Discovery of Potent Coumarin-based Kinetic Stabilizers of Amyloidogenic Immunoglobulin Light Chains Using Structure-based Design

Abstract

In immunoglobulin light chain (LC) amyloidosis, transient unfolding, or unfolding and proteolysis enable aggregation of LC proteins causing potentially fatal organ damage. A drug that kinetically stabilizes LCs could suppress aggregation; however, LC sequences variable and have no natural ligands, hindering drug development efforts. We previously identified high-throughput screening hits that bind to a site at the interface between the two variable domains of the LC homodimer. We hypothesized that extending the stabilizers beyond this initially characterized binding site would improve affinity. Here, using protease sensitivity assays, we identified stabilizers that can be divided into four substructures. Some stabilizers exhibitnanomolar EC₅₀ values, a 3,000-fold enhancement over the screening hits. Crystal structures reveal a key π-π stacking interaction with a conserved tyrosine residue that was not utilized by the screening hits. These data provide a foundation for developing LC stabilizers with improved binding selectivity and enhanced physicochemical properties.

Graphical Abstract

Introduction

The processes of misfolding, aberrant proteolysis and aggregation, or misfolding and misassembly of human proteins results in the widespread degeneration of post-mitotic tissues and organ systems in the systemic amyloidoses, which eventually become fatal diseases if untreated^1–2. Small molecule “kinetic stabilizers”, which bind to and stabilize the native state of an aggregation-prone protein, have been shown to be clinically effective at stopping or slowing the progression of the systemic amyloidoses^{1, 3–5}, as well as other protein misfolding diseases^6–7. In the case of the transthyretin amyloidoses, the kinetic stabilizer tafamidis discovered by our group is now widely used to treat transthyretin polyneuropathy and cardiomyopathy^{1, 4–5, 8}.

Immunoglobulin light chain amyloidosis (AL) is one of the most commonly diagnosed systemic amyloid diseases^{2, 9–10}. Organ damage is associated with the process of aggregation of antibody light chain (LC) proteins, which are secreted by an expanded population of monoclonal plasma cells^9–10. LCs that cause amyloidosis are generally less stable than other LCs, which allows them to undergo some combination of misfolding, aberrant proteolysis and misassembly to form a range of non-native species, including amyloid fibrils^11–16. While it is established that these processes are associated with organ degeneration^17–19, the precise non-native LC structure(s) that cause this proteotoxicity remain unclear¹.

Antibodies normally contain two heavy chains and two light chains, in which each light chain engages in a heterodimeric association with each heavy chain. In AL, clonal plasma cells secrete “free”, homodimeric full-length (FL) LCs, with each LC protomer containing an N-terminal variable (V_L) domain and a C-terminal constant (C_L) domain, sometimes in the absence of antibody secretion²⁰. The LCs may be covalently joined by an inter-chain disulfide bind between the C214 residue of each C_L domain. AL-associated LCs more rapidly undergo conformational changes in comparison to non-amyloidogenic FL LC dimers^{13–14, 21}. Both amyloidogenic and non-amyloidogenic FL LC dimers adopt a common and well-defined “native” structure, although the function of LC dimers, if it exists, is unknown^22–24. Amyloid fibrils isolated from AL patients can be composed of FL LCs and/or fragments thereof; the latter are thought to arise from FL LC misfolding and aberrant endoproteolysis²⁵.

Each AL patient has a unique LC sequence, causing AL to be a somewhat heterogeneous disease, which creates major challenges for early AL diagnosis and treatment^26–27. Existing treatments typically consist of cytotoxic chemotherapy regimens that aim to eradicate the plasma cell clones (the proteasome inhibitor bortezomib being one common component)²⁸. Despite substantial progress in treating AL, many patients, especially those with cardiac involvement, are often too frail to tolerate these demanding treatment regimens. Emerging strategies for treating AL include the monoclonal antibody CAEL101 (Caelum), which promotes clearance of amyloid fibrils²⁹, and protein disulfide isomerase inhibitors that reduce the secretion of amyloidogenic LCs from plasma cells^30–31. The anti-CD38 antibody daratumumab, which induces apoptosis in plasma cell clones, met its endpoints in a recent Phase II AL trial and is the first drug specifically FDA-approved for AL (Jan. 2021)³².

Mechanistically novel AL treatments could further improve prognosis for a disease that still kills thousands of patients each year. One such approach is to use kinetic stabilizers, which are small molecules that bind to the LC dimer native state preferentially over the misfolding transition state. Thus, kinetic stabilizers slow the unfolding/misfolding steps that can directly lead to pathology^18–19, or slow misfolding leading to misassembly, with or without endoproteolysis, also linked to pathology³³. An ideal Fl LC kinetic stabilizer would render the unfolding/misfolding steps slow enough that newly synthesized FL LCs would be cleared before they could misfold, which would also preclude LC-seeded aggregation. However, if such a high kinetic barrier cannot be achieved by small molecule binding, these kinetic stabilizers would still reduce the equilibrium population of aggregation-prone LC species. The kinetic stabilizer strategy does not require a detailed understanding of the non-native LC structure(s) causing AL, as it lowers the concentration of all non-native structures^{1, 17}. Kinetic stabilizers, hereafter called stabilizers, would be complementary to the therapies that aim to eliminate the clonal plasma cells, as complete elimination of the clonal plasma cells and thus the amyloidogenic light chains is difficult to achieve. If administered yearly in the course of AL, stabilizers could potentially reduce cardiotoxicity sufficiently to allow the patient to tolerate cytotoxic chemotherapy. We also envision using a FL LC stabilizer drug that exhibits a benign safety profile for maintenance therapy, to enhance and prolong remission. Relapse after anti-plasma cell treatments is often associated with slowly progressing AL, which could be treated with a LC stabilizer³⁴.

Full length immunoglobulin LCs are “non-traditional” drug targets because they do not have enzymatic activity, natural ligands, or recognized binding pockets prior to our work. In addition, the amino acid sequences of FL LCs are very diverse, because they are antibody-derived proteins³⁵. These issues have caused LCs to be considered “undruggable” proteins and create significant challenges for designing small molecule FL LC stabilizers having the potential to become a new treatment for AL. A clinically successful LC stabilizer must satisfy the following criteria: (1) exhibit adequate affinity and selectivity for binding LCs versus other plasma proteins, especially albumin, because LCs are found in micromolar concentrations in blood²⁰ compared to albumin, present at millimolar concentrations; (2) be sufficiently safe to exist at micromolar concentrations in blood necessary for saturable LC binding; (3) demonstrate the ability to stabilize LCs with mulitple sequences; and (4) display good metabolic stability and an acceptable half-life, comparable to marketed small molecule drugs. By satisfying these criteria, FL LC•stabilizer complexes should be incompetent for misfolding, proteolysis, and aggregation, and have sufficiently long lifetimes to facilitate removal from the body, which occurs mainly through the kidneys³⁶.

Because FL LC dimers do not have known natural ligands, we previously conducted a high-throughput screen to identify FL LC small molecule stabilizers, using a destabilized amyloidogenic FL LC dimer called WIL-FL^{11, 37}. In this “protease-coupled fluorescence polarization” assay, the WIL-FL LC dimer, covalently labeled with a fluorescein dye, is incubated with the broad-spectrum protease proteinase K. Transient misfolding events in the WIL-FL LC create protease-sensitive conformations, enabling proteolysis that results in release of smaller fluorescein-labeled peptides exhibiting decreased fluorescence polarization. Small molecules that bind to and kinetically stabilize WIL-FL decrease the rate at which the LC is proteolyzed, and therefore slow the decrease of the initially high fluorescence polarization signal³⁷. Using this assay, we identified and validated five distinct FL LC stabilizer chemotypes from a screening library of approximately 650,000 small molecules. These hit chemotypes included coumarins (1 and 2), an aryl cyanoacrylamide (3), a diaryl hydrazone (4), a hydantoin (5), and a sulfone (6) (Figure 1)³⁷.

Figure 1.

Examples of FL LC stabilizers that were identified from our earlier high-throughput screen³⁷.

We previously obtained high-resolution crystallographic data on two structurally distinct stabilizers, coumarin 1 and hydantoin 5, bound to the JTO-FL LC dimer, which is a non-amyloidogenic LC that is structurally very similar to the amyloidogenic WIL-FL LC dimer, but more amenable to crystallization (Figure 2)^{11, 37–38}. Importantly, the aromatic substructures of both 1 and 5 occupy a hydrophobic pocket formed in the V_L-V_L domain interface of the FL LC dimer. Notably, this pocket is composed of residues conserved between most FL LCs, suggesting that it can be utilized to stabilize a variety of AL-associated FL LC sequences^37–38. The pocket is not observed in the unbound structures of JTO-FL³⁷ or most other FL LC dimers²² and may be formed through induced fit. The trifluoromethyl phenylacetamide substructure of 5 and the entirety of 1 bind this pocket (Figure 2). Notably, the hydantoin and isobutyl substructures of 5 extend into a nearby region that is unoccupied by 1. Despite these additional interactions that 5 forms with the FL LC dimer compared to 1 (Figure 2), both stabilizers have similar binding affinities and potencies, exhibiting dissociation constants on the order of 1–10 μM³⁷. However, lead molecules with nanomolar affinity for LCs are likely required to satisfy the above-mentioned requirements for a clinically successful LC stabilizer.

Figure 2.

Modular design of FL LC stabilizers. A) Representation of the JTO-FL LC dimer kinetic stabilizer binding site, showing two distinct pockets that can be occupied by ligand, the core hydrophobic pocket (left) and the distal aromatic pocket (right). A LC stabilizer that occupies both pockets can be divided into four different regions that are color-coded. Interactions of the LC stabilizer substructures with FL LC residues are shown. Potential hydrogen bonds between the kinetic stabilizer and LC protein are indicated with dashed lines. The prime labels for some residues denote residues comprising the second polypeptide chain of the LC homodimer. B) The A, B, C and D substructures are color-coded on the two LC stabilizers 1 and 5 that have published crystallographic data (PDB: 6MG5 and 6W4Y, respectively). The coumarin ring of 1 is numbered for reference. The isobutyl group of 5 serves only as a prototype “distal substructure” because it does not fully engage with Y49’. Only the R enantiomer of 5 is observed in the crystal structure, but the compound was used as a racemate. C) The color-coded substructures of stabilizers 1 and 5 are indicated in the superimposed crystal structures of JTO•1 (PDB: 6MG5) and JTO•5 (PDB: 6W4Y).

From these crystal structures and additional molecular modeling, we envisioned that more potent FL LC stabilizers could be composed of four substructures differentially colored in Figure 2, and depicted in the graphical abstract. Here, we used structure-based design to identify four substructures that complement the four FL LC dimer-binding sub-sites. We designed, synthesized, and tested 225 candidate LC stabilizers (Table S1). The “four-substructure hypothesis” afforded the highest affinity FL LC coumarin-based stabilizers known, with up to 3,000-fold improved potency compared to the screening hits, which is a critical milestone towards generating a lead LC stabilizer for the treatment of AL. These molecules will be carried forward into lead optimization with the end goal of identifying a preclinical candidate.

Results

A four-substructure design for FL LC kinetic stabilizers

From our existing crystallographic data and molecular modeling on structures of coumarin 1 and hydantoin 5 bound to the JTO-FL LC dimer (Figure 2)^37–38, we identified four substructures in the FL LC stabilizers that we hypothesized could be optimized either individually, or in pairs, to generate stabilizers with higher potency. The “anchor substructure” of a FL LC stabilizer (Figure 2A; red substructure A) occupies the “anchor cavity” or the deepest part of the core hydrophobic pocket. The anchor cavity comprises the side chains of residues Q38 and P44 from both protomers and the Y87 and Y36’ side chains from different protomers (residues are numbered according to the Kabat system, and the prime notation indicates the second polypeptide chain of the dimer). While residues lining the anchor cavity have hydrogen-bonding potential, all potential hydrogen bonds are satisfied by other residues in the FL LC dimer. The “aromatic core” of a FL LC stabilizer (Figure 2A; blue substructure B) occupies the solvent-facing region of the core hydrophobic pocket (the “aromatic slit”), made up principally by the Y87, F98, and P44’ side chains. The aromatic core can hydrogen-bond with the backbone of residue T46’. The “linker module” of a FL LC stabilizer (Figure 2A; green substructure C) occupies the “solvent-exposed linker binding site”, where the linker module hydrogen-bonds with the backbone of residues V96 and F98. The “distal substructure” of a FL LC stabilizer (Figure 2A; magenta substructure D) occupies the “distal aromatic pocket”, comprising conserved residue Y49’ and non-conserved residue 96 (Val in the case of JTO-FL). The ideal distal substructure should interact primarily with Y49’ and less so with V96 as the latter is not conserved.

Structure-activity relationships of 7-aminocoumarin derivatives

Our initial structure-activity relationship (SAR) studies aimed to optimize the “anchor substructure” in the context of the 7-aminocoumarin scaffold. We prepared a series of 7-aminocoumarin analogs, replacing one or both ethyl groups of 1 with hydrogen, methyl, n-propyl, isopropyl, propargyl, benzyl, etc., in various combinations (exemplified by 9 – 13 in Table 1; see Table S2 for a comprehensive list of the 45 analogs synthesized). Stabilizer potency was assessed using a protease sensitivity assay, which measures the cleavage of WIL-FL LC dimer in buffer and measures the fold protection of the FL LC dimer (2.5 μM) at 37 °C by proteinase K (50 nM) over 2 h at a single stabilizer dose (10 μM in 1% DMSO)^{14, 37–38}. We report “fold protection”, the fraction of un-proteolyzed FL LC dimer in the presence of small molecule corrected for the residual un-proteolyzed FL LC dimer in the vehicle (1% DMSO) control. We also employed a proteolysis-coupled fluorescence polarization (PCFP) assay, which was used initially in our high-throughput screen, with fluorescein-labeled WIL-FL, i.e., WIL-FL* (2.5 to 10 nM), to measure the concentration of each stabilizer at which 50% of the LC was protected from proteolysis (EC₅₀)³⁷. Most 7-dialkylamino analogs prepared exhibited a decrease in potency compared to the screening hit (1), i.e., the diethylamino coumarin. Substitution of one or both ethyl groups by propargyl (12 and 13, respectively) resulted in an ≈ 1.6-fold potency increase compared to 1; however, we decided that the potential metabolic liabilities of terminal alkynes³⁹ outweighed the small increase in potency. The synthetic efforts to optimize the anchor substructure and thus improve the potency of the 7-aminocoumarin series were generally unsuccessful (Table 1 and Table S2). However, later efforts on 7-ether coumarin analogs (see below) revealed that the anchor substructure is more tolerant of modification than suggested by our initial work on 7-aminocoumarin analogs.

Table 1.

SAR on dialkyl-aminocoumarins 9 to 13. Compound 1 is shown for comparison. Each data value is the mean of at least three replicates ± 1 standard deviation.

Compound	R1 =	R2 =	WIL-FL fold protectiona at 10 μM small molecule	WIL-FL* EC50 (nM)
1	Et	Et	0.28 ± 0.02	3240 ± 270
9	Et	Me	0.05 ± 0.00	18000 ± 1900
10	Et	n-propyl	0.11 ± 0.00	7890 ± 630
11	Et	isopropyl	0.10 ± 0.00	9700 ± 1200
12	Et	propargyl	0.26 ± 0.03	2650 ± 120
13	propargyl	propargyl	0.32 ± 0.01	2000 ± 320

^a“Fold protection” refers to the fraction of unproteolyzed FL LC dimer in the presence of small molecule, corrected for the residual unproteolyzed FL LC dimer in the vehicle (1% DMSO) control.

Design, SAR, and structural analysis of 3-substituted coumarin-carbamate-distal substructure derivatives

We recently reported the structure of the 3,5-substituted hydantoin 5 (Figure 2B) bound to a FL LC dimer (Figure 2C)³⁸. Notably, the hydantoin ring extends beyond the binding site mapped out by the simple coumarin hit 1 (Figure 2C). The hydantoin linker module hydrogen bonds with both LC monomers simultaneously. The isobutyl substructure at the hydantoin 5-position serves as a prototype “distal substructure”, and while the isobutyl substructure does not fully engage with Y49’, larger distal substructures could do so. We hypothesized that a distal substructure terminating in an aromatic ring could engage in π-π stacking with Y49’. This so-called “distal aromatic pocket” partially comprising Y49’ is more solvent-exposed than the aromatic slit and is envisioned to be more amenable to structural elaboration to increase affinity. To test this hypothesis, we first chose an anchor substructure and aromatic core pair to which we could attach diverse linker modules and distal substructures. The trifluoromethyl phenylacetamide anchor substructure and aromatic core of 5 by itself (i.e., with the hydantoin and isobutyl substructures removed, 5a) was inactive. Removal of the distal isobutyl substructure of 5, while retaining the hydantoin linker module, also led to a sharp drop in potency (5b, Table S3). Therefore, we decided to use the diethylamino anchor substructure and the coumarin aromatic core pair of 1, which lacks a linker module and distal substructure. Using 1 to optimize the linker module and distal substructure pairing is ideal because its potency as a FL LC stabilizer is modest (EC₅₀ ≈ 3300 nM³⁷).

To identify a linker module complementary to the V_L-V_L domain interface of the FL LC dimer, we initially focused on carbamate-based linkers rather than the hydantoin of 5 attached to 1 at the 3-position. 3-carbamate-substituted coumarins are accessible by a one-step synthesis from a common intermediate, greatly facilitating synthesis of many linker module–distal substructure pairs attached to 1. Carbamates joined to the coumarin with a two-carbon spacer were likely to achieve optimal geometry for hydrogen-bonding to the backbone of residues F98 and V96, analogous to the hydrogen bonding displayed by the hydantoin linker module³⁸.

