biological activity, and hence for target binding. On the other hand, structural alterations on the eastern sulfonamide part were better tolerated (Hill et al., 2017(Hill et al., , 2020bAgarwal et al., 2020).
Accordingly, our molecular design toward NLRP3 probes was based on the introduction of a molecular handle for attaching the label at the eastern part of the diarylsulfonylurea. A network of convergent synthetic routes was employed to assemble the envisaged probes. As depicted in Scheme 1, we synthesized an aromatic sulfonamide (8) containing a Boc-protected amino group to be used as exit vector at a later stage of the synthesis. Next, we transformed the hexahydro-s-indacen-4-amine 5 in the presence of triphosgene (BTC, [bis(trichloromethyl) carbonate] into the isocyanate 6. Following the pretreatment of sulfonamide 8 with sodium hydride, the deprotonated intermediate was reacted with isocyanate 6 to obtain the NLRP3 ligand 9. Compound 9 was deprotected to the trifluoroacetic acid (TFA) salt 10 prior to each coupling reaction.
For two final compounds, a short polyethylene glycol linker was chosen to connect the NLRP3 binding moiety and coumarin 343 or biotin. 2-(2-Aminoethoxy)ethanol (11) was Cbz-protected and the resulting alcohol 12 was O-alkylated with tert-butyl bromoacetate to obtain the orthogonally protected linker 13 (Scheme 1), which, in turn, was N-deprotected under palladium on carbon (Pd/C)-catalyzed conditions to 14 (Scheme 2).
To obtain the biotin-tagged probes, we first activated D-biotin with N-hydroxysuccinimid (NHS) to the active ester 20. This was reacted with the deprotected linker 14. After deprotection of 21, a further HATU coupling was performed to yield the biotin-tagged probe 3. Direct linkage of 20 with the NLRP3 ligand 10 delivered the short biotin-tagged probe 4.
Surface plasmon resonance (SPR) is a powerful and versatile spectroscopic method, but less commonly used for the analysis of low-molecular weight probes for functional proteins. SPR spectroscopy allows a real-time measurement and the labelfree detection of biomolecular interactions through a surfacesensitive response. In order to assess the suitability of our probes, they were subjected to an SPR analysis and compounds 1-4, together with CRID3/MCC950, were analyzed for their binding behavior to human NLRP3. For that purpose, recombinant biotinylated human NLRP3-PYD-NACHT domain protein derived from HEK293T cells was immobilized on a streptavidinfunctionalized sensor chip (sensor chip SA). After a stabilization period, the analytes, CRID3/MCC950 and the respective test SCHEME 1 | Synthesis of NLRP3 ligand 10 and polyethylene glycol linker 13.
compound, were injected on single channels using single-cycle kinetics (Figure 3 for compounds 2 and 4; for compounds 1 and 3, see Supplementary Figure 1). The SPR data (Table 1) revealed a dissociation constant K D of 322 nM for the fluorescently labeled compound 2 and a K D value of 31 nM for CRID3/MCC950. The dissociation constant K D for the fluorescently labeled probe with the PEG linker (1) was determined to be 813 nM, whereas the two biotinylated probes (3, 4) bound with K D values of 1,030 and 611 nM, respectively. These data confirm that the molecular probes 1-4 attached with fluorescent or biotin labels to the scaffold of CRID3/MCC950 bind to recombinant NLRP3 protein. The smaller compounds (2 and 4) exhibited lower dissociation constants than their linker-connected counterparts (1 and 3). To confirm that compound 2 and 4 have the same binding site as CRID3/MCC950, we injected two ligands 120 s after the first injection ( Figure 3D). No significant response change, indicative of a second binding site, was observed in either succession of the injections (left and right), which points to the assumption that CRID3/MCC950 and both compounds interact with the same binding site on the target protein NLRP3. Although CRID3/MCC950 possessed a higher affinity to NLRP3, our compounds bound sufficiently tight, making them promising candidates for a further characterization as suitable NLRP3 probes. However, the reversible binding mode may obstruct the use for Western blotting analysis.
Next, the probes were examined in cell culture experiments. First, we determined the cytotoxicity of probes 1-4 at different concentrations between 625 nM and 40 µM in immortalized murine bone marrow-derived macrophages (Figure 4). While biotin probes 3 and 4 were proved to be non-toxic up to 40 µM, and the coumarin probe 1 was non-toxic up to 10 µM, the coumarin probe 2 exhibited some cytotoxicity at concentrations higher than 2.5 µM.
The following experiments were performed with probes 1-4 in immortalized murine bone marrow-derived macrophages. Nigericin, an antibiotic carboxylate ionophore, was used as the activator of the NLRP3 inflammasome, owing to its ability to induce a potassium efflux. In addition, we addressed the absent in melanoma 2 (AIM2) inflammasome, a distinct inflammasome complex, which, upon assembly, can also activate caspase-1 and cause cells to release IL-1β (Hornung et al., 2009). For this purpose, AIM2 was stimulated by transfection with poly(deoxyadenylic-deoxythymidylic) acid [poly(dA:dT)] after pretreatment with LPS. A Homogeneous Time Resolved Fluorescence (HTRF) assay was employed to determine the release of IL-1β ( Figure 5). Probe 1 was used at 2.5 µM and 10 µM, i.e., with concentrations found to be non-toxic in the same cells. We observed a complete inhibition of IL-1β release at a concentration of 10 µM, but not at 2.5 µM. Probe 2 was only investigated at 2.5 µM due to toxic effects at higher concentration. Notably, this probe showed complete inhibition at 2.5 µM, just like CRID3/MCC950 at this concentration. The biotin-tagged probes 3 and 4, at 20 µM each, did not significantly affect the cytokine release (see Supplementary Figure 2). For reasons to be clarified, these cellular data only partially agree with the affinity of the entire subset of compounds 1-4 for NLRP3 as determined in the SPR assay. The coumarin-containing compound 2 at 20 µM also reduced AIM2-dependent IL-1β release initiated by poly(dA:dT) (data not shown), obviously due to cytotoxicity. However, at 2.5 µM, only a slightly reduced AIM2-dependent release of IL-1β was observed ( Figure 5). As expected, CRID3/MCC950 and probe 1 did not affect this process.
To assess pyroptosis, cytosolic lactate dehydrogenase (LDH) release was measured as a marker of cell lysis. In contrast to priming with lipopolysaccharide (LPS) alone, priming with LPS followed by treatment with nigericin and poly(dA:dT), respectively, clearly increased cell lysis. CRID3/MCC950 and probe 2, each at a concentration of 2.5 µM, as well as probe 1 at 10 µM prevented the NLRP3dependent, but not the AIM2-dependent cell death (Figure 5). This rescue effect can be attributed to the anti-pyroptotic activity of the compounds and was in agreement with the inhibitory potency of the investigated compounds regarding IL-1β release.
The most suitable probe, 2, was subjected to confocal microscopy for detection of NLRP3 in non-stimulated, immortalized murine macrophages overexpressing NLRP3 fused to mCitrine, a monomeric variant of the yellow fluorescent protein. The confocal images show that NLRP3 staining was colocalized with the yellow fluorescence ( Figure 6A). On the other hand, the presence of the NLRP3 fusion protein was a prerequisite for colocalization as shown in the negative control ( Figure 6B). This incipient result indicated a successful application of this coumarin-labeled NLRP3 ligand for future studies.
In conclusion, we designed and synthesized four probes based on the structure of the prototypical NLRP3 inhibitor CRID3/MCC950. An appropriate exit vector was established for the attachment of a fluorescent or biotin label. All of our probes were demonstrated to bind to the NLRP3 protein. Two coumarin-labeled probes were shown to reach the cytosolic target and to block the final cellular response. One of them, probe 2, is expected to serve as an appropriate tool compound in the field of inflammasome research which is currently ongoing in our laboratories.