The distal aromatic pocket of JTO-FL, with which we aimed to form stabilizing interactions, comprises residues R94, N95, and V96 from the non-conserved CDR3 loop of one protomer, and the conserved Y49’ of the other protomer. Y49’ appears in approximately 86% of AL-associated sequences, including JTO. Residue 49’ is a phenylalanine in 10% of sequences, including that of WIL^{35, 38}. We hypothesized that if the distal substructure could make strong interactions primarily with the aromatic sidechain of residue 49’ and less so with the non-conserved CDR3 residues, it would result in FL LC stabilizers that bind with high affinity to the majority of AL-associated FL LC sequences. Therefore, we evaluated a large series of distal substructures positioned by a carbamate linker module. Most distal substructures tested featured a basic nitrogen and/or an aromatic ring(s), to enable cation-π or π-π stacking interactions with Y49’ or F49’ (exemplified by 18 – 27, Table 2; see Table S4 for a list of all 78 analogs synthesized and evaluated). Piperidine (18) or pyridine (20, 21) as distal substructures resulted in up to an 8-fold increase in potency (EC₅₀) in comparison to 1. A further increase in potency (up to 22-fold compared to 1) could be achieved by utilizing a meta-acetanilide (23) or meta-benzamide (25) as the distal substructures. We also evaluated a meta substituted-biaryl subseries. In this biaryl context, an imidazole (26) or 2-pyridone (27) as the terminal ring (Table 2) afforded the largest gains in potency–a 42-fold increase compared to 1 for 26 (EC₅₀ = 77.1 nM). Additional meta substituted-biaryls were also active stabilizers (e.g., S92-S104; Table S4), suggesting that this secondary carbamate linker module and distal substructure pairing is a general solution for designing potent FL LC stabilizers. Methylation of the carbamate nitrogen of the linker module of 21 affords 22 (EC₅₀ = 5280 nM), which exhibits 70-fold reduced potency compared to the secondary carbamate 21, an EC₅₀ below even that of 1. All tertiary carbamates tested had poor to modest activity, including several N-cyclic carbamates (S57-S69, Table S4), indicating that a hydrogen bond donor in the linker module is likely important for activity.

Table 2.

SAR on coumarin-carbamate analogs 18 to 27. Compound 1 is shown for comparison. Each data value is the mean of at least three replicates ± 1 standard deviation.

Compound	R1 =	R2 =	WIL-FL fold protection at 10 μM small molecule	WIL-FL* EC50 (nM)	WIL-FL* T46L/F49Y EC50 (nM)
1	-	-	0.28 ± 0.02	3240 ± 270	6880 ± 880
18	H		0.47 ± 0.01	1690 ± 110	8800 ± 2440
19	H		0.34 ± 0.01	302 ± 79	317 ± 32
20	H		0.63 ± 0.02	415 ± 47	944 ± 200
21	H		0.54 ± 0.02	1110 ± 110	3680 ± 1470
22	Me		0.14 ± 0.02	5280 ± 210	>20000
23	H		0.71 ± 0.01	289 ± 19	194 ± 56
24	H		0.56 ± 0.02	816 ± 41	1170 ± 480
25	H		0.80 ± 0.03	150 ± 13	228 ± 75
26	H		0.91 ± 0.04	77.1 ± 4.6	27.9 ± 9.4
27	H		> 0.99	151 ± 20	26.2 ± 5.1

Sequence analysis of the WIL-FL binding site indicates that two residues, T46’ and F49’, predicted to interact with the stabilizers featuring the aforementioned carbamate linker module and distal substructures, are not the consensus residues amongst AL sequences³⁸. Therefore, we constructed the WIL-FL T46L/F49Y sequence, introducing the residues found in the majority of amyloidogenic FL LCs. Comparison of stabilizer EC₅₀ values between the native WIL-FL and WIL-FL T46L/F49Y indicates that these two amino acid substitutions generally do not greatly affect the potency of the stabilizers in Table 2. The more potent stabilizers 26 and 27 exhibit up to a six-fold potency increase with the WIL-FL T46L/F49Y consensus sequence in comparison to the parent WIL-FL sequence (e.g., for 27, WIL-FL EC₅₀ = 151 nM vs. WIL-FL T46L/F49Y EC₅₀ = 26.2 nM). Importantly, the T46L/F49Y mutations do not affect the inherent kinetic stability of the LC, as both the parent and mutant WIL-FL sequences are proteolyzed to a similar extent (Figure S1). Thus, the higher mutant potency is most likely due to the consensus amino-acid side chain changes enabling higher affinity stabilizer binding.

To understand the structural basis for the improved potency of the coumarin-carbamate-distal substructure series in comparison to 1, we obtained a 2.0 Å resolution crystal structure of JTO-FL (the crystallization-amenable FL LC dimer we used previously, which contains the Y49 consensus residue^37–38) (Table 3) in complex with carbamate 26, comprising a phenyl-imidazole distal substructure (Figures 3 and S2).The coumarin core of 26 is nearly aligned with 1 from our earlier structure³⁷, indicating that the presence of the linker module and the distal substructure of 26 does not affect the binding mode of the anchor substructure and the aromatic core. The carbamate linker of 26 forms hydrogen bonds with the backbone of residues F98 and V96. Notably, the carbamate is in the higher-energy syn conformation, which directs the phenyl-imidazole moiety towards Y49’. This substructure interacts with Y49’ likely using imperfect π-π stacking interactions, with a 4.5 Å distance between the centers of the imidazole and Y49’ phenol.

Table 3.

Crystallography data collection and refinement statistics.

Structure	JTO-FL•26	JTO-FL•34	JTO-FL•36	JTO-FL•62	JTO-FL•63
PDB code	7LMN	7LMO	7LMP	7LMQ	7LMR
Data collection
Beamline	ALS 5.0.3	APS-23-ID-B	APS-23-ID-B	ALS 5.0.3	APS-23-ID-B
Space group	P212121	P212121	P212121	P212121	P212121
(a, b, c) (Å)	64.03, 82.13, 95.45	63.88, 82.85, 95.38	64.30, 81.72, 95.00	64.13, 81.09, 96.30	63.74, 81.94, 95.83
(α, ß, γ) (°)	90, 90, 90	90, 90, 90	90, 90, 90	90, 90, 90	90, 90, 90
Resolution range (Å)	44.74 – 2.01 (2.12 – 2.01)	47.69 – 1.99 (2.10 – 1.99)	47.50 – 2.29 (2.42 – 2.29)	44.59 – 1.91 (2.01 – 1.91)	47.91 – 2.09 (2.20 – 2.09)
Unique reflections	34,240 (4,798)	35,344 (5,011)	23,063 (3,282)	39,458 (5,381)	30,392 (4,309)
Completeness (%)	99.6 (97.2)	99.8 (98.7)	99.9 (99.1)	99.1 (94.1)	99.7 (98.1)
Rsym	0.091 (0.720)	0.117 (1.008)	0.130 (1.033)	0.096 (0.344)	0.085 (0.649)
Rpim	0.033 (0.273)	0.035 (0.298)	0.037 (0.278)	0.037 (0.132)	0.034 (0.251)
CC(1/2)	1.00 (0.85)	1.00 (0.85)	1.00 (0.82)	1.00 (0.95)	1.00 (0.98)
I/σ (I)	14.1 (2.8)	13.3 (2.5)	13.9 (2.3)	12.6 (4.8)	12.3 (2.9)
Redundancy	7.8 (7.8)	12.3 (12.4)	13.3 (13.6)	7.7 (7.5)	7.2 (7.6)
Wilson B factor (Å2)	28	22	25	17	30
Refinement
Resolution range (Å)	44.78 – 2.01	47.74 – 1.99	47.54 – 2.29	44.63 – 1.91	47.96 – 2.09
No. reflections - work	32,449	33,544	21,851	37,376	28,513
No. reflections - free	1,645	1,656	1,100	1,916	1,464
Rwork	0.187	0.176	0.203	0.161	0.202
Rfree	0.239	0.225	0.264	0.203	0.260
RMS bond length (Å)	0.009	0.010	0.008	0.012	0.009
RMS bond angle (°)	1.59	1.54	1.56	1.72	1.58
Mean B value (Å2)
overall	41	39	40	27	50
protein	40	38	40	25	49
water	47	46	41	39	51
buffer	76	62	70	44	93
ligand	50	35	34	27	57
Ramachandran favored (%)	95.5	96.0	96.2	95.8	95.8
Ramachandran allowed (%)	100.0	100.0	100.0	100.0	100.0
Clashscore	4.14	3.37	1.85	3.05	3.38
No. atoms
total	3,686	3,740	3,593	3,863	3,577
protein	3,299	3,308	3,288	3,314	3,297
water	342	385	257	511	241
buffer	10	10	10	10	10
ligand	35	37	38	28	29

Figure 3.

Crystal structure of carbamate stabilizer 26 (magenta) in complex with JTO-FL (PDB: 7LMN). A) Line drawing of 26. B) Comparison of the binding modes of the “anchor substructure” and the “aromatic core” of 1 (orange) and 26 (magenta; PDB: 6MG5). C) Interactions between the carbamate linker module and distal substructure of 26 with the solvent-exposed linker binding site and the distal aromatic pocket. Potential hydrogen-bonding interactions and measured distances are shown in black dashed lines. D) A focused view of the distal substructure-Y49’ interaction.

Design, SAR, and structural analysis of 3-substituted coumarins with rigidified linker modules

Based on the crystallographic data, we hypothesized that further increases in potency could be achieved by rigidifying the linker module, to lock it into a binding-competent conformation. This change would reduce the entropic cost of binding associated with an enthalpically costly isomerization to the syn carbamate conformation, which is approximately 2 kcal/mol higher in energy than the trans conformation according to density functional theory (DFT) calculations. Therefore, we revisited the hydantoins as potential linker modules, evaluating both the spiro-hydantoin-pyrrolidine and the spiro-hydantoin-piperidine series. We hypothesized that the rigid spirocyclic linker modules could direct the distal group towards Y49’ with minimal entropic penalty. Using the pyrrolidine and piperidine nitrogens as functionalization handles to install different distal substructures, we evaluated a variety of aromatic amides, sulfonamides, and N-aryl substructures (exemplified by stabilizers 34 – 46 of both series, Table 4; see Table S5 for a full list of the 41 spiro-hydantoins synthesized and evaluated). In this context, aromatic amide and sulfonamide distal substructures resulted in up to 11-fold increase potency compared to 1, but still below that of the carbamate-phenyl-imidazole 26 (Table 2). In contrast, N-aryl distal substructures, in the context of the spiro-hydantoin piperidine linker module, resulted in up to 230-fold gains in potency. Stabilizers with certain nitrogen-rich bicyclic aromatic distal substructures—purine 43, pyrazolo-pyrimidine 44, and imidazo-pyrazine 45—afforded EC₅₀ values of 46 nM, 124 nM, and 29.8 nM, respectively, for the WIL-FL T46L/F49Y variant.

Table 4.

SAR on spiro-hydantoins 34 to 46. Each data value is the mean of at least three replicates ± 1 standard deviation.

Compound	n =	R =	WIL-FL fold protection at 10 μM small molecule	WIL-FL* EC50 (nM)	WIL-FL* T46L/F49Y EC50 (nM)
34	1		0.69 ± 0.01	855 ± 80	729 ± 150
35	1		0.40 ± 0.01	2600 ± 170	3960 ± 480
36	2		0.46 ± 0.01	1880 ± 90	1430 ± 390
37	2		0.27 ± 0.02	3970 ± 210	2700 ± 830
38	1		0.78 ± 0.04	207 ± 22	648 ± 210
39	1		0.78 ± 0.02	313 ± 54	4370 ± 830
40	1		0.62 ± 0.03	1370 ± 260	688 ± 170
41	1		0.74 ± 0.01	551 ± 73	355 ± 83
42	2		0.85 ± 0.01	292 ± 49	165 ± 52
43	2		> 0.99	110 ± 19	46 ± 12
44	2		> 0.99	103 ± 10	124 ± 41
45	2		> 0.99	62.4 ± 14	29.8 ± 8.1
46	2		0.85 ± 0.05	71.1 ± 15	55.2 ± 7.9

We solved crystal structures of JTO-FL in complex with stabilizers harboring a spiro-hydantoin pyrrolidine (34) and spiro-hydantoin piperidine (36) linker module (Figure 4 and Figure S2), each with an imidazole amide distal substructure, at 2.0 Å and 2.3 Å resolution, respectively. In each structure, the hydantoin forms hydrogen bonds with the F98 and V96 residues of the polypeptide backbone. For 36, the spiro-hydantoin piperidine linker module adopts a “downwards chair” conformation (Figure S3) which directs the imidazole sidechain to stack directly upon Y49’. For 34, the spiro-hydantoin pyrrolidine adopts a “chair-sideways” (Figure S3) instead of downwards conformation. The orientation of the distal substructure amide carbonyl in 34 is reversed compared to that in 36, but the distal imidazoles occupy a similar location near Y49’. Notably, the observed conformation of the spiro-hydantoin piperidine appears to be strained, because the C4 carbonyl of the hydantoin engages in significant 1,3-diaxial interactions with the piperidine chair. We hypothesized that this apparently high-energy conformation was likely limiting the potential potency increase from rigidifying the linker module.

Figure 4.

Crystal structures of JTO-FL in complex with spiro-hydantoin stabilizers 34 and 36. Potential hydrogen-bonding interactions with measured distances are shown in black dashed lines. A) Line drawings of 34 (R enantiomer) and 36. Only the R enantiomer of 34 is observed in the electron density (see Figure S2), but the compound was used as a racemate. B) JTO-FL•34 crystal structure (with stabilizer colored yellow, PDB: 7LMO). C) JTO-FL•36 crystal structure (with stabilizer colored light blue, PDB: 7LMP). D) Overlay of JTO-FL•34 and JTO-FL•36 structures with the coloring scheme as in panels B and C.

If this hypothesis has merit, removal of the C4 carbonyl would allow the active conformation of the spiro-hydantoin piperidine linker module to become more thermodynamically favored due to the absence of additional 1,3-diaxial interactions. We performed DFT calculations that predicted the energy difference between each chair conformer of the spiro-hydantoin piperidine linker module. In contrast to our expectations, the DFT calculations predicted that the active, “chair-down” conformation of the spiro-hydantoin would be energetically favored over the “chair-sideways” conformation by 1.3 kcal/mol in the gas phase (0.7 kcal/mol with implicit water) (Figure S3).

To experimentally scrutinize this hypothesis, we assessed the spiro-urea-piperidine linker modules in the context of 3 different N-aryl distal substructures corresponding to stabilizers 51 – 53 (Table 5) that have a CH₂ group in place of a carbonyl in the analogous spiro-hydantoin-piperidine series (i.e., compounds 42, 44, and 45 respectively; Table 4). Because most of the stabilizers (10 μM) in Table 5 provided complete protease K resistance for WIL-FL, we utilized the destabilized WIL T46L/F49Y/C214S variant to distinguish between these more potent stabilizers. The C214S mutation removes the C_L-C_L interchain disulfide bond, destabilizing the FL LC dimer and greatly increasing its protease susceptibility¹⁴. We also used the PCFP assay employing fluorescein-labeled WIL-FL or fluorescein-labeled WIL-FL T46L/F49Y dimer (2.5 nM). In each case, the spiro-urea stabilizer exhibits up to a 26-fold higher potency than the analogous spiro-hydantoin compound (Table 5). Notably, the spiro-urea stabilizers containing a pyrazolo-pyrimidine (52) or imidazo-pyrazine distal substructure (53) exhibit the highest potency of any stabilizers in the spiro series (WIL T46L/F49Y; EC₅₀ = 4.7 and 4.9 nM, respectively) and are the first FL LC stabilizers to exhibit single-digit nanomolar EC₅₀ values in any assay. The higher potencies of the spiro-urea linker module compared to the spiro-hydantoin may not only be due to conformational energy differences of the piperidine and may involve other factors such as desolvation differences, and intramolecular interactions between the hydantoin and the coumarin carbonyl, although we have not further investigated these factors. We also evaluated a variety of other bicyclic aromatic distal substructures in the context of the spiro-urea-piperidine linker module, but these 20 compounds exhibited markedly lower potencies than 51 – 53 (S125 – S153, Table S5).

Table 5.

SAR on spiro-hydantoins and spiro-ureas utilizing the destabilized WIL-FL T46L/F49Y/C214S mutant. Each data value is the mean of at least three replicates ± 1 standard deviation.

Compound	X =	R =	WIL-FL fold protection at 10 μM small molecule	WIL-FL T46L/F49Y/C214S fold protection at 10 μM	WIL-FL* EC50 (nM)	WIL-FL* T46L/F49Y EC50 (nM)
42			0.85 ± 0.01	< 0.01	292 ± 49	165 ± 52
44			> 0.99	< 0.01	103 ± 10	124 ± 41
45			> 0.99	0.04 ± 0.03	62.4 ± 14	29.8 ± 8.1
51			> 0.99	0.18 ± 0.03	48.1 ± 26	13.5 ± 1.7
52			> 0.99	0.03 ± 0.01	8.8 ± 0. 7	4.7 ± 0.4
53			> 0.99	0.41 ± 0.07	35.6 ± 2.7	4.9 ± 1.0

Design, SAR, and structural analysis of 7-substituted coumarin ethers

We next returned to the anchor substructure, as the diethylamino substructure of 1 could be prone to N-dealkylation in vivo, resulting in potentially toxic metabolites via subsequent oxidation to nitroso compounds⁴⁰. Moreover, since we had not successfully optimized the diethylamino anchor substructure, we considered 7-ether anchor substructures similar to 7-(1)-phenylethoxy coumarin (2; Table 5), a validated hit from the high-throughput screen (Figure 1)³⁷. We were unable to dock stabilizer 2 into the “anchor cavity” and the “aromatic slit” created by the binding of 1 to the FL LC dimer JTO-FL³⁸, but hypothesized that conformational flexibility of the LC would allow 2 to bind analogously to 1. We prepared a series of ether analogs of coumarin 2 (exemplified by stabilizer candidates 55 to 59 (Table 6; for a comprehensive list of the 33 coumarins with 7-ether anchor substructures tested, see Table S6). We initially focused on inserting methylene groups at different positions of the original 7-phenylethoxy anchor substructure from 2. Many analogs in this series have similar or slightly increased potency compared to 2 (WIL-FL T46L/F49Y EC₅₀ = 3960 nM). Notably, insertion of a methylene between the ether oxygen and chiral carbon of 2, affording 57, resulted in a 9-fold potency increase (444 nM) relative to 2 (Table 6).

Table 6.

SAR on coumarin ethers 55 – 59. Each compound was tested as a racemic mixture. Each data value is the mean of at least three replicates ± 1 standard deviation.

Compound	R =	WIL-FL fold protection at 10 μM small molecule	WIL-FL* EC50 (nM)	WIL-FL* T46L/F49Y EC50 (nM)
2		0.24 ± 0.01	1960 ± 160	3960 ± 1600
55		0.12 ± 0.01	3960 ± 1100	1670 ± 310
56		0.14 ± 0.01	2970 ± 310	1380 ± 270
57		0.40 ± 0.00	558 ± 84	444 ± 40
58		0.27 ± 0.01	N.D.a	N.D.a
59		0.18 ± 0.03	2860 ± 270	2580 ± 210

^aN.D. = not determined. Poor solubility of 58 at concentrations past 10 μM prevented the determination of an accurate dose-response curve.

We were unable to obtain a crystal structure of JTO-FL in complex with 2 using our existing method of soaking the stabilizer into preformed FL LC dimer crystals. We attributed this to the low aqueous solubility of 2 caused by its high lipophilicity. Therefore, we synthesized analogs of 2 and 57 that contained a solubilizing N-methyl morpholine substituent at the 3-position (i.e., stabilizers 62 and 63, respectively; see Figure 5A for their structures), which also slightly increased their potency (Table S6). We successfully solved crystal structures of 62 and 63 in complex with JTO-FL at 1.9 Å and 2.1 Å resolution, respectively (Figures 5 and S2), revealing that the ether and the diethylamino substructures both bind in the “anchor cavity”. In the binding of both 62 and 63 to JTO-FL relative to 1, we observed that the anchor cavity expands to be complementary to the 7-(1)-phenylethoxy anchor substructure. Minor conformational changes in residues lining the anchor cavity enable this expansion (Figure 5B). First, the Q38 sidechain shifts downwards, away from the cavity entrance, in both the JTO•62 and JTO•63 crystal structures compared to JTO•1, extending the cavity by at least 1.2 Å. In the other protomer, the Y87’ aromatic ring rotates by 12°, further enlarging the cavity. Our initial rigid-body docking experiment did not allow for these conformational changes, explaining why 2 could not be docked in the JTO-FL•1 “anchor cavity” and “aromatic slit” with 1 removed.

Figure 5.

Crystal structures of JTO-FL in complex with coumarin stabilizers 62 (green, PDB: 7LMQ) and 63 (blue, PDB: 7LMR), harboring ether anchor substructures. A) Line drawings of 62 (R enantiomer) and 63 (S enantiomer), which are the only stereoisomers observed in the electron density (Figure S2), although the stabilizers were synthesized and tested as racemates. B) Expansion of the anchor cavity is observed in the presence of an ether-containing anchor substructure also comprising a phenyl ring (for example, 62) compared to a diethylamino substructure (1, orange, PDB: 6MG5). Residues Q38 and Y87’, which undergo conformational changes enabling the cavity enlargement, are shown. C-E) Close-ups of the core hydrophobic cavity of JTO-FL with an ether bound in the anchor cavity. Residues that comprise the binding site, consisting of the “anchor cavity” and “aromatic slit”, are labeled. C) JTO-FL•62 D) JTO-FL•63 E) Overlay of JTO-FL•62 and JTO-FL•63 F) Alignment of 1, 62, and 63 in their binding poses in JTO-FL from the crystal structures.

The phenyl ring of each ether-based anchor substructure interacts with the P44 residues on each monomer and is positioned upon the Y87’ sidechain and Q38-Q38’ bridge at the base of the cavity. The methyl group of each anchor extends towards Y36. For 62, only the R enantiomer can be observed in the electron density, suggesting that binding may be stereoselective (Figure S2). Based on the electron density for 63, the opposite (S) enantiomer appears to preferentially bind. This difference is likely a consequence of the additional methylene group in 63 altering the conformation of the aliphatic portion of the anchor substructure, even though the methyl groups for both stabilizers occupy similar regions in the pocket (Figures 5C–5E). In the case of 62, the coumarin core is rotated slightly compared to 1 due to the bulkier anchor substructure compared to the diethylamino of 1. In the case of 63, the phenyl ring of the 2-phenylpropoxyl substructure occupies a similar location as in 62; however, the additional methylene in 63 causes the coumarin scaffold to be displaced by approximately 1.5 Å toward the solvent-exposed linker binding site (Figure 5F). The orientation of the coumarin ring relative to F98 reveals the coumarin comprising 63 is moved half the width of a benzene ring toward the solvent-exposed linker binding site (Figures 5E and 5F). The binding position of the aromatic core relative to F98 is therefore dependent on the identity of the anchor substructure.

Substructure combination results in highly potent FL LC stabilizers.

Having validated the alternative 1-phenylethoxy or 2-phenylpropoxy ether anchor substructures, we synthesized three analogs (83 – 85; Table 7) of the highly potent spiro-urea piperidine 53 (Table 5), wherein we replaced the diethylamino anchor substructure with the 1-phenylethoxy or 2-phenylpropoxy ether anchor substructures. The spiro-urea-piperidine-imidazo[1,2-a]pyrazine linker module and the imidazo-pyrazine distal substructure were common to all analogs (83 – 85, Table 7). In the case of the 2-phenylpropoxy anchor, we also tested an analog containing a truncated, 1-carbon spacer between the coumarin and linker module (85), considering that the longer anchor substructure may shift the remaining regions of the stabilizer and likely affect binding to LCs. Each analog was tested as isolated enantiomers. The presence of the (R)-1-phenylethoxy anchor substructure ((R)-83) afforded the most potent FL LC stabilizer identified to date (WIL T46L/F49Y EC₅₀ < 2.5 nM; Figure S4). The opposite enantiomer ((S)-83) was slightly less potent (EC₅₀ = 5.5 nM), but still highly active. It is of note that the EC₅₀ of (R)-83 is at the limit of the working range for the PCFP assay, which is defined by the concentration of fluorescein-labeled WIL-FL dimer (2.5 nM). Therefore, it is not possible to distinguish between more potent stabilizers using these assay conditions.

Table 7.

SAR on coumarin ethers 83 – 85 with a potency-increasing substituent at the 3-position. The individual enantiomers of each compound were evaluated separately. Each data value is the mean of at least three replicates ± 1 standard deviation. The absolute configurations of 84a, 84b, 85a, and 85b were not determined.

Compound	R1 =	n =	WIL-FL T46L/F49Y/C214S fold protection at 10 μM small molecule	WIL-FL* EC50 (nM)	WIL-FL* T46L/F49Y EC50 (nM)
(R)-83		2	0.63 ± 0.06	4.1 ± 1.0	<2.5
(S)-83		2	0.34 ± 0.08	25.2 ± 1.4	5.5 ± 0.3
84a		2	0.43 ± 0.02	8.6 ± 0.7	8.3 ± 0.8
84b		2	0.43 ± 0.03	11.5 ± 0.6	3.9 ± 0.2
85a		1	< 0.01	299 ± 3.5	187 ± 13
85b		1	< 0.01	477 ± 23	143 ± 9.9

The differential activities of each enantiomer are consistent with the crystallographic data for JTO•62, which suggests that the R enantiomer is preferred, although this does not translate to strongly enantioselective binding. Both enantiomers of the 2-phenylpropoxy stabilizer (84a and 84b; absolute configuration not determined) were similar in potency, even though the crystal structure of JTO•63 suggests preferential binding of the S enantiomer. Stabilizers containing the 2-phenylpropoxy anchor substructure and a truncated (1 methylene) spacer (85a, 85b) exhibited a significant decrease in potency (EC₅₀ > 100 nM) compared to the two-methylene analogs (EC₅₀ < 10 nM).

To verify that these highly potent stabilizers do not inhibit proteinase K, we assessed if they could protect the WIL-FL F98D variant from endoproteolysis by proteinase K. The F98D mutation disrupts the dimer interface⁴¹ and removes the critical π-π stacking interaction with the aromatic core. We have previously shown that the F98D mutation obviates binding to 1³⁷. Stabilizers 27, 53, and (R)-83 provided no appreciable proteinase K resistance to WIL-FL F98D (Figure S5), indicating that the compounds still require stacking with F98 to kinetically stabilize WIL-FL and do not inhibit proteinase K.

Thus far, we have been unable to obtain crystal structures of 53 or any of the highly potent coumarin ethers (e.g., 83 or 84) in complex with JTO-FL. Therefore, we turned to computational docking to gain structural insight into their binding, as we did in our previous study for the high-throughput screening hits³⁸. We docked the most potent stabilizer to date, (R)-83, using the FL LC dimer binding site conformation present in JTO•62, which has the enlarged anchor cavity (Figure 6). We recapitulated the crystal structure pose of the (1)-phenylethoxy anchor substructure and coumarin aromatic core (Figure 6A), with a root mean square deviation (RMSD) of 0.6 Å, comparing only the two substructures. As expected, the spiro-urea linker module extends towards the terminal aromatic binding site and forms hydrogen bonds with the peptide backbone of residues V96 and F98 (Figure 6B). The conformation of the linker module is similar to that in the crystallized pose of 36, which has a hydantoin linker module instead of a cyclic urea. The imidazo-pyrazine distal substructure could adopt a variety of different conformations exhibiting π−π stacking with Y49’, suggesting that this group may be partially mobile in the binding pocket.

Figure 6.

Docking model of stabilizer (R)-83 in complex with JTO-FL, using the protein conformation observed in the crystal structure of JTO•62. A) Line drawing of (R)-83. B) Comparison of a docked pose of (R)-83 (red) with the crystallized pose of 62 (light gray, PDB: 7LMQ), emphasizing the anchor substructures and aromatic cores. C) Comparison of a docked pose of (R)-83 with the pose of 36 from the crystal structure (dark gray, PDB: 7LMP), emphasizing the linker modules and distal substructures.

To strengthen our confidence in the docked poses of the imidazo-pyrazine distal substructure, we carried out molecular dynamics (MD) simulations of the JTO•53 complex, which differs from (R)-83 only in the anchor substructure (Figure S6, Table S7). Three distinct docked poses were selected to be the starting coordinates of a set of MD simulations. Each of the starting coordinates was simulated both with and without restrains on the protein α-carbons (the restraints prevent the protein from deviating from the x-ray structure). A total of six independent MD trajectories were produced, each exceeding 80 nanoseconds of simulated time. Analysis of the MD trajectories revealed that the imidazo-pyrazine distal substructure converged to a common position despite the different starting conformations (Figure 7), increasing our confidence in the simulations, because it is highly unlikely that independent trajectories find the same binding pose by chance⁴². While the distal substructure is more mobile than the rest of the molecule, its oscillations are not as large as the docked models initially suggested, and π−π stacking with Y49’ is the only salient feature. Structure-based optimization of the distal substructure may be possible by growing the molecule towards the vicinity of Y36’.

Figure 7.

Molecular dynamics (MD) simulations of JTO-FL•53. A) Line drawing of stabilizer 53. B-C) The three docked poses of 53 that were used as starting coordinates for the MD simulations are depicted in red, orange, and green. Representative structures of the consensus binding mode seen in all independent MD simulations are depicted in gray. The B) and C) panels show different views of the same data.

Clinically viable LC stabilizers must be able to selectively bind to and stabilize FL LCs in patient plasma, in the presence of thousands of other proteins. We measured the plasma protein binding of potent LC coumarin-based stabilizers (e.g., 53, (R)-83; Table S8). All small molecules tested had very high (> 99.5%) plasma protein binding, which was most likely derived from the highly lipophilic anchor substructure and coumarin core. In support of this hypothesis, coumarin-based stabilizers lacking a linker module and distal substructure (1, 2, and 57) also exhibited very high plasma protein binding. Stabilizers 53 and (R)-83 (10 μM) provided only marginal proteinase K protection for fluorescein-labeled WIL-FL T46L/F49Y (0.5 μM) in 93% healthy human plasma, even though they have EC₅₀ values of < 5 nM in buffer (Figure S7). Efforts to design potent stabilizers with improved FL LC binding selectivity over other plasma proteins are ongoing.

Discussion and Conclusions

Previous FL LC dimer• stabilizer crystal structures^37–38 of hits identified by our high-throughput screen suggested that we could design and synthesize FL LC stabilizers composed of four substructures: an “anchor substructure”, an “aromatic core”, a “linker module” and a “distal substructure” that bind to the “anchor cavity”, the “aromatic slit”, the “solvent-exposed linker binding site”, and the “distal aromatic pocket”, of FL LC dimers, respectively (Figure 2). The FL LC small molecule stabilizer SAR data presented herein validate our approach. By designing FL LC stabilizers comprising four components that occupy these binding sites, we have increased LC binding affinity by 3,000-fold compared to the screening hits, yielding kinetic stabilizer leads with single-digit nanomolar potency, which are the highest-potency LC stabilizers described to date. These stabilizers comprise a 1-phenylethoxy or 2-phenylpropoxy ether “anchor substructure”, a coumarin “aromatic core”, a “linker module” comprising a two-methylene spacer followed by spiro-urea-piperidine, and an imidazo[1,2-a]pyrazine “distal substructure” (83 – 84). We confirmed that conserved FL LC amino acid residues beyond the previously-identified binding site³⁷ can be engaged to bind highly potent FL LC stabilizers. Further improving the potency of LC stabilizers may not be necessary because LCs are typically found at micromolar concentrations in patients.²⁰

The approach we used is analogous to the “fragment-growing” strategies in hit-to-lead efforts⁴³. Our existing leads (e.g., 53, (R)-83)) were derived from identifying a suitable growth vector from screening hits 1 and 2, which have low molecular weight (< 300 Da), modest affinity for LCs, and low complexity. Our results demonstrate that structure-based design starting from fragment-like hits can be successful for identifying high-affinity lead molecules for proteins that are not evolved to bind small molecules, such as LCs.

The low binding selectivity of stabilizers reported herein greatly diminishes their efficacy in binding to and stabilizing FL LCs in human plasma due to competition with other plasma proteins. Thus, we will reduce lipophilicity and carbocyclic aromatic content to improve selective binding to FL LCs in blood^42–43, which will entail replacing the coumarin aromatic core. We will primarily optimize for improving proteinase K resistance of LCs in plasma, which is a more informative metric than plasma protein binding measurements alone. Future design, synthesis, and selection of more soluble stabilizers will also facilitate important additional in vitro characterization assays, such as measurement of dissociation constants and assessing direct inhibition of FL LC aggregation, which were not successful with the most potent compounds described here, due to the difficulty with solubility. In future studies, we will also test the hypothesis that optimization of the binding of stabilizers to highly conserved FL LC residues enables stabilization of several different AL patient-derived LC sequences in patient plasma.

In summary, we have discovered the first FL LC stabilizers with single-digit nanomolar potency through structure-based design, starting from high-throughput screening hits with micromolar potency. We have also further elaborated the conserved binding surface of FL LC dimers that can bind stabilizers. Our data establish the feasibility of pharmacologically targeting LCs, which were previously thought to be “undruggable”. This work is a critical milestone towards the next goal of identifying leads exhibiting selective FL LC binding in plasma, ultimately focused on identifying a clinical candidate for the treatment of AL.

Chemistry

Preparation of 7-dialkylamino coumarin analogs 9 to 13 was achieved by nucleophilic substitution with the appropriate alkyl halide, starting from commercially available 7-amino-4-methylcoumarin (7, Scheme 1). For 9 to 12, the mono-ethyl intermediate 8 was isolated after reaction with iodoethane and subjected to the same reaction conditions with a different alkyl halide. For 13, an excess of propargyl bromide was used to prepare the di-propargyl compound from 7.

Scheme 1.

Synthesis of 7-dialkyl-4-methylcoumarins 9 – 13^a.

^aReagents and conditions: a) alkyl bromide or iodide, DMF, 80 °C, 2–16 h.

Syntheses of carbamate analogs 18 to 27 each derived from the key intermediate 17 which was obtained in a multi-step route from commercially available 1 (Scheme 2). First, selective bromination at the 3-position with N-bromosuccinimide (NBS) afforded 14, which underwent a subsequent Suzuki reaction with potassium trifluoro(2-((tetrahydro-2H-pyran-2-yl)oxy)ethyl)borate to prepare the THP-protected intermediate 15. Deprotection under acidic conditions gave alcohol 16, followed by conversion into the activated carbonate 17 by reaction with p-nitrophenyl chloroformate. Displacement of p-nitrophenolate with a variety of primary or secondary amines resulted in carbamates 18 to 26. To synthesize the meta-phenyl-pyridone 27, intermediate 17 was first converted into boronic acid pinacol ester 17a by coupling with commercially available 3-aminomethylphenylboronic acid pinacol ester, followed by a Suzuki reaction with 3-bromopyridin-2(1H)-one.

Scheme 2.

Synthesis of carbamates 18 – 27^a.

^aReagents and conditions: a) N-bromosuccinimide, MeCN, rt, 3 h; b) potassium trifluoro(2-((tetrahydro-2H-pyran-2-yl)oxy)ethyl)borate, Pd(OAc)₂, cataCXium® A, Cs₂CO₃, 1,4-dioxane, H₂O, 100 °C, 12 h; c) HCl, 1,4-dioxane, 25 °C, 30 min; d) p-nitrophenyl chloroformate, pyridine, DCM, rt, 3 h; e) R-NH₂, DIPEA, DMF, 25 °C, 2 h; f) 3-bromopyridin-2(1H)-one, Pd(PPh₃)₄, K₂CO₃, 3:1 dioxane/H₂O, 90 °C, 8 h.

To synthesize spiro-hydantoin coumarins 34 to 46, the respective spiro-hydantoins (30 and 31) were first prepared using a Bucherer-Bergs reaction from commercially available Boc-pyrrolidone 28 or Boc-piperidone 29 (Scheme 3). A Mitsunobu reaction with alcohol 16, followed by Boc deprotection, resulted in intermediate amines 32 and 33. These intermediates were subjected to amide coupling conditions using HATU to prepare amides 34 – 37, reaction with sulfonyl chlorides to prepare sulfonamides 38 – 39, or nucleophilic aromatic substitution with electron-deficient aryl halides to prepare 40 – 46.

Scheme 3.

Synthesis of spiro-hydantoin analogs 34 – 46^a.

^aReagents and conditions: a) KCN, (NH₄)₂CO₃, EtOH, H₂O, 60 °C, 16 h; b) 16, PPh₃, diisopropyl azodicarboxylate, THF, 0 °C to 25 °C, 4 h; c) HCl, MeOH, H₂O, 25 °C, 16 h; d) R₁-CO₂H, HATU, DIPEA, DMF, 25 °C, 16 h; e) R₂-SO₂Cl, DIPEA, DMF, 0 °C, 3 h; f) R₃-Cl, DIPEA, DMF, 100 °C, 16 h.