General Experimental Procedures
Chemicals were purchased from ABCR, BLDpharm, Fisher Chemical, Sigma Aldrich and Tokyo Chemical Industry. Thin  Table 1. Data were fitted to a 1:1 binding model. (D) SPR sensorgrams of CRID3/MCC950, compound 2, and compound 4 injected in competition mode on the chip surface loaded with human NLRP3-PYD-NACHT. Either CRID3/MCC950 (left) or compound 2 or 4 (right) were provided in the first injection step, followed by a second injection step after 120 s, where equimolar concentrations of the second compound were injected in presence of the first compound.
TABLE 1 | Second-order on-rate constants for association (k a ), first-order off-rate constants for dissociation (k d ), and kinetic dissociation constants (K D ) for NLRP3 ligands.

Expression of Recombinant NLRP3 Protein
HEK293T cells were transiently co-transfected with human avi-FLAG-His10-NLRP3-PYD-NACHT domain (pIRES puro 3 vector) and BirA (pcDNA3.1 vector) for in-cell biotinylation at the lysine amino group of the Avi-tag. After 24 h expression, cells were chemically disrupted and avi-FLAG-His 10 -NLRP3-PYD-NACHT domain was isolated using anti-FLAG M2 affinity beads (Sigma Aldrich). Eluted protein was stored snap frozen in liquid nitrogen, aliquoted, and stored at −80 • C.

Cell Viability Assay
NLRP3 KO immortalized murine bone marrow-derived macrophages overexpressing mouse NLRP3-FLAG and human ASC-mCerulean were seeded in 96 well plates and incubated overnight. Cells were treated with the fluorescent probes (1 and 2) and the biotin-tagged probes (3 and 4) at different concentrations in Opti-MEM for 4 h. Subsequently, cell viability was assessed using CellTiter-Blue R cell viability assay according to the manufacturer's instructions.

Inflammasome and LDH Release Assays
NLRP3 KO Immortalized murine bone marrow-derived macrophages overexpressing mouse NLRP3-Flag and human ASC-mCerulean were seeded in 96-well plates and incubated for 16 h. Cells were treated with 400 ng/mL ultrapure LPS for 3 h. After priming, the probes or CRID3/MCC950 at various concentrations or DMSO were added to the cells for 30 min. Subsequently, to induce NLRP3 inflammasome activation, cells were treated with 10 µM nigericin for 1.5 h. As controls, to activate the AIM2 inflammasome, cells were transfected with 200 ng of poly(dA:dT) using 0.5 µL lipofectamine 2,000 per well for 4 h. Supernatants were collected to measure cytosolic LDH release by an LDH assay and IL-1ß release by Homogeneous Time Resolved Fluorescence (HTRF). All assays were performed according to the manufacturer's instructions. A two-tailed unpaired t-test was performed using GraphPad Prism 9.

Confocal Microscopy
Immortalized murine bone marrow-derived macrophages overexpressing NLRP3-mCitrine or mCitrine were seeded in Ibidi µ-slides. After 16 h, the fluorescent probe 2 was diluted in complete DMEM and added to the cells at 5 µM concentration for 30 min. Cells were fixed with 4% methanol-free formaldehyde for 10 min at room temperature. They were then washed three times with PBS and directly imaged with a Leica SP8 lightning confocal microscope.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
TK and MGü designed compounds and wrote the manuscript. TK synthesized compounds. KG and MM performed SPR experiments. AA and ML performed cellular assays and microscopy. All authors analyzed data. MGe, EL, and MGü supervised the study. TK and KG contributed equally. MGü conceived the study.