Syntheses of spiro-ureas 51 to 53 and 83 to 85 started from the respective aldehydes (Schemes 4 and 5). The aldehydes were prepared using a variety of methods (Scheme 4). Oxidation of aminocoumarin alcohol 16 afforded aldehyde 47. The coumarin ether aldehydes 70 and 71 were prepared using a multistep route. Commercially available 7-methoxy-4-methylcoumarin 64 was first brominated using NBS to afford 65, which was subsequently demethylated using BBr₃ to afford 66. This intermediate was subjected to a Suzuki coupling to give vinyl ether 67, which underwent Mitsunobu reactions with optically pure (R) or (S)-1-phenylethanol to give ethers (R)-68 and (S)-69, or racemic 1-phenyl-1-propanol to give 70. Subsequent acidic cleavage resulted in aldehydes (R)-70, (R)-70, and 71. The truncated aldehyde 73 was prepared from racemic coumarin ether 61 (see Scheme 6 for synthesis) in two steps, starting with a Stille coupling to afford hydroxymethyl coumarin 72, followed by oxidation with MnO₂ to give aldehyde 73.

Scheme 4.

Synthesis of aldehyde intermediates 47, (R)-70, (S)-70, 71, and 73^a.

^aReagents and conditions: a) Dess-Martin Periodinane, DCM, 0 °C to 25 °C, 2 h; b) NBS, MeCN, 50 °C, 2 h; c) BBr₃, DCM, −78 to 30 °C, 6 h; d) 2-[(E)-2-ethoxyvinyl]-4,4,5,5-tetramethyl-1,3,2-dioxaborolane, Pd(dppf)Cl₂, K₂CO₃, dioxane, H₂O, 100 °C, 2 h; e) R₁-OH, DIAD, PPh₃, THF, 30 °C, 1 h; f) 3M HCl in acetone, 30 °C, 30 min; g) (tributylstannyl)methanol, Pd(PPh₃)₄, dioxane, 100 °C, 12 h; h) MnO₂, DCM, 30 °C, 24 h.

Scheme 5.

Synthesis of spiro-ureas 51 – 53 and 83 – 85 (separate enantiomers). (R)-83 and (S)-83 were prepared from enantiomerically pure alcohols, while 84 and 85 were prepared from racemic 1-phenyl-1-propanol and then enantiomerically resolved using supercritical fluid chromatography^a.

^aReagents and conditions: a) tert-butyl 4-amino-4-(aminomethyl)piperidine-1-carboxylate, NaBH(OAc)₃ or NaBH₃CN, AcOH, 1,2-DCE or MeOH, 25 °C, 16 h; b) carbonyldiimidazole, DCM, 25 °C, 24 h; c) HCl, MeOH, H₂O, 25 °C, 16 h; d) R-Cl, DIPEA, DMF, 100 °C, 16 h.

Scheme 6.

Synthesis of coumarin ethers 55 – 59 and morpholine analogs 62 – 63 for crystallography^a.

^aReagents and conditions: a) R-Br or R-OTs, DMF, K₂CO₃, 80 °C, 3 to 16 h; b) N-bromosuccinimide, MeCN, 25 °C, 16 h; c) potassium (morpholin-4-yl)methyltrifluoroborate, Pd(dba)₂, RuPhos, Cs₂CO₃, toluene, H₂O, 90 °C, 16 hr.

The aldehydes 47, (R)-70, (S)-70, 71, and 72 were subjected to reductive amination with commercially available tert-butyl 4-amino-4-(aminomethyl)piperidine-1-carboxylate, followed by intramolecular urea formation with carbonyldiimidazole and subsequent Boc deprotection, to give amines 50, (R)-80, (S)-80, 81, and 82, respectively (Scheme 5). These advanced intermediates were subjected to nucleophilic aromatic substitution conditions, analogous to Scheme 3, to prepare 51 – 53 and 83 – 85. Specifically, diethylamino coumarin 50 was reacted with 2-chloropyrimidine, 4-chloro-1H-pyrazolo[3,4-d]pyrimidine, or 8-chloroimidazo[1,2-a]pyrazine to prepare 51 – 53, respectively. Coumarin ethers (R)-80, (S)-80, 81, and 82 were reacted with 8-chloroimidazo[1,2-a]pyrazine to give (R)-83, (S)-83, 84, and 85, respectively. Compounds 84 and 85 were enantiomerically resolved to 84a, 84b, 85a, and 85b using chiral supercritical fluid chromatography (SFC). The absolute configuration of each isolated enantiomer of 84 and 85 was not determined.

Coumarin ethers 55 – 59 were readily accessible in a one-step Williamson ether synthesis, starting with commercially available 7-hydroxyl-4-methylcoumarin 54 and various alkyl bromides or tosylates (Scheme 6). To prepare morpholine analogs 62 and 63 for crystallography, coumarin ethers 2 and 57 were first brominated at the 3-position with NBS. Suzuki coupling with potassium (morpholin-4-yl)methyltrifluoroborate resulted in 62 and 63.

Experimental Section

Protein expression, purification, and labeling with fluorescein-maleimide

Full-length LCs were expressed and purified as previously described¹⁴. The WIL-FL T46L/F49Y and T46L/F49Y/K79C variants were constructed using the Q5® site-directed mutagenesis kit (NEB) according to the manufacturer’s instructions. Labeling of WIL-FL T46L/F49Y/K79C with fluorescein-maleimide was performed as previously described for fluorescein-labeled native WIL-FL at residue 79³⁷.

Crystallography

Generation of unliganded JTO-FL crystals and subsequent soaking of ligands, along with data collection, processing, and refinement (Table 3), were performed as previously described^37–38.

Protease sensitivity assay with unlabeled LCs

WIL-FL (5 μM) in phosphate-buffered saline (PBS, pH 7.4) was incubated with small molecule stabilizer candidates (10 μM in 1% DMSO) and proteinase K (50 nM) for 2 h at 37 °C. Reactions were quenched with phenylmethylsulfonyl fluoride (PMSF, 2 mM) and analyzed using size-exclusion chromatography as previously described³⁸. Proteinase K sensitivity assays for WIL T46L/F49Y/C214S used similar conditions, but the DMSO concentration was 0.5% and the incubation time was 1 h at 37 °C. Fold protection was calculated using the following:

$$\mathit{\text{Fold protection}}= \frac{A_c- A_v}{A_{-PK}- A_v}$$

“A” represents the peak integration of the respective treatments: c, compound; v, DMSO vehicle; -PK, no proteinase K.

Proteolysis-coupled fluorescence polarization (PCFP) assay

Determination of stabilizer EC₅₀ values using the PCFP assay on fluorescein-labeled WIL-FL (native and T46L/F49Y) was performed similarly as previously described³⁷. Serial dilutions of stabilizers (10 mM to 500 nM) in DMSO were dry-spotted (0.2 μL) onto black polystyrene 384-well microplates (Greiner catalog #788076) using a Labcyte Echo 550. To the treated wells were added 10 μL of 5 nM fluorescein-labeled dimeric WIL-FL in PBS with 0.04% pluronic F-127, followed by 10 μL of 400 nM proteinase K (Thermo) in PBS, achieving final concentrations of 2.5 nM LC dimer, 0.02% pluronic F-127, and 200 nM proteinase K. The plates were spun at 400 × g for 10 seconds, sealed in foil, and incubated at 22 °C for 24 h. Fluorescence polarization was measured using a PHERAstar plate reader (FP 485 520 520 optic module, focal height = 10.0 mm, channel A gain = 400, channel B gain = 360, settling time = 0.3 s, number of flashes per well = 300, target mP = 35). Note that the initial dose-response evaluation used a final concentration of 10 nM LC dimer. The concentration was lowered to 2.5 nM dimer to re-test stabilizers with EC₅₀ values of 10 nM or lower (e.g., 51 – 53, 83, 84). The EC₅₀ values of less potent compounds (e.g., 1) is unaffected regardless of the LC dimer concentration used.

Molecular docking

Docking was performed using AutoDock Vina⁴⁴ (hereafter referred to as Vina) v1.1.2 as well as an unpublished version of Vina that uses the AutoDock 4 scoring function⁴⁵. Three-dimensional coordinates of the small molecules were generated using OpenBabel⁴⁶ v2.4.1. Atom typing, assignment of rotatable bonds, merging of non-polar hydrogens, and calculation of Gasteiger partial charges⁴⁷ was also performed with OpenBabel v2.4.1.

Molecular dynamics of JTO-FL•53

Three docked poses of JTO-FL•53 were manually selected to be the starting points of independent molecular dynamics (MD) trajectories. Each of the docked poses was simulated both with and without restraints on the protein α-carbons (0.1 kcal/mol/Å) leading to a total of six independent MD trajectories. The simulated dimer was a truncated version of JTO-FL: only residues F1 through L109 from each chain were included in the simulation. The starting coordinates of protein atoms corresponded to the crystallographic structure of JTO-FL•1. The Amber ff14SB forcefield⁴⁸ was used for protein atoms, GAFF2 for compound 53 with AM1-BCC partial charges⁴⁹, and TIP3P water⁵⁰. Periodic boundary conditions were used employing a rectangular box while leaving at least 12 Å between protein atoms and any face of the box. The net charge was neutralized with 5 Na⁺ ions.

Prior to the production MD simulation, all atoms were minimized for 1000 steps with 25 kcal/mol/Å restraints on protein atoms, followed by another 1000 steps with softer restraints of 10 kcal/mol/Å. Then, the temperature was increased to 300 K during 140 picoseconds with 10 kcal/mol/Å restraints on α-carbons. Two 20 picosecond-long constant-volume (NVT) simulations were performed, first with 5 kcal/mol/Å restraints and then with 2 kcal/mol/Å restraints on α-carbons. A 40 picosecond constant pressure (NPT) simulation with 2 kcal/mol/Å restraints on α-carbons brought the pressure to 1 bar using a 2 picosecond pressure relaxation time. The production constant-pressure MD simulations were at least 80 nanoseconds long and were continued up to 200 ns depending on the availability of computational resources. Soft restraints of 0.1 kcal/mol/Å were used on α-carbons. Coordinates were recorded every 10 picoseconds. Simulations were performed using the PMEMD CUDA binary of Amber 16.

Structural analysis of MD trajectories

MD trajectories were aligned with cpptraj⁵⁰ using protein residues at the surface of the binding pocket: Y36, Q38, P44, Y87, V96, V97, and F98 from the A chain, and Y36’, P44’, T46’, Y49’, Q89’, and Y91’ from the B chain. For each trajectory, the RMSD between all pairs of recorded frames were plotted in a bidimensional matrix, where the RMSD value for each pair of frames is color coded (Figure S6). These RMSD values were calculated for the heavy atoms of molecule 53 excluding the carbons in the diethylamino anchor substructure. Using this graphical representation of pairwise RMSD values, segments of each MD trajectory for which 53 remained in a well-defined binding mode for an extended period were visually identified. A few short segments in which 53 adopted a clearly distinct conformation were also selected. The number of visually identified segments varied between one and three per MD trajectory, for a total of fourteen segments. The representative frame for each of the segments identified was the frame with the lowest median RMSD from all the other frames in the same segment. The percentage of frames within a 1.0 Å RMSD (using heavy atoms of 53 but excluding diethylamino carbons) of each of the fourteen representative frames was calculated (Table S7).

Measurement of plasma protein binding (PPB)

Small molecule LC stabilizer binding to human or mouse plasma (BioIVT) was measured using a 96-well HT equilibrium dialysis device from Dialysis LLC (Gales Ferry, CT), which was assembled according to the manufacturer’s instructions. Small molecules (2 μM) in plasma were dialyzed against 100 mM sodium phosphate, 150 mM NaCl, pH 7.4 buffer for 4 h at 37 °C. Aliquots from both the buffer and plasma components were analyzed using UPLC-MS/MS. All measurements were done in triplicate.

Protease sensitivity assay with fluorescein-labeled LCs in plasma

Healthy human plasma (Scripps Normal Blood Donor Service) was freshly thawed at ambient temperature. Proteolysis reactions were prepared as follows: fluorescein-labeled WIL-FL T46L/F49Y (500 nM), stabilizer (10 μM), and proteinase K (2 μM) were sequentially added to plasma for a final plasma concentration of approximately 93%. Reactions were incubated at 37 °C for 2 h, quenched with 1 mM PMSF, and diluted tenfold with PBS (pH 7.4) to a final plasma concentration of less than 10%. Samples were analyzed by SDS-PAGE and imaged using the fluorescein signal. Note that a higher proteinase K concentration (2 μM) compared to proteolysis assays in buffer (50 to 200 nM) is necessary because the presence of plasma slows the proteolysis of LCs.

General Synthetic Procedures

All reagents and anhydrous solvents were obtained from commercial suppliers and used without further purification. ¹H and ¹³C NMR spectra were recorded on a Varian Mercury-400, Bruker DPX-400, or DRX-500. Normal-phase chromatography was performed on a Teledyne Isco CombiFlash NextGen 300+ using Luknova SuperSep SiO₂ columns. Preparative-scale reverse phase high performance liquid chromatography (RP-HPLC) was performed on an Agilent 1260 Infinity LC system using a Gemini NX-C18 column (110 Å pore size, 5 μm particle size, 150 × 21.2 mm dimensions, mobile phase A = 0.1% TFA in H₂O, mobile phase B = 0.1% TFA in MeCN). Final compound purities were determined by analytical RP-HPLC (254 nm) and were > 95% in purity. High-resolution mass spectrometry data were collected at The Scripps Research Institute Center for Mass Spectrometry (ESI-HRMS; Agilent Technologies, LC/MSD TOF 6230).

7-(ethylamino)-4-methyl-2H-chromen-2-one (8).

The title compound was prepared from 7 and iodoethane according to a literature procedure⁵¹. Spectral data matched that reported in the literature.

General Method A: aminocoumarin alkylation.

A suspension of 8, alkyl bromide or iodide (5 to 10 eq), and K₂CO₃ (3 eq) in DMF (1 mL) was heated at 80 °C for at least 16 h. The mixture was diluted with water (20 mL) and extracted with EtOAc (3 × 15 mL). The combined organics were washed once with brine, dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography (SiO₂, 0 to 25% EtOAc in hexanes).

7-(ethyl(methyl)amino)-4-methyl-2H-chromen-2-one (9).

The title compound was prepared using General Method A using methyl iodide to afford 9 as a viscous, dark yellow oil.

¹H NMR (500 MHz, Chloroform-d) δ 7.41 (d, J = 8.9 Hz, 1H), 6.63 (dd, J = 8.9, 2.5 Hz, 1H), 6.52 (d, J = 2.5 Hz, 1H), 5.97 (q, J = 1.2 Hz, 1H), 3.48 (q, J = 7.1 Hz, 2H), 3.02 (s, 3H), 2.36 (d, J = 1.3 Hz, 3H), 1.19 (t, J = 7.1 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 162.20, 155.88, 152.92, 151.68, 125.40, 109.41, 109.05, 108.68, 98.09, 46.77, 37.59, 18.46, 11.41. LC-MS [ESI, M+1]: 218.1

7-(ethyl(propyl)amino)-4-methyl-2H-chromen-2-one (10).

The title compound was prepared using General Method A using n-propyl iodide to afford 10 as a viscous yellow oil.

¹H NMR (600 MHz, Acetone-d₆) δ 7.50 (d, J = 9.0 Hz, 1H), 6.72 (dd, J = 9.0, 2.6 Hz, 1H), 6.49 (d, J = 2.6 Hz, 1H), 5.89 (q, J = 1.2 Hz, 1H), 3.53 (q, J = 7.1 Hz, 2H), 3.43 – 3.37 (m, 2H), 2.37 (d, J = 1.2 Hz, 3H), 1.74 – 1.64 (m, 2H), 1.21 (t, J = 7.1 Hz, 3H), 0.99 (t, J = 7.4 Hz, 3H). ¹³C NMR (151 MHz, Acetone-d₆) δ 204.77, 160.21, 155.71, 152.42, 150.54, 125.38, 108.01, 107.74, 96.65, 51.25, 44.34, 19.92, 16.94, 11.13, 10.09. LC-MS [ESI, M+1]: 246.1

7-(ethyl(isopropyl)amino)-4-methyl-2H-chromen-2-one (11).

The title compound was prepared using General Method A using isopropyl iodide to afford 11 as a viscous yellow oil.

¹H NMR (600 MHz, Acetone-d₆) δ 7.52 (d, J = 9.0 Hz, 1H), 6.82 (dd, J = 9.0, 2.6 Hz, 1H), 6.58 (d, J = 2.7 Hz, 1H), 5.90 (q, J = 1.2 Hz, 1H), 4.27 (hept, J = 6.6 Hz, 1H), 3.44 (q, J = 7.1 Hz, 2H), 2.38 (d, J = 1.2 Hz, 3H), 1.28 (d, J = 6.6 Hz, 6H), 1.23 (t, J = 7.1 Hz, 3H). ¹³C NMR (151 MHz, Acetone) δ 204.79, 160.22, 155.64, 152.37, 150.87, 125.32, 108.66, 107.97, 97.45, 47.83, 37.28, 18.91, 16.92, 13.58. LC-MS [ESI, M+1]: 246.1

7-(ethyl(prop-2-yn-1-yl)amino)-4-methyl-2H-chromen-2-one (12).

The title compound was prepared using General Method A using propargyl bromide to afford 12 as a light yellow solid.

¹H NMR (500 MHz, Chloroform-d) δ 7.45 (d, J = 8.8 Hz, 1H), 6.74 (dd, J = 8.9, 2.6 Hz, 1H), 6.68 (d, J = 2.5 Hz, 1H), 6.02 (q, J = 1.2 Hz, 1H), 4.10 (d, J = 2.4 Hz, 2H), 3.54 (q, J = 7.2 Hz, 2H), 2.37 (d, J = 1.2 Hz, 3H), 2.25 (t, J = 2.4 Hz, 1H), 1.28 (t, J = 7.1 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 161.94, 155.66, 152.77, 150.36, 125.48, 110.50, 109.92, 109.54, 99.46, 79.21, 72.26, 46.00, 39.73, 18.47, 12.27. LC-MS [ESI, M+1]: 242.1.

7-(di(prop-2-yn-1-yl)amino)-4-methyl-2H-chromen-2-one (13).

The title compound was prepared analogously to 8, but using an excess of propargyl bromide (10 eq) to afford 13 as a dark brown oil.

¹H NMR (600 MHz, Chloroform-d) δ 7.51 (d, J = 8.8 Hz, 1H), 6.87 (dd, J = 8.8, 2.6 Hz, 1H), 6.84 (d, J = 2.5 Hz, 1H), 6.10 (q, J = 1.2 Hz, 1H), 4.21 (d, J = 2.4 Hz, 4H), 2.40 (d, J = 1.2 Hz, 3H), 2.32 (t, J = 2.4 Hz, 2H). ¹³C NMR (151 MHz, Chloroform-d) δ 161.22, 154.82, 152.14, 149.75, 125.01, 111.48, 110.60, 110.37, 100.86, 77.70, 72.74, 39.90, 18.05. LC-MS [ESI, M+1]: 252.1

3-bromo-7-(diethylamino)-4-methyl-2H-chromen-2-one (14).

To a solution of 1 (24 g, 104 mmol, 1 eq) in MeCN (300 mL) was added N-bromosuccinimide (20.3 g, 114 mmol, 1.1 eq). The mixture was stirred at 20 °C for 1 h. The mixture was diluted with ethyl acetate (600 mL), and then washed with water (1 × 300 mL) and brine (1 × 200 mL). The separated organic layer was dried over sodium sulfate, filtered and concentrated under vacuum. The mixture was purified by column chromatography over silica gel (petroleum ether/ethyl acetate 10/1 to 4/1). The desired fractions were collected and concentrated under vacuum to give 14 (31 g, 96%) as a yellow solid. Spectral data matched that reported in the literature⁵².

7-(diethylamino)-4-methyl-3-(2-((tetrahydro-2H-pyran-2-yl)oxy)ethyl)-2H-chromen-2-one (15).

A mixture of 14 (3.91 g, 12.6 mmol, 1 eq), potassium trifluoro(2-((tetrahydro-2H-pyran- 2-yl)oxy)ethyl)borate (4.46 g, 18.9 mmol, 1.5 eq), Pd(OAc)₂ (848 mg, 3.78 mmol, 0.3 eq), CATACXIUM(R)A Pd G3 (1.83 g, 2.52 mmol, 0.2 eq) and Cs₂CO₃ (12.3 g, 37.8 mmol, 3 eq) in dioxane (120 mL) and H₂O (40 mL) was degassed under vacuum and purged with N₂ for 3 times, and then the mixture was stirred at 100 °C for 12 h under N₂ atmosphere to give a black suspension. The mixture was filtered and the filtrate was concentrated in vacuo. The residue was purified by reversed-phase flash (0.1% FA condition). The desired fractions were collected and lyophilized to give 7-(diethylamino)-4-methyl-3-(2-tetrahydropyran-2-yloxyethyl) chromen-2-one (2 g, 44%) as a yellow solid.

¹H NMR (400 MHz, Chloroform-d) δ 7.44 – 7.39 (m, 1H), 6.64 – 6.58 (m, 1H), 6.51 (d, J=2.8 Hz, 1H), 4.59 (t, J=4.0 Hz, 1H), 3.94 – 3.77 (m, 2H), 3.59 (td, J=6.9, 9.6 Hz, 1H), 3.53 – 3.45 (m, 1H), 3.41 (q, J=6.8 Hz, 4H), 2.98 – 2.89 (m, 2H), 2.42 – 2.36 (m, 3H), 1.81 – 1.72 (m, 1H), 1.72 – 1.62 (m, 1H), 1.59 – 1.43 (m, 4H), 1.20 (t, J=6.8 Hz, 6H). LC-MS [ESI, M+1]: 360.2

7-(diethylamino)-3-(2-hydroxyethyl)-4-methyl-2H-chromen-2-one (16).

A solution of 15 (1.9 g, 5.29 mmol, 1 eq) in HCl/dioxane (4 M, 50 mL) was stirred at 20 °C for 30 min. The mixture was concentrated under reduced pressure. The residue was purified by reverse-phase preparative HPLC. The desired fractions were collected and lyophilized to give 16 (734 mg, 45%) as a brown gum.

¹H NMR (400 MHz, DMSO-d₆) δ 7.58 – 7.48 (m, 1H), 6.75 – 6.63 (m, 1H), 6.52 – 6.42 (m, 1H), 4.61 – 4.53 (m, 1H), 3.75 – 3.59 (m, 2H), 3.46 – 3.37 (m, 6H), 2.84 – 2.73 (m, 2H), 2.37 – 2.27 (m, 3H), 1.73 – 1.52 (m, 2H), 1.49 – 1.34 (m, 4H), 1.16 – 1.07 (m, 6H). LC-MS [ESI, M+1]: 276.1

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (4-nitrophenyl) carbonate (17).

To a stirred solution of 16 (500 mg, 1.82 mmol, 1 eq) and pyridine (440 μL, 5.46 mmol, 3 eq) in DCM (2 mL) under argon was added a solution of 4-nitrophenyl chloroformate (440 mg, 2.18 mmol, 1.2 eq) in DCM (2 mL), dropwise, at 25 °C. The reaction was stirred for 3 h at 25 °C, during which the solution turned orange and a crystalline precipitate appeared. The mixture was partitioned between DCM (20 mL) and 0.1 M aqueous HCl (20 mL), then extracted with DCM (3 × 20 mL). The combined organics were washed with brine, dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography (SiO₂, 0 to 25% EtOAc in hexanes) to afford 17 as a yellow-orange solid (621.1 mg, 78%), which was stored at −20 °C.

¹H NMR (400 MHz, Chloroform-d) δ 8.27–8.23 (m, 2H), 7.43 (d, J = 8.8 Hz, 1H), 7.36 – 7.31 (m, 2H), 6.61 – 6.56 (m, 1H), 6.51 (d, J = 2.8 Hz, 1H), 4.48 (t, J = 6.8 Hz, 2H), 3.42 (q, J = 7.2 Hz, 4H), 3.11 (t, J = 6.8 Hz, 2H), 2.41 (s, 3H), 1.22 (t, J = 7.2 Hz, 6H). LC-MS [ESI, M+1]: 441.4

General Method B: Carbamate synthesis.

To a stirred solution of 17 (10 to 200 mg, 1 eq) and DIPEA (3 eq) in DMF (1 mL) was added primary or secondary amine (1.5 eq). Typically, the solution immediately turned a bright yellow upon addition of amine. The reaction was stirred at 25 °C for 2 h, then diluted with 20 mL H₂O and basified with addition of solid Na₂CO₃. The suspension was extracted with EtOAc (3 × 10 mL). The combined organics were washed once with brine, dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by normal-phase flash column chromatography or with reverse-phase preparative HPLC to afford the carbamate products.

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (3-(4,4,5-trimethyl-1,3,2-dioxaborolan-2-yl)benzyl)carbamate (17a).

The title compound was prepared according to General Method B, using (3-(4,4,5-trimethyl-1,3,2-dioxaborolan-2-yl)phenyl)methanamine, to afford 17a as a yellow solid (204.4 mg, 86%).

LC-MS [ESI, M+1]: 521.4

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (piperidin-3-ylmethyl)carbamate (18).

The Boc-protected amine was prepared according to General Method B, using tert-butyl 3-(aminomethyl)piperidine-1-carboxylate, and purified by flash column chromatography (SiO₂, EtOAc/hexanes). To a stirred solution of the Boc-protected amine in DCM (2 mL) was added neat TFA (1 mL) and the mixture stirred at 25 °C for 16 h. The mixture was concentrated under reduced pressure and purified by reverse-phase preparative HPLC to afford 18 as a yellow gum (10.2 mg, 71% over 2 steps).

¹H NMR (400 MHz, Chloroform-d) δ 7.42 (d, J = 9.0 Hz, 1H), 6.60 (dd, J = 9.1, 2.6 Hz, 1H), 6.50 (d, J = 2.6 Hz, 1H), 5.13 (dd, J = 7.6, 4.9 Hz, 1H), 4.21 (t, J = 7.0 Hz, 2H), 3.42 (q, J = 7.1 Hz, 4H), 3.30 – 3.03 (m, 7H), 2.96 (t, J = 7.0 Hz, 2H), 2.65 (t, J = 11.5 Hz, 1H), 2.43 (s, 1H), 2.39 (s, 3H), 1.87 – 1.67 (m, 3H), 1.67 – 1.51 (m, 1H), 1.21 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 416.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl benzylcarbamate (19).

The title compound was prepared according to General Method B, using benzylamine, to afford 19 as a yellow solid (26.1 mg, 80%).

¹H NMR (400 MHz, Chloroform-d) δ 7.41 (d, J = 9.0 Hz, 1H), 7.37 – 7.27 (m, 5H), 6.61 (dd, J = 9.0, 2.6 Hz, 1H), 6.51 (d, J = 2.6 Hz, 1H), 4.97 (br t, 1H), 4.39 (d, J = 6.0 Hz, 2H), 4.28 (t, J = 7.0 Hz, 2H), 3.43 (q, J = 7.1 Hz, 4H), 3.01 (t, J = 7.0 Hz, 2H), 2.39 (s, 3H), 1.23 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 409.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (pyridin-3-ylmethyl)carbamate (20).

The title compound was prepared according to General Method B, using pyridin-3-ylmethanamine, to afford 20 as a tan solid (18.7 mg, 71%).

¹H NMR (400 MHz, Chloroform-d) δ 8.53 (s, 1H), 8.52 (d, J = 4.4 Hz, 1H), 7.65 (d, J = 7.9 Hz, 1H), 7.40 (d, J = 9.0 Hz, 1H), 7.25 (dd, J = 7.8, 5.0 Hz, 1H), 6.60 (dd, J = 9.0, 2.6 Hz, 1H), 6.49 (d, J = 2.6 Hz, 1H), 5.25 (br t, 1H), 4.38 (d, J = 6.2 Hz, 2H), 4.27 (t, J = 7 Hz, 2H), 3.42 (q, J = 7.1 Hz, 4H), 2.98 (t, J = 7.0 Hz, 2H), 2.36 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 410.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (2-(pyridin-3-yl)ethyl)carbamate (21).

The title compound was prepared according to General Method B, using 2-(pyridin-3-yl)ethan-1-amine, to afford 21 as a tan solid (20.7 mg, 78%).

¹H NMR (400 MHz, Chloroform-d) δ 8.49 (dd, J = 4.7, 1.4 Hz, 1H), 8.47 (br s, 1H), 7.53 (d, J = 7.9 Hz, 1H), 7.41 (d, J = 9 Hz, 1H), 7.24 (dd, J = 7.8, 4.8 Hz, 1H), 6.60 (dd, J = 9.0, 2.6 Hz, 1H), 6.50 (d, J = 2.6 Hz, 1H), 4.84 (br t, 1H), 4.23 (t, J = 7.0 Hz, 2H), 3.48 – 3.36 (m, 6H, overlapping peaks), 2.97 (dd, J = 8.3, 5.7 Hz, 2H), 2.83 (t, J = 7.2 Hz, 2H), 2.37 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 424.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl methyl(2-(pyridin-3-yl)ethyl)carbamate (22).

The title compound was prepared according to General Method B, using N-methyl-2-(pyridin-3-yl)ethan-1-amine, to afford 22 as a viscous yellow oil (15.1 mg, 84%).

¹H NMR (400 MHz, Chloroform-d) δ 8.53 (br d, J = 3.5 Hz, 1H), 7.65 – 7.54 (m, 1H), 7.46 – 7.35 (m, 1H), 7.24 – 7.05 (m, 2H), 6.60 (dd, J = 9.0, 2.6 Hz, 1H), 6.50 (d, J = 2.6 Hz, 1H), 4.29 – 4.13 (m, 2H), 3.69 – 3.57 (m, 2H), 3.42 (q, J = 7.1 Hz, 4H), 3.07 – 2.92 (m, J = 4H), 2.87 (s, 1.5H, amide conformer methyl), 2.80 (s, 1.5H, amide conformer methyl), 2.39 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 438.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (3-acetamidobenzyl)carbamate (23).

The title compound was prepared according to General Method B, using N-(3-(aminomethyl)phenyl)acetamide, to afford 23 as a white solid (11.0 mg, 59%).

¹H NMR (500 MHz, Chloroform-d) δ 7.91 (s, 1H), 7.66 (d, J = 7.6 Hz, 1H), 7.41 (d, J = 9.0 Hz, 1H), 7.30 – 7.22 (m, 2H, overlapping peaks), 6.97 (d, J = 7.6 Hz, 1H), 6.60 (dd, J = 9.0, 2.6 Hz, 1H), 6.47 (d, J = 2.6 Hz, 1H), 5.31 (br t, 1H) 4.35 (d, J = 6.2 Hz, 2H), 4.27 (t, J = 5 Hz, 2H), 3.42 (q, J = 7.1 Hz, 4H), 3.01 (t, J = 6.9 Hz, 2H), 2.38 (s, 3H), 2.18 (s, 3H), 1.22 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 466.3

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl ((5-acetamidopyridin-3-yl)methyl)carbamate (24).

The title compound was prepared according to General Method B, using N-(5-(aminomethyl)pyridin-3-yl)acetamide, to afford 24 as a light yellow solid (10.3 mg, 65%).

¹H NMR (500 MHz, Chloroform-d) δ 8.80 (s, 1H), 8.44 (s, 1H), 8.24 (s, 1H), 7.85 (s, 1H), 7.40 (d, J = 9.0 Hz, 1H), 6.60 (dd, J = 9.1, 2.6 Hz, 1H), 6.46 (d, J = 2.6 Hz, 1H), 5.66 (br t, 1H), 4.36 (d, J = 6.1 Hz, 2H), 4.25 (t, J = 6.9 Hz, 2H), 3.42 (q, J = 7.1 Hz, 4H), 2.99 (t, J = 6.9 Hz, 2H), 2.37 (s, 3H), 2.22 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 467.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (3-carbamoylbenzyl)carbamate (25).

The title compound was prepared according to General Method B, using 3-(aminomethyl)benzamide, to afford 25 as a light yellow solid (13.1 mg, 66%).

¹H NMR (500 MHz, Chloroform-d) δ 7.89 (s, 1H), 7.82 (d, J = 4.8 Hz, 1H), 7.42 (m, 4H), 6.61 (dd, J = 9.0, 2.6 Hz, 1H), 6.47 (d, J = 2.6 Hz, 1H), 5.71 (br s, 1H), 5.15 (br t, 1H), 4.45 (d, J = 6.2 Hz, 2H), 4.30 (t, J = 6.5 Hz, 2H), 3.43 (q, J = 7.1 Hz, 4H), 3.01 (t, J = 6.5 Hz, 2H), 2.38 (s, 3H), 1.22 (t, J = 7.1 Hz, 6H). LC-MS [ESI, M+1]: 452.2

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (3-(1H-imidazol-4-yl)benzyl)carbamate (26).

The title compound was prepared according to General Method B, using (3-(1H-imidazol-4-yl)phenyl)methanamine, to afford 26 as a yellow solid (13.1 mg, 61%).

¹H NMR (500 MHz, Chloroform-d) δ 7.76 (d, J = 5.6 Hz, 2H), 7.61 (d, J = 7.8 Hz, 1H), 7.42 (d, J = 9.0 Hz, 1H), 7.39 (s, 1H), 7.32 (t, J = 7.7 Hz, 1H), 7.12 (d, J = 7.6 Hz, 1H), 6.61 (dd, J = 9.0, 2.6 Hz, 1H), 6.44 (d, J = 2.6 Hz, 1H), 5.16 (t, J = 6.1 Hz, 1H), 4.43 (d, J = 6.1 Hz, 2H), 4.30 (t, J = 6.4 Hz, 2H), 3.41 (q, J = 7.1 Hz, 4H), 3.03 (t, J = 6.4 Hz, 2H), 2.38 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). ESI-HRMS: calc’d for C₂₇H₃₀N₄O₄ [M+H]⁺ 475.2340, found 475.2341

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl (3-(2-oxo-1,2-dihydropyridin-3-yl)benzyl)carbamate (27).

A solution of 17a (12 mg, 0.022 mmol, 1 eq), 3-bromopyridin-2(1H)-one (5.7 mg, 0.033 mmol, 1.5 eq), and K₂CO₃ (12.5 mg, 0.09 mmol, 4 eq) in 1,4-dioxane (0.9 mL) and H₂O (0.3 mL) in a microwave vial was sparged with argon for 2 min. while stirring. To the mixture was added Pd(PPh₃)₄ (5.1 mg, 0.0044 mmol, 0.2 eq) and further sparged for an additional minute. The vial was sealed and heated at 90 °C for 8 h. The reaction mixture was then diluted with water (20 mL) and extracted with EtOAc (3 × 20 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography (SiO₂, 10 to 100% EtOAc/hexanes) to afford 27 as a white solid (4.4 mg, 40%).

¹H NMR (500 MHz, Chloroform-d) δ 12.15 (s, 1H), 7.64 (s, 1H), 7.59 (m, 2H), 7.44 – 7.37 (m, 2H), 7.35 (m, J = 5.6 Hz, 1H), 7.29 (s, 1H), 6.59 (dd, J = 9.0, 2.6 Hz, 1H), 6.49 (d, J = 2.6 Hz, 1H), 6.38 (t, J = 6.7 Hz, 1H), 5.13 (t, J = 5.6 Hz, 1H), 4.43 (d, J = 5.8 Hz, 2H), 4.27 (t, J = 7.0 Hz, 2H), 3.42 (q, J = 7.1 Hz, 4H), 3.00 (t, J = 7.0 Hz, 2H), 2.38 (s, 3H), 1.21 (t, J = 7.1 Hz, 6H). ¹³C NMR (126 MHz, CDCl₃) δ 163.39, 162.68, 156.65, 154.69, 149.91, 149.15, 139.78, 138.48, 136.83, 133.69, 131.53, 128.69, 127.77, 127.69, 127.17, 125.68, 115.39, 109.54, 108.49, 107.06, 97.47, 63.39, 45.12, 44.70, 29.71, 27.32, 14.81, 12.47. ESI-HRMS: calc’d for C₂₉H₃₁N₃O₅ [M+H]⁺ 502.2337, found 502.2338

Tert-butyl 2,4-dioxo-1,3,7-triazaspiro[4.4]nonane-7-carboxylate (30).

To a stirred suspension of KCN (293 mg, 4.5 mmol, 1.5 eq) and (NH₄)₂CO₃ (2.8 g, 30 mmol, 10 eq) in H₂O (6 mL) was added a solution of N-Boc-3-pyrrolidinone (556 mg, 3 mmol, 1 eq) in EtOH (4 mL). The white suspension was heated at 60 °C for 16 h (warning: significant gas evolution). The mixture was diluted with H₂O (40 mL) and extracted with EtOAc (3 × 30 mL). The combined organics were washed once with brine, dried over MgSO₄, filtered, and concentrated under reduced pressure to afford 30 as a white solid (650.9 mg, 85%), which was used without further purification. Spectral data matched that in the literature⁵³.

Tert-butyl 2,4-dioxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate (31).

The title compound was prepared analogously as for 30, using 1-Boc-4-piperidone as the starting material instead, to afford 31 as a light yellow solid (690.5 mg, 85%). Spectral data matched that in the literature⁵⁴.

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (32).

A suspension of 16 (200 mg, 0.73 mmol, 1 eq), 30 (206 mg, 0.80 mmol, 1.1 eq), and PPh₃ (382 mg, 1.46 mmol, 2 eq) in anhydrous THF (5 mL) was placed under argon and cooled on an ice bath. To the stirred suspension was added diisopropyl azodicarboxylate (288 μL, 1.46 mmol, 2 eq) dropwise over 2 minutes. The orange suspension was stirred on ice for 10 minutes, followed by at 25 °C for an additional 4 h, during which the hydantoin starting material gradually dissolved. The mixture was concentrated under reduced pressure and purified by flash column chromatography (SiO₂, 0 to 50% EtOAc/hexanes) to afford a mixture of the Boc-protected spiro-hydantoin coumarin and Ph₃PO as a viscous yellow oil. To a stirred solution of this residue in MeOH (3 mL) was then added concentrated HCl (1 mL) dropwise and stirred overnight at 25 °C. The reaction was diluted with water (30 mL) and washed once with EtOAc (15 mL) to remove Ph₃PO. Afterwards, the aqueous layer was basified with addition of solid Na₂CO₃ to give a bright yellow solution, which was extracted with DCM (6 × 20 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated to afford 32 as a yellow solid (274.3 mg, 91% over 2 steps) which was used without further purification.

LC-MS [ESI, M+1]: 413.4

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (33).

The title compound was prepared analogously as for 32, using 31 as the starting material, to afford 33 as a white solid (257.5 mg, 71% over 2 steps), which was used without further purification.

LC-MS [ESI, M+1]: 427.4

General Method C: Amide coupling on spiro-hydantoin amines.

To a stirred solution of carboxylic acid (2 eq) and DIPEA (3 eq) in DMF (1 mL) was added HATU (2 eq), which typically causes a color change to orange/brown, and the mixture was stirred for 5 min. To the mixture was then added 32 or 33 (10 to 12 mg, 1 eq) and stirred for 16 h at 25 °C. The reaction mixture was concentrated under reduced pressure, and the residue was purified by reverse-phase preparative HPLC.

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(1H-imidazole-4-carbonyl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (34).

The title compound was prepared according to General Method C, using 32 and 1H-imidazole-4-carboxylic acid, to afford 34 as a yellow solid (8.6 mg, 50%).

¹H NMR (400 MHz, Methanol-d₄) δ 9.15 – 9.09 (m, 1H), 8.21 – 8.08 (m, 1H), 8.00 (d, J = 9.4 Hz, 1H), 7.49 (s, 2H), 4.20 – 3.76 (m, 6H), 3.71 (q, J = 7.1 Hz, 4H), 3.06 (q, J = 6.5 Hz, 2H), 2.55 (d, J = 3.4 Hz, 3H), 2.52 – 2.13 (m, 2H), 1.21 (t, J = 7.1 Hz, 6H); 2 exchangeable protons not seen. LC-MS [ESI, M+1]: 507.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(2-oxo-1,2-dihydropyridine-3-carbonyl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (35).

The title compound was prepared according to General Method C, using 32 and 2-oxo-1,2-dihydropyridine-3-carboxylic acid, to afford 35 as a yellow solid (10.6 mg, 62%).

¹H NMR (400 MHz, DMSO-d₆) δ 8.72 (d, J = 18.7 Hz, 1H), 7.61 – 7.48 (m, 2H), 7.45 (d, J = 9.1 Hz, 1H), 6.67 (ddd, J = 11.8, 9.1, 2.6 Hz, 1H), 6.47 (dd, J = 12.2, 2.5 Hz, 1H), 6.27 (td, J = 6.6, 1.4 Hz, 1H), 3.70 – 3.37 (m, 10H), 3.34 (s, 3H), 2.77 (dt, J = 20.0, 6.5 Hz, 2H), 2.22 – 2.12 (m, 1H), 2.02 – 1.91 (m, 1H), 1.13 (t, J = 6.9, 6H); 1 exchangeable proton not seen. LC-MS [ESI, M+1]: 534.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(1H-imidazole-4-carbonyl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (36).

The title compound was prepared according to General Method C, using 33 and 1H-imidazole-4-carboxylic acid, to afford 36 as a yellow solid (6.6 mg, 37%).

¹H NMR (400 MHz, Methanol-d₄) δ 7.75 (s, 1H), 7.57 (s, 1H), 7.55 (d, J = 9.1 Hz, 1H), 6.74 (dd, J = 9.1, 2.6 Hz, 1H), 6.49 (d, J = 2.6 Hz, 1H), 4.54 – 4.07 (br, 2H), 3.90 – 3.41 (br, 2H, overlapping) 3.76 (t, J = 7 Hz, 2H), 3.47 (q, J = 7.0 Hz, 4H), 2.95 (t, J = 6.3 Hz, 2H), 2.41 (s, 3H), 1.90 (ddd, J = 13.6, 9.4, 4.1 Hz, 2H), 1.67 (dt, J = 14.0, 4.8 Hz, 2H), 1.21 (t, J = 7.0 Hz, 6H), 2 exchangeable protons not seen. LC-MS [ESI, M+1]: 521.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(2-oxo-1,2-dihydropyridine-3-carbonyl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (37).

The title compound was prepared according to General Method C, using 32 and 2-oxo-1,2-dihydropyridine-3-carboxylic acid, to afford 37 as a yellow solid (8.1 mg, 46%).

¹H NMR (500 MHz, DMSO-d₆) δ 8.82 (s, 1H), 7.53 (d, J = 8.9 Hz, 1H), 7.50 (dd, J = 6.8, 2.1 Hz, 1H), 7.47 (dd, J = 6.5, 2.2 Hz, 1H), 6.79 – 6.67 (m, 1H), 6.52 (br s, 1H), 6.24 (t, J = 6.6 Hz, 1H), 3.59 – 3.51 (m, 2H), 3.50 – 3.35 (m, 6H), 3.31 – 3.13 (m, 2H), 2.78 (t, J = 6.4 Hz, 2H), 2.30 (s, 3H), 1.86 – 1.63 (m, 2H), 1.52 (dd, J = 28.3, 13.3 Hz, 2H), 1.12 (t, J = 7.0, 6H). LC-MS [ESI, M+1]: 548.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(phenylsulfonyl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (38).

To a stirred solution of 32 (10 mg, 0.023 mmol, 1 eq) and DIPEA (12 μL) in DMF (1 mL) was added benzenesulfonyl chloride (3.8 μL, 0.03 mmol, 1.3 eq) on an ice-water bath. The reaction was stirred at 0 °C for 3 h, then concentrated under reduced pressure. The residue was purified by flash column chromatography (SiO₂, EtOAc/hexanes) to afford 38 as a yellow solid (12.7 mg, quant).

¹H NMR (400 MHz, DMSO-d₆) δ 8.70 (s, 1H), 7.78 – 7.70 (m, 3H), 7.66 – 7.60 (m, 2H), 7.52 (d, J = 9.1 Hz, 1H), 6.70 (dd, J = 9.1, 2.6 Hz, 1H), 6.50 (d, J = 2.6 Hz, 1H), 3.54 – 3.47 (m, 2H), 3.47 – 3.38 (m, 4H), 3.35 – 3.13 (m, 4H), 2.79 – 2.69 (m, 2H), 2.26 (s, 3H), 1.89 (td, J = 7.0, 2.2 Hz, 2H), 1.13 (t, J = 7.0 Hz, 6H). LC-MS [ESI, M+1]: 553.4

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(pyridin-3-ylsulfonyl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (39).

The title compound was prepared analogously as for 38, using pyridine-3-sulfonyl chloride, to afford 39 as a yellow solid (11.8 mg, 93%).

¹H NMR (400 MHz, DMSO-d₆) δ 8.93 (dd, J = 2.4, 0.8 Hz, 1H), 8.90 (dd, J = 4.8, 1.6 Hz, 1H), 8.65 (s, 1H), 8.15 (ddd, J = 8.0, 2.4, 1.6 Hz, 1H), 7.66 (ddd, J = 8.0, 4.8, 0.8 Hz, 1H), 7.52 (d, J = 9.1 Hz, 1H), 6.69 (dd, J = 9.1, 2.6 Hz, 1H), 6.49 (d, J = 2.6 Hz, 1H), 3.57 – 3.36 (m, 8H), 3.29 (d, J = 6.1 Hz, 2H), 2.73 (t, J = 6.4 Hz, 2H), 2.27 (s, 3H), 1.96 (ddt, J = 15.7, 13.1, 7.0 Hz, 2H), 1.13 (t, J = 7.0 Hz, 6H). LC-MS [ESI, M+1]: 554.3

General Method D: Nucleophilic aromatic substitution.

A solution of secondary amine (32, 33, or 50; 10 mg, 1 eq), aryl chloride (2 eq), and DIPEA (3 eq) in DMF (1 mL) was stirred at 100 °C for 16 h. The dark yellow to brown mixtures were concentrated under reduced pressure and purified by flash column chromatography (SiO₂, 10 to 100% EtOAc in hexanes) or reverse-phase preparative HPLC.

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(pyrimidin-2-yl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (40).

The title compound was prepared according to General Method D, using 32 and 2-chloropyrimidine, to afford 40 as a yellow solid (8.6 mg, 68%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.80 – 8.50 (br s, 2H), 7.96 – 7.85 (br d, 1H), 7.50 – 7.10 (br m, 2H), 7.07 (t, J = 5.3 Hz, 1H), 4.06 – 3.94 (m, 2H), 3.94 – 3.86 (m, 2H), 3.83 (t, J = 6.1 Hz, 2H), 3.65 (q, J = 7.3 Hz, 4H), 3.05 (td, J = 6.0, 3.7 Hz, 2H), 2.60 – 2.54 (m, 1H), 2.53 (s, 3H), 2.36 (ddd, J = 13.1, 7.6, 5.2 Hz, 1H), 1.21 (t, J = 7.1 Hz, 6H), 1 exchangeable proton not seen. LC-MS [ESI, M+1]: 491.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-7-(9H-purin-6-yl)-1,3,7-triazaspiro[4.4]nonane-2,4-dione (41).

The title compound was prepared according to General Method D, using 32 and 6-chloropurine, to afford 41 as a yellow solid (10.1 mg, 74%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.53 (d, J = 31.3 Hz, 1H), 8.48 – 8.40 (br s, 1H), 8.01 (d, J = 8.6 Hz, 1H), 7.59 – 7.39 (br s, 2H), 4.62 – 4.30 (m, 2H), 4.11 (dd, J = 38.3, 10.8 Hz, 2H), 3.85 (t, J = 6.1 Hz, 2H), 3.70 (q, J = 7.2 Hz, 4H), 3.08 (t, J = 6.0 Hz, 2H), 2.70 – 2.57 (br m, 1H), 2.57 (s, 3H), 2.49 – 2.33 (br m, 1H), 1.25 – 1.13 (t, J = 6.0 Hz, 6H), 2 exchangeable protons not seen. LC-MS [ESI, M+1]: 531.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(pyrimidin-2-yl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (42).

The title compound was prepared according to General Method D, using 33 and 2-chloropyrimidine, to afford 42 as a yellow solid (8.7 mg, 75%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.32 (d, J = 4.8 Hz, 2H), 7.56 (d, J = 9.1 Hz, 1H), 6.75 (dd, J = 9.1, 2.6 Hz, 1H), 6.60 (t, J = 4.8 Hz, 1H), 6.51 (d, J = 2.6 Hz, 1H), 4.35 (dt, J = 13.9, 4.8 Hz, 2H), 3.77 (t, J = 6.0 Hz, 2H), 3.57 – 3.51 (m, 2H), 3.49 (q, J = 7.1 Hz, 4H), 2.96 (t, J = 6.3 Hz, 2H), 2.42 (s, 3H), 1.84 (ddd, J = 13.9, 9.9, 4.2 Hz, 2H), 1.61 (dt, J = 13.2, 3.9 Hz, 2H), 1.22 (t, J = 7.0 Hz, 6H), 1 exchangeable proton not seen. LC-MS [ESI, M+1]: 505.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(9H-purin-6-yl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (43).

The title compound was prepared according to General Method D, using 33 and 6-chloropurine, to afford 43 as a yellow solid (9.2 mg, 69%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.50 (s, 1H), 8.33 (s, 1H), 7.96 – 7.8 (br d, J = 5.5 Hz, 1H), 7.50 – 7.05 (br m, 2H), 4.40 – 4.00 (br s, 2H), 3.81 (dd, J = 7.0, 5.3 Hz, 2H), 3.65 (q, J = 7.1 Hz, 4H), 3.04 (t, J = 6.2 Hz, 2H), 2.52 (s, 3H), 2.13 (ddd, J = 13.5, 9.2, 4.0 Hz, 2H), 1.92 (ddd, J = 14.0, 6.9, 3.8 Hz, 2H), 1.21 (t, J = 7.1 Hz, 6H), 2 exchangeable protons not seen, 2 protons obscured by H₂O peak. LC-MS [ESI, M+1]: 545.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(1H-pyrazolo[3,4-d]pyrimidin-4-yl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (44).

The title compound was prepared according to General Method D, using 33 and 4-chloro-1H-pyrazolo[3,4-d]pyrimidine, to afford 43 as a yellow solid (8.1 mg, 61%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.96 (s, 1H), 8.55 (s, 1H), 7.95 – 7.80 (br d, J = 8.6 Hz, 1H), 7.45 – 6.90 (br m, 2H), 4.82 – 4.65 (br s, 1H), 4.55 – 4.40 (br s, 1H), 4.22 – 4.07 (br s, 2H), 3.81 (t, J = 6.2 Hz, 2H), 3.64 (q, J = 7.1 Hz, 4H), 3.03 (t, J = 6.3 Hz, 2H), 2.52 (s, 3H), 2.30 – 2.07 (br d, J = 53.3 Hz, 2H), 2.07 – 1.82 (br d, J = 40 Hz, 2H), 1.21 (t, J = 7.1 Hz, 6H), 2 exchangeable protons not seen. LC-MS [ESI, M+1]: 545.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(imidazo[1,2-a]pyrazin-8-yl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (45).

The title compound was prepared according to General Method D, using 33 and 8-chloroimidazo[1,2-a]pyrazine, to afford 45 as a yellow solid (6.4 mg, 48%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.09 (d, J = 1.1 Hz, 1H), 8.06 (d, J = 5.5 Hz, 1H), 7.84 (d, J = 1.1 Hz, 1H), 7.77 – 7.64 (br s, 1H), 7.23 (d, J = 5.5 Hz, 1H), 7.07 – 6.83 (br s, 1H), 6.83 – 6.60 (s, 1H), 4.50 – 4.18 (br s, 2H), 3.79 (t, J = 6.2 Hz, 2H), 3.54 (q, J = 7.5 Hz, 4H), 3.01 (t, J = 6.2 Hz, 2H), 2.48 (s, 3H), 2.21 (ddd, J = 13.3, 8.9, 3.9 Hz, 2H), 1.98 (ddd, J = 14.0, 7.0, 3.5 Hz, 2H), 1.21 (t, J = 7.1 Hz, 6H), 1 exchangeable proton not seen, 2 protons obscured by H₂O peak. LC-MS [ESI, M+1]: 544.3

8-([1,2,4]triazolo[1,5-a]pyrazin-8-yl)-3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decane-2,4-dione (46).

The title compound was prepared according to General Method D, using 33 and 8-chloro-[1,2,4]triazolo[1,5-a]pyrazine, to afford 46 as a light yellow solid (10.9 mg, 76%).

¹H NMR (500 MHz, Methanol-d₄) δ 9.39 (s, 1H), 8.07 – 8.01 (m, 2H), 7.62 – 7.50 (m, 2H), 7.28 (d, J = 5.5 Hz, 1H), 4.55 – 4.35 (br s, 2H), 3.83 (dd, J = 7.1, 5.2 Hz, 2H), 3.73 (q, J = 7.2 Hz, 4H), 3.08 (dd, J = 7.1, 5.2 Hz, 2H), 2.57 (s, 3H), 2.26 (ddd, J = 13.3, 8.8, 3.9 Hz, 2H), 2.06 (ddd, J = 14.2, 7.3, 3.8 Hz, 2H), 1.21 (t, J = 7.2 Hz, 6H), 1 exchangeable proton not seen, 2 protons obscured by H₂O peak. LC-MS [ESI, M+1]: 545.3

2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)acetaldehyde (47).

To a stirred solution of 16 (100 mg, 0.36 mmol, 1 eq) in DCM (4 mL) was added Dess-Martin periodinane (233 mg, 0.55 mmol, 1.5 eq) on an ice bath. The mixture was stirred at 25 °C for 2 h and turned dark brown. The reaction was quenched with 20 mL saturated NaHCO₃ and extracted with EtOAc (3 × 20 mL). The combined organics were washed once with brine, dried over MgSO₄, filtered, and concentrated. The residue was quickly purified by flash column chromatography (SiO₂, 0 to 70% EtOAc/hexanes, within 5 min.) to afford 47 and minor unidentified decomposition products as a viscous, bright yellow oil (108 mg), which was used immediately in the next step.

LC-MS [ESI, M+1]: 274.2

Tert-butyl 4-amino-4-(((2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)amino)methyl)piperidine-1-carboxylate (48).

To a stirred solution of 47 (108 mg, 0.39 mmol, 1 eq), tert-butyl 4-amino-4-(aminomethyl)piperidine-1-carboxylate (108 mg, 0.47 mmol, 1.2 eq), and AcOH (67 μL, 1.18 mmol, 3 eq) in 1,2-dichloroethane (3 mL) was added NaBH(OAc)₃ (250 mg, 1.18 mmol, 3 eq), and the yellow suspension was stirred at 25 °C for 2 h, during which it turned a dark orange. The reaction was quenched with 20 mL saturated NaHCO₃ and extracted with EtOAc (3 × 20 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated. The residue was purified by reverse-phase preparative HPLC to afford the product as a TFA salt. The desired fractions were basified with NaOH and extracted with EtOAc (3 × 20 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated to afford the freebase of 48 as a viscous yellow oil (55.4 mg, 29%).

LC-MS [ESI, M+1]: 487.5

Tert-butyl 3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-2-oxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate (49).

To a stirred solution of 48 (69.2 mg, 0.14 mmol, 1 eq) in DCM (2 mL) was added a solution of carbonyldiimidazole (27.7 mg, 1.2 eq) in DCM (2 mL) dropwise over 3 min. The reaction was stirred at 25 °C for 16 h. Afterwards, another solution of carbonyldiimidazole (27.7 mg, 1.2 eq) in DCM (0.5 mL) was added at once, and the mixture was stirred for another 8 h. The mixture was concentrated under reduced pressure and purified by flash column chromatography (SiO₂, 10 to 100% EtOAc in hexanes) to afford 49 as a yellow gum.

LC-MS [ESI, M+1]: 513.4

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one (50).

To a stirred solution of 49 in MeOH (3 mL) was added concentrated HCl (1 mL) dropwise, and the reaction stirred at 25 °C for 16 h. The colorless solution was diluted with water (15 mL) and basified with addition of solid Na₂CO₃ to give a bright yellow solution, which was extracted with DCM (6 × 20 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated to afford 50 as a yellow solid (40 mg, 69% over 2 steps) which was used without further purification.

LC-MS [ESI, M+1]: 413.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(pyrimidin-2-yl)-1,3,8-triazaspiro[4.5]decan-2-one (51).

The title compound was prepared according to General Method D, using 50 and 2-chloropyrimidine, to afford 51 as a yellow solid (8.1 mg, 40%).

¹H NMR (400 MHz, Methanol-d₄) δ 8.50 (d, J = 5.1 Hz, 2H), 7.69 (d, J = 8.8 Hz, 1H), 7.05 – 6.87 (br s, 1H), 6.85 (t, J = 5.1 Hz, 1H), 6.82 – 6.60 (br s, 1H), 4.05 (ddd, J = 13.9, 7.0, 4.1 Hz, 2H), 3.80 (ddd, J = 13.9, 8.0, 3.9 Hz, 2H), 3.55 (q, J = 7.0 Hz, 4H), 3.52 (s, 2H), 3.43 (t, J = 6.8 Hz, 2H), 2.92 (t, J = 6.7 Hz, 2H), 2.51 (s, 3H), 1.90 – 1.75 (m, 4H), 1.22 (t, J = 7.1 Hz, 6H), 1 exchangeable proton not seen. LC-MS [ESI, M+1]: 491.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(1H-pyrazolo[3,4-d]pyrimidin-4-yl)-1,3,8-triazaspiro[4.5]decan-2-one (52).

The title compound was prepared according to General Method D, using 50 and 4-chloro-1H-pyrazolo[3,4-d]pyrimidine, to afford 52 as a light yellow solid (8.4 mg, 38%).

¹H NMR (500 MHz, Methanol-d₄) δ 8.98 (s, 1H), 8.55 (s, 1H), 8.02 (d, J = 8.6 Hz, 1H), 7.62 – 7.40 (overlapping br s, 2H), 4.50 – 4.27 (m, 2H), 4.27 – 4.12 (m, 2H), 3.73 (q, J = 7.2 Hz, 4H), 3.62 (d, J = 2.3 Hz, 2H), 3.49 (t, J = 6.6 Hz, 2H), 3.00 (t, J = 6.7 Hz, 2H), 2.60 (s, 3H), 2.15 – 1.86 (m, 4H), 1.21 (t, J = 7.2 Hz, 6H), 2 exchangeable protons not seen. LC-MS [ESI, M+1]: 531.3

3-(2-(7-(diethylamino)-4-methyl-2-oxo-2H-chromen-3-yl)ethyl)-8-(imidazo[1,2-a]pyrazin-8-yl)-1,3,8-triazaspiro[4.5]decan-2-one (53).

The title compound was prepared according to General Method D, using 50 and 8-chloroimidazo[1,2-a]pyrazine, to afford 53 as a brown solid (4.3 mg, 20%).

¹H NMR (400 MHz, CDCl₃) δ = 7.54 (s, 1H), 7.52 (d, J = 4.8 Hz, 1H), 7.49 (d, J = 8.8 Hz, 1H), 7.41 (d, J = 7.2 Hz, 1H), 7.34 (d, J = 4.9 Hz, 1H), 6.60 (dd, J = 2.4, 8.8 Hz, 1H), 6.49 (d, J = 2.4 Hz, 1H), 6.86 (s, 1H), 5.67 (d, J = 2.4 Hz, 1H), 4.67 (s, 1H), 4.35 – 4.31 (m, 2H), 4.20 – 4.18 (m, 2H), 3.44 (s, 3H), 3.42 – 3.39 (m, 5H), 2.88 (t, J = 7.2 Hz, 2H), 2.43 (s, 3H), 1.87–1.82 (m, 4H), 1.21 (t, J = 7.2 Hz, 6H). ¹³C NMR (126 MHz, CDCl₃) δ 162.88, 160.54, 154.65, 149.93, 149.45, 149.01, 134.23, 131.35, 127.87, 125.65, 116.31, 114.16, 111.14, 109.58, 108.60, 97.46, 56.22, 53.31, 44.70, 43.16, 42.27, 37.37, 25.74, 14.81, 12.49. ESI-HRMS: calc’d for C₂₉H₃₅N₇O₃ [M+H]⁺ 530.2874, found 530.2874

General Method E: Williamson ether synthesis.

To a mixture of 7-hydroxy-4-methylcoumarin (50 to 175 mg, 1 eq) and K₂CO₃ (3 eq) in DMF (0.3 M) was added alkyl bromide or tosylate (1.5 eq) and stirred at 80 °C for at least 4 h. The suspension typically became yellow within 30 min. and turned into a white suspension when the reaction was complete. The mixture was diluted ten-fold with water and extracted with diethyl ether (3 × 15 mL). The combined organics were washed once with brine, dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography to provide the coumarin ether products.

4-methyl-7-(1-phenylpropoxy)-2H-chromen-2-one (55).

The title compound was prepared according to General Method E, using (1-bromopropyl)benzene, to afford 55 as a white solid (158.4 mg, 54%).

¹H NMR (400 MHz, Chloroform-d) δ 7.44 (d, J = 8.8 Hz, 1H), 7.40 – 7.24 (m, 5H), 6.88 (dd, J = 8.8, 2.5 Hz, 1H), 6.74 (d, J = 2.5 Hz, 1H), 6.10 (d, J = 0.9 Hz, 1H), 5.09 (dd, J = 7.3, 5.6 Hz, 1H), 2.36 (d, J = 1.3 Hz, 3H), 2.13 – 2.10 (m, 1H), 2.00 – 1.87 (m, 1H), 1.03 (t, J = 7.4 Hz, 3H). LC-MS [ESI, M+1]: 295.2

4-methyl-7-((1-phenylpropan-2-yl)oxy)-2H-chromen-2-one (56).

The title compound was prepared according to General Method E, using 1-phenylpropan-2-yl 4-methylbenzenesulfonate, to afford 56 as a white solid (154.5 mg, 70%).

¹H NMR (500 MHz, Chloroform-d) δ 7.49 (d, J = 8.5 Hz, 1H), 7.38 – 7.21 (m, 5H), 6.88 – 6.81 (m, 2H), 6.14 (q, J = 1.2 Hz, 1H), 4.67 (q, J = 6.1 Hz, 1H), 3.13 (dd, J = 13.8, 6.1 Hz, 1H), 2.89 (dd, J = 13.8, 6.5 Hz, 1H), 2.41 (d, J = 1.2 Hz, 3H), 1.38 (d, J = 6.1 Hz, 3H). LC-MS [ESI, M+1]: 295.2

4-methyl-7-(2-phenylpropoxy)-2H-chromen-2-one (57).

The title compound was prepared according to General Method E, using 2-phenylpropyl 4-methylbenzenesulfonate, to afford 57 as a white solid (60.3 mg, 71%).

¹H NMR (400 MHz, Chloroform-d) δ 7.49 (d, J = 8.7 Hz, 1H), 7.41 – 7.25 (m, 5H), 6.86 (dd, J = 8.8, 2.5 Hz, 1H), 6.82 (d, J = 2.5 Hz, 1H), 6.15 (q, J = 1.2 Hz, 1H), 4.16 (dd, J = 9.2, 6.1 Hz, 1H), 4.06 (dd, J = 9.2, 7.5 Hz, 1H), 3.30 (sextet, J = 7.0 Hz, 1H), 2.41 (d, J = 1.2 Hz, 3H), 1.46 (d, J = 7.0 Hz, 3H). LC-MS [ESI, M+1]: 295.2

4-methyl-7-(1-(o-tolyl)ethoxy)-2H-chromen-2-one (58).

The title compound was prepared according to General Method E, using 1-(1-bromoethyl)-2-methylbenzene, to afford 58 as a white solid (172.7 mg, 78%).

¹H NMR (500 MHz, Chloroform-d) δ 7.45 (d, J = 8.8 Hz, 1H), 7.39 – 7.33 (m, 1H), 7.23 – 7.13 (m, 3H), 6.84 (dd, J = 8.8, 2.5 Hz, 1H), 6.61 (d, J = 2.5 Hz, 1H), 6.10 (q, J = 1.3 Hz, 1H), 5.53 (q, J = 6.4 Hz, 1H), 2.46 (s, 3H), 2.37 (d, J = 1.2 Hz, 3H), 1.67 (d, J = 6.4 Hz, 3H). LC-MS [ESI, M+1]: 295.2

7-(1-(2-fluorophenyl)ethoxy)-4-methyl-2H-chromen-2-one (59).

The title compound was prepared according to General Method E, using 1-(1-bromoethyl)-2-fluorobenzene, to afford 59 as a white solid (185.6 mg, 83%).

¹H NMR (400 MHz, Chloroform-d) δ 7.46 (d, J = 8.8 Hz, 1H), 7.40 (td, J = 7.6, 1.8 Hz, 1H), 7.33 – 7.23 (m, 1H), 7.22 – 7.05 (m, 2H), 6.88 (dd, J = 8.8, 2.5 Hz, 1H), 6.76 (d, J = 2.5 Hz, 1H), 6.12 (q, J = 1.3 Hz, 1H), 5.72 (q, J = 6.4 Hz, 1H), 2.38 (d, J = 1.2 Hz, 3H), 1.71 (d, J = 6.4 Hz, 3H). LC-MS [ESI, M+1]: 298.1

3-bromo-4-methyl-7-(1-phenylethoxy)-2H-chromen-2-one (60).

To a stirred solution of 2 (50 mg, 0.18 mmol, 1 eq) in anhydrous MeCN (2 mL) was added (35.6 mg, 0.2 mmol, 1.1 eq) and the reaction stirred at room temperature for 16 hours. The colorless solution was concentrated under reduced pressure and purified by flash column chromatography (SiO₂, 0 to 20% EtOAc in hexanes) to afford 60 as a white solid (58.2 mg, 90%).

3-bromo-4-methyl-7-(2-phenylpropoxy)-2H-chromen-2-one (61).

The title compound was prepared analogously as for 60, using 57 as the starting material, to afford 61 as a white solid (55.2 mg, 87%).

4-methyl-3-(morpholinomethyl)-7-(1-phenylethoxy)-2H-chromen-2-one (62).

A mixture of 60 (20 mg, 0.056 mmol, 1 eq), potassium (morpholin-4-yl)methyltrifluoroborate (18 mg, 0.084 mol, 1.5 eq), RuPhos (5 mg, 0.011 mmol, 0.2 eq) and Cs₂CO₃ (55 mg, 0.17 mmol, 3 eq) in toluene (0.375 mL) and H₂O (0.125 mL) in a microwave vial was sparged with argon for 2 min. while stirring. To the mixture was added Pd(dba)₂ (3.2 mg, 0.056 mmol, 0.1 eq) and further sparged for an additional minute. The vial was sealed and heated at 90 °C for 16 h. The initially red biphasic mixture turned dark yellow-orange, and a black precipitate appeared. The reaction was then diluted with water (15 mL) and extracted with EtOAc (3 × 10 mL). The combined organics were washed once with brine, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. The residue was purified by reverse-phase preparative HPLC to afford 62 as a viscous, colorless oil (3.5 mg, 17%).

¹H NMR (500 MHz, Chloroform-d) δ 7.51 (d, J = 8.9 Hz, 1H), 7.39 – 7.34 (m, 5H), 6.89 (dd, J = 8.9, 2.5 Hz, 1H), 6.73 (d, J = 2.5 Hz, 1H), 5.37 (q, J = 6.4 Hz, 1H), 3.73 – 3.62 (br s, 4H), 3.60 – 3.47 (br s, 2H), 2.64 (s, 2H), 2.60 – 2.50 (br s, 2H), 2.48 (s, 3H), 1.70 (d, J = 6.4 Hz, 3H). LC-MS [ESI, M+1]: 380.3

4-methyl-3-(morpholinomethyl)-7-(2-phenylpropoxy)-2H-chromen-2-one (63).

The title compound was prepared analogously as for 62, using 61 as the starting material, to afford 63 as a white solid.

¹H NMR (500 MHz, Chloroform-d) δ 7.61 (d, J = 8.9 Hz, 1H), 7.42 – 7.34 (m, 2H), 7.34 – 7.26 (m, 3H), 6.92 (dd, J = 9.0, 2.5 Hz, 1H), 6.80 (d, J = 2.5 Hz, 1H), 4.33 (s, 2H), 4.18 (dd, J = 9.1, 6.1 Hz, 1H), 4.08 (dd, J = 9.2, 7.5 Hz, 1H), 4.05 – 3.87 (br s, 4H), 3.65 – 3.45 (br s, 2H), 3.31 (sextet, J = 7.0 Hz, 1H), 3.27 – 3.08 (br s, 2H), 2.59 (s, 3H), 1.46 (d, J = 7.0 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 163.43, 162.48, 157.52, 154.78, 142.80, 128.67, 127.35, 127.14, 126.96, 113.81, 112.99, 109.94, 101.30, 74.07, 63.85, 52.62, 52.02, 39.35, 18.03, 16.30. LC-MS [ESI, M+1]: 394.2

3-bromo-7-methoxy-4-methyl-2H-chromen-2-one (65).

The title compound was prepared analogously to 60, using commercially available 7-methoxy-4-methyl-2H-chromen-2-one (64) as starting material, to afford 65 as a yellow solid (25 g, 75%).

¹H NMR (400 MHz, DMSO-d₆) δ = 7.80 (d, J = 8.8 Hz, 1H), 7.04 (d, J = 2.4 Hz, 1H), 7.00 (dd, J = 2.4, 8.8 Hz, 1H), 3.87 (s, 3H), 2.58 (s, 3H). LC-MS [ESI, M+1]: 269.1

3-bromo-7-hydroxy-4-methyl-2H-chromen-2-one (66).

To a solution of 65 (20.0 g, 74.3 mmol, 1.00 eq) in DCM (200 mL) was added BBr₃ (548 g, 372 mmol, 211 mL, 5.00 eq) at −78 °C. The mixture was stirred at 30 °C for 6 h, and then poured into ice water (200 mL) and extracted with ethyl acetate (300 mL × 3). The combined organic phase was washed with brine (500 mL × 2), dried with anhydrous sodium sulfate, filtered and concentrated under vacuum to give 66 (15.0 g, 76%) as a yellow solid, which was used without further purification.

¹H NMR (400 MHz, DMSO-d6) δ = 10.68 (s, 1H), 7.70 (d, J = 8.8 Hz, 1H), 6.83 (dd, J = 2.4, 8.8 Hz, 1H), 6.73 (d, J = 2.4 Hz, 1H), 2.54 (s, 3H). LC-MS [ESI, M+1]: 255.0

(E)-3-(2-ethoxyvinyl)-7-hydroxy-4-methyl-2H-chromen-2-one (67).

A mixture of 66 (1.50 g, 5.88 mmol, 1.00 eq), 2-[(E)-2-ethoxyvinyl]-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (1.75 g, 8.82 mmol, 1.50 eq), Pd(dppf)Cl₂ (430 mg, 588 μmol, 0.10 eq) and K₂CO₃ (2.44 g, 17.6 mmol, 3.00 eq) in dioxane (12.0 mL) and H₂O (3.00 mL) was degassed and purged with N₂ three times, and then the mixture was stirred at 100 °C for 2 h under N₂ atmosphere. The mixture was diluted with H₂O (20 mL) and extracted with ethyl acetate (20 mL × 3). The combined organic phase was washed with brine (50 mL × 2), dried with anhydrous sodium sulfate, filtered and concentrated under vacuum. The residue was purified by silica gel column chromatography (PE/EA = 20/1 to 5/1) to give 67 (1.50 g, 89%) as a yellow solid.

LC-MS [ESI, M+1]: 247.1

(R,E)-3-(2-ethoxyvinyl)-4-methyl-7-(1-phenylethoxy)-2H-chromen-2-one [(R)-68].

To a solution of 67 (1 g, 4.06 mmol, 1 eq) and (S)-1-phenylethanol (744 mg, 6.09 mmol, 737 μL, 1.50 eq) in THF (10.0 mL) was added PPh₃ (1.60 g, 6.09 mmol, 1.50 eq) and DIAD (1.23 g, 6.09 mmol, 1.18 mL, 1.50 eq). The mixture was stirred at 30 °C for 1 h. The mixture was diluted with H₂O (20 mL) and extracted with ethyl acetate (20 mL × 3). The combined organic phase was washed with brine (50 mL × 1), dried with anhydrous sodium sulfate, filtered and concentrated under vacuum. The residue was purified by silica gel column chromatography (PE/EA = 10/1 to 3/1) to give (R)-68 (1.00 g, 55%) as a yellow oil.

¹H NMR (400 MHz, DMSO-d₆) δ = 7.64 (d, J = 8.8 Hz, 1H), 7.53 (d, J = 12.4 Hz, 1H), 7.33 – 7.24 (m, 5H), 6.93 (dd, J = 2.4, 8.8 Hz, 1H), 6.87 (d, J = 2.4 Hz, 1H), 5.79 (d, J = 12.4 Hz, 1H), 5.65 (q, J = 6.4 Hz, 1H), 3.92 (q, J = 7.2 Hz, 2H), 2.35 (s, 3H), 1.58 (d, J = 6.4 Hz, 3H), 1.26 – 1.22 (m, 1H), 1.24 (t, J = 7.2 Hz, 3H).

LC-MS [ESI, M+1]: 351.2

(S,E)-3-(2-ethoxyvinyl)-4-methyl-7-(1-phenylethoxy)-2H-chromen-2-one [(S)-68].

The title compound was prepared analogously to 60, using (R)-1-phenylethanol, to afford (S)-68 as a yellow solid (1.4 g, 37%).

¹H NMR (400 MHz, DMSO-d₆) δ = 7.65 (d, J = 8.8 Hz, 1H), 7.53 (d, J = 12.4 Hz, 1H), 7.43 (d, J = 7.2 Hz, 2H), 7.35 (t, J = 7.6 Hz, 2H), 7.27 (d, J = 7.2 Hz, 1H), 6.93 (dd, J = 2.4, 8.8 Hz, 1H), 6.87 (d, J = 2.4 Hz, 1H), 5.79 (d, J = 12.4 Hz, 1H), 5.66 (q, J = 6.4 Hz, 1H), 3.92 (q, J = 7.2 Hz, 2H), 2.35 (s, 3H), 1.58 (d, J = 6.4 Hz, 3H), 1.24 (t, J = 7.2 Hz, 3H).

LC-MS [ESI, M+1]: 351.2

(E)-3-(2-ethoxyvinyl)-4-methyl-7-(2-phenylpropoxy)-2H-chromen-2-one (69).

The title compound was prepared analogously to 60, using racemic 2-phenyl-1-propanol, to afford racemic 69 as a yellow solid (1.16 g, 69%).

LC-MS [ESI, M+1]: 365.1

(R)-2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)acetaldehyde [(R)-70].

To a solution of (R)-68 (1.10 g, 2.61 mmol, 1.00 eq) in acetone (10.0 mL) was added HCl (12.0 M, 2.34 mL, 10.7 eq). The mixture was stirred at 30 °C for 0.5 h. The reaction mixture was diluted with water (10.0 mL) and extracted with ethyl acetate (20 mL × 3). The combined organic layers were washed with brine (10 mL), dried over anhydrous sodium sulfate, filtered, and concentrated under reduced pressure to give a residue. The residue was purified by column chromatography (SiO₂, Petroleum ether/Ethyl acetate = 10/1 to 2/1) to give (R)-70 (400 mg, 33%) as a yellow oil.

LC-MS [ESI, M+1]: 323.2

(S)-2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)acetaldehyde [(S)-70].

The title compound was prepared analogously to (R)-70, using (S)-68 as starting material, to give (S)-70 (380 mg, 41%) as a yellow oil.

LC-MS [ESI, M+1]: 323.1

2-(4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)acetaldehyde (71).

The title compound was prepared analogously to (R)-70, using 69 as starting material, to give 71 (560 mg, 44%) as a yellow oil.

LC-MS [ESI, M+1]: 337.0

¹H NMR (400 MHz, DMSO-d₆) δ = 9.65 (s, 1H), 7.70 (d, J = 8.8 Hz, 1H), 7.38 – 7.28 (m, 5H), 6.98 (d, J = 2.4 Hz, 1H), 6.94 (dd, J = 2.4, 8.8 Hz, 1H), 4.24 – 4.16 (m, 2H), 3.78 (s, 2H), 3.27 – 3.20 (m, 1H), 2.32 (s, 3H), 1.33 (d, J = 7.0 Hz, 3H).

3-(hydroxymethyl)-4-methyl-7-(2-phenylpropoxy)-2H-chromen-2-one (72).

A mixture of 61 (1.50 g, 4.02 mmol, 1.00 eq), tributylstannylmethanol (2.58 g, 8.04 mmol, 2.00 eq) and Pd(PPh₃)₄ (464 mg, 402 μmol, 0.10 eq) in dioxane (10.0 mL) was degassed and purged with N₂ three times, and then the mixture was stirred at 100 °C for 16 h under a N₂ atmosphere. The mixture was diluted with H₂O (20 mL) and extracted with ethyl acetate (20 mL × 3). The combined organic phase was washed with brine (50 mL × 1) and dried with anhydrous sodium sulfate, filtered and concentrated under vacuum. The residue was purified by silica gel column chromatography (PE/EA = 10/1 to 3/1) to give 72 (350 mg, 22%) as a light yellow oil.

LC-MS [ESI, M+1]: 325.0

4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromene-3-carbaldehyde (73).

To a solution of 3-(hydroxymethyl)-4-methyl-7-(2-phenylpropoxy)chromen-2-one (300 mg, 925 μmol, 1.00 eq) in DCM (2.00 mL) was added MnO₂ (804 mg, 9.25 mmol, 10.0 eq). The mixture was stirred at 30 °C for 24 h. The mixture was diluted with H₂O (10 mL) and extracted with ethyl acetate (10 mL × 3). The combined organic phase was washed with brine (20 mL × 2), dried with anhydrous sodium sulfate, filtered and concentrated under vacuum. The residue was purified by column chromatography (SiO₂, DCM/MeOH = 100/1 to 10/1) to give 73 (150 mg, 49%) as a yellow oil.

LC-MS [ESI, M+1]: 323.1

Tert-butyl (R)-4-amino-4-(((2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)amino)methyl)piperidine-1-carboxylate [(R)-74].

To a solution of (R)-70 (350 mg, 1.09 mmol, 1.10 eq) and tert-butyl 4-amino-4-(aminomethyl)piperidine-1-carboxylate (226 mg, 987 μmol, 1.00 eq) in MeOH (5.00 mL) was added AcOH (59.3 mg, 987 μmol, 56.45 μL, 1.00 eq) and NaBH₃CN (186 mg, 2.96 mmol, 3.00 eq) at 0 °C. The mixture was stirred at 30 °C for 3 h. The pH of the mixture was adjusted with saturated NaHCO₃ aqueous solution to pH 8. Then the solution was extracted with DCM (10 mL × 3), and the combined organic layer was washed with brine (20 mL × 2), dried over anhydrous sodium sulfate. The mixture was filtered and the filtrate was concentrated. The residue was partially purified by column chromatography (SiO₂, DCM/MeOH = 200/1 to 20/1) to give (R)-74 (200 mg, crude) as a white solid.

LC-MS [ESI, M+1]: 536.4.

Tert-butyl (S)-4-amino-4-(((2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)amino)methyl)piperidine-1-carboxylate [(S)-74].

The title compound was prepared analogously to (R)-74, using (S)-70 as starting material, to give (S)-74 (600 mg, 57%) as a white solid.

LC-MS [ESI, M+1]: 536.4.

Tert-butyl 4-amino-4-(((2-(4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)ethyl)amino)methyl)piperidine-1-carboxylate (75).

The title compound was prepared analogously to (R)-74, using 71 as starting material, to give 75 (150 mg, 33%) as a light yellow gum.

LC-MS [ESI, M+1]: 550.4

¹H NMR (400 MHz, DMSO-d₆) δ = 7.68 (d, J = 8.6 Hz, 1H), 7.38 – 7.30 (m, 4H), 7.26 – 7.20 (m, 1H), 6.98 – 6.91 (m, 2H), 4.28 – 4.09 (m, 2H), 3.52 – 3.44 (m, 4H), 3.26 – 3.19 (m, 4H), 2.82 – 2.64 (m, 6H), 2.42 (s, 3H), 1.68 – 1.58 (m, 2H), 1.57 – 1.44 (m, 2H), 1.40 (s, 9H), 1.33 (d, J = 7.0 Hz, 3H).

Tert-butyl 4-amino-4-((((4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)methyl)amino)methyl)piperidine-1-carboxylate (76).

The title compound was prepared analogously to (R)-74, using 73 as starting material, to give 76 (150 mg, 60%) as a light yellow gum.

LC-MS [ESI, M+1]: 536.3

Tert-butyl (R)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-2-oxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate [(R)-77].

To a solution of (R)-74 (200 mg, 373 μmol, 1.00 eq) in DMF (2.00 mL) was added CDI (182 mg, 1.12 mmol, 3.00 eq) at 0 °C. The mixture was stirred at 30 °C for 1 hr. The reaction mixture was diluted with water (10 mL) and extracted with ethyl acetate (10 mL × 3). The combined organic layers were washed with brine (10 mL × 1), dried over anhydrous sodium sulfate, filtered and concentrated under reduced pressure to give a residue. The residue was purified by column chromatography (SiO₂, Petroleum ether/Ethyl acetate = 10/1 to 1/1) to give (R)-77 (150 mg, 15% over two steps) as a white solid.

LC-MS [ESI, M+1]: 562.3

Tert-butyl (S)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-2-oxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate [(S)-77].

The title compound was prepared analogously to (R)-77, using (S)-74 as starting material, to give (S)-77 (310 mg, 45%) as a white solid.

LC-MS [ESI, M+1]: 562.3

Tert-butyl 3-(2-(4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)ethyl)-2-oxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate (78).

The title compound was prepared analogously to (R)-77, using 75 as starting material, to give 78 (80 mg, 62%) as a white solid.

LC-MS [ESI, M+1]: 576.3

Tert-butyl 3-((4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)methyl)-2-oxo-1,3,8-triazaspiro[4.5]decane-8-carboxylate (79).

The title compound was prepared analogously to (R)-77, using 76 as starting material, to give 79 (130 mg, 73%) as a light yellow oil.

LC-MS [ESI, M+1]: 562.3

(R)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one [(R)-80].

To a solution of (R)-77 (150 mg, 267 μmol, 1.00 eq) in DCM (0.50 mL) was added HCl/dioxane (4.00 M, 0.50 mL, 7.49 eq) at 0 °C. The mixture was stirred at 30 °C for 0.1 h. The mixture was diluted with H₂O (2 mL × 1), extracted with ethyl acetate (5 mL × 3). The combined organic phase was washed with brine (10 mL × 1), dried with anhydrous sodium sulfate, filtered and concentrated under vacuum to give (R)-80 (80.0 mg, crude) as a white solid.

LC-MS [ESI, M+1]: 462.3

(S)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one [(S)-80].

The title compound was prepared analogously to (R)-80, using (S)-77 as starting material, to give (S)-80 (80 mg, crude) as a white solid.

LC-MS [ESI, M+1]: 462.2

3-(2-(4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one (81).

The title compound was prepared analogously to (R)-80, using 78 as starting material, to give 81 (50 mg, crude) as a white solid.

LC-MS [ESI, M+1]: 476.3

3-((4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)methyl)-1,3,8-triazaspiro[4.5]decan-2-one (82).

The title compound was prepared analogously to (R)-80, using 79 as starting material, to give 82 (70 mg, crude) as a light yellow solid.

LC-MS [ESI, M+1]: 462.2

(R)-8-(imidazo[1,2-a]pyrazin-8-yl)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one [(R)-83].

The title compound was prepared according to General Method D, using (R)-80 and 8-chloroimidazo[1,2-a]pyrazine, to afford (R)-83 as a brown solid (6 mg, 15%)

¹H NMR (400 MHz, DMSO-d₆) δ = 7.92 (d, J = 0.8 Hz, 1H), 7.89 (d, J = 4.4 Hz, 1H), 7.65 (d, J = 8.8 Hz, 1H), 7.54 (d, J = 0.8 Hz, 1H), 7.46 – 7.41 (m, 2H), 7.38 – 7.30 (m, 3H), 7.28 – 7.23 (m, 1H), 6.94 (dd, J = 2.4, 8.8 Hz, 1H), 6.88 (d, J = 2.5 Hz, 1H), 6.85 (s, 1H), 5.67 (q, J = 6.4 Hz, 1H), 4.38 – 4.28 (m, 2H), 4.18 – 4.06 (m, 2H), 3.27 (br s, 2H), 3.20 (br t, J = 6.8 Hz, 2H), 2.71 (br t, J = 6.8 Hz, 2H), 2.38 (s, 3H), 1.67 – 1.61 (m, 2H), 1.58 (br d, J = 6.4 Hz, 5H). ¹³C NMR (101 MHz, DMSO-d₆) δ = 161.00, 159.96, 159.67, 152.98, 148.48, 142.30, 133.21, 131.15, 128.67, 127.69, 127.40, 126.46, 119.84, 115.15, 113.66, 113.36, 111.55, 102.33, 75.39, 52.67, 42.40, 36.79, 25.43, 24.01. ESI-HRMS: calc’d for C₃₃H₃₄N₆O₄ [M+H]⁺ 579.2714, found 579.2717

(S)-8-(imidazo[1,2-a]pyrazin-8-yl)-3-(2-(4-methyl-2-oxo-7-(1-phenylethoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one [(S)-83].

The title compound was prepared according to General Method D, using (S)-80 and 8-chloroimidazo[1,2-a]pyrazine, to afford (S)-83 as a brown solid (0.92 mg, 0.9%).

1H NMR (400 MHz, DMSO-d₆) δ = 7.93 (s, 1H), 7.90 (d, J = 4.8 Hz, 1H), 7.65 (d, J = 8.8 Hz, 1H), 7.56 (s, 1H), 7.44 (d, J = 7.2 Hz, 2H), 7.37 – 7.31 (m, 3H), 7.28 – 7.23 (m, 1H), 6.95 (dd, J = 2.4, 8.8 Hz, 1H), 6.88 (d, J = 2.4 Hz, 1H), 6.86 (s, 1H), 5.67 (d, J = 6.6 Hz, 1H), 4.39 – 4.26 (m, 2H), 4.19 – 4.07 (m, 2H), 3.28 – 3.26 (m, 2H), 3.21 (br t, J = 7.2 Hz, 2H), 2.72 (br t, J = 7.2 Hz, 2H), 2.38 (s, 3H), 1.66 – 1.57 (m, 7H). ESI-HRMS: calc’d for C₃₃H₃₄N₆O₄ [M+H]⁺ 579.2714, found 579.2714

8-(imidazo[1,2-a]pyrazin-8-yl)-3-(2-(4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)ethyl)-1,3,8-triazaspiro[4.5]decan-2-one (84).

The title compound was prepared according to General Method D, using 81 and 8-chloroimidazo[1,2-a]pyrazine, to afford 84 as a white solid (10 mg, 23% over two steps). The racemate was resolved into the individual enantiomers (84a and 84b) using chiral SFC (Daicel CHIRALPAK® AD, 250 mm × 30 mm, 10 μm pore size, eluent 0.08% NH₄OH, 80% iPrOH/CO₂). The absolute stereochemistry of each enantiomer was not determined.

¹H NMR (400 MHz, DMSO-d₆) δ = 7.92 (d, J = 1.2 Hz, 1H), 7.89 (d, J = 4.8 Hz, 1H), 7.69 (d, J = 8.8 Hz, 1H), 7.54 (d, J = 1.2 Hz, 1H), 7.38 – 7.28 (m, 5H), 7.25 – 7.20 (m, 1H), 6.96 (d, J = 2.4 Hz, 1H), 6.92 (dd, J = 2.4, 8.8 Hz, 1H), 6.86 (s, 1H), 4.44 – 4.01 (m, 6H), 3.30 – 3.28 (m, 2H), 3.25 – 3.19 (m, 3H), 2.79 – 2.72 (m, 2H), 2.45 – 2.39 (m, 3H), 1.73 – 1.52 (m, 4H), 1.32 (d, J = 6.8 Hz, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ = 161.08, 160.83, 159.98, 153.26, 148.48, 148.07, 143.52, 133.22, 131.14, 128.41, 127.43, 126.52, 119.75, 115.13, 113.66, 112.52, 111.54, 101.06, 73.11, 55.23, 52.68, 42.41, 41.62, 38.69, 36.78, 18.35, 14.90. ESI-HRMS: calc’d for C₃₄H₃₆N₆O₄ [M+H]⁺ 593.2871, found 593.2870 for 84a, found 598.2868 for 84b

8-(imidazo[1,2-a]pyrazin-8-yl)-3-((4-methyl-2-oxo-7-(2-phenylpropoxy)-2H-chromen-3-yl)methyl)-1,3,8-triazaspiro[4.5]decan-2-one (85).

The title compound was prepared according to General Method D, using 82 and 8-chloroimidazo[1,2-a]pyrazine, to afford 85 as an off-white solid (10 mg, 29%). The racemate was resolved into the individual enantiomers (85a and 85b) using chiral SFC (Daicel CHIRALPCEL® OJ, 250 mm × 30 mm, 10 μm pore size, eluent 0.06% NH₄OH, 80% EtOH/CO₂). The absolute stereochemistry of each enantiomer was not determined.

¹H NMR (400 MHz, DMSO-d₆) δ = 7.90 (d, J = 0.8 Hz, 1H), 7.87 (d, J = 4.6 Hz, 1H), 7.73 (d, J = 8.8 Hz, 1H), 7.52 (d, J = 0.8 Hz, 1H), 7.37 – 7.28 (m, 5H), 7.25 – 7.20 (m, 1H), 7.01 (s, 1H), 6.98 (d, J = 2.4 Hz, 1H), 6.94 (dd, J = 2.4, 8.8 Hz, 1H), 4.33 – 4.11 (m, 8H), 3.26 – 3.21 (m, 1H), 3.16 (s, 2H), 2.48 (s, 3H), 1.68 – 1.54 (m, 4H), 1.32 (d, J = 6.8 Hz, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ = 161.41, 161.35, 159.61, 153.73, 151.13, 148.45, 143.45, 133.19, 131.09, 128.40, 127.40, 127.35, 117.69, 115.08, 113.57, 112.66, 111.48, 101.08, 73.15, 55.19, 52.87, 42.31, 38.66, 36.69, 18.31, 14.74. LC-MS [ESI, M+1]: 579.

Abbreviations Used

AL

Light chain amyloidosis

CDR3

Complementarity-determining region 3

C_L

Constant domain of light chain

FL

Full-length

JTO

A non-amyloidogenic light chain named after the multiple myeloma patient it was originally derived from¹¹. The full name is unknown

LC

Light chain

PCFP

Proteolysis-coupled fluorescence polarization

V_L

Variable domain of light chain

WIL

An amyloidogenic light chain named after the AL patient it was originally derived from¹¹. The full name is unknown

WIL-FL*

Fluorescein-labeled WIL full-length light chain at residue 79