2.1. Assays

Table 1 summarizes the details for the assays that drove this probe project.

2.2. Probe Chemical Characterization

b. Probe chemical structure including stereochemistry if known

The probe ML233 was obtained as a single observed anti isomer. The assignment is consistent with precedent literature (19).

ML233

c. Synthesis and Structural Verification Information of probe SID99361200 corresponding to CID46905036

Scheme 1Synthesis of ML233, conditions

a. Sodium nitrite (1.5 eq.), HCl (conc.), EtOH; b. benzenesulfonyl chloride (1 eq.), DMAP (cat.), pyridine.

For detailed procedures see Section 2.3. Images of spectral data (¹HNMR, ¹³CNMR, and LC/MS) used to support the structural assignment of ML233 can be found in Section 6 (Supplementary Information).

d. If available from a vendor, please provide details

This probe is not commercially available. A 25 mg sample of ML233 synthesized at SBCCG has been deposited in the MLSMR (see Probe Submission Table 3 below).

Table 3

Probe and Analog Submissions to MLSMR (BioFocus DPI) for APJ Agonists.

e. Solubility and Stability of probe in PBS at room temperature

The stability and solubility of ML233 was investigated in PBS buffer at room temperature (Figure 2). Initial experiments examining the stability of ML233 revealed no quantifiable level of compound after 1 hr in PBS buffer due to low solubility. Subsequent examination in 50% acetonitrile:PBS (1:1) showed the compound was mostly stable out to 48 hr (71% remaining).

Figure 2

Stability of ML233 in PBS.

2.3. Probe Preparation

Step 1: 2-cyclohexyl-5-methylphenol (0.5g, 2.63mmol) was dissolved in 10 ml ethanol followed by addition of 10 ml concentrated HCl. The resultant mixture was cooled to 0 °C and sodium nitrite (272mg, 3.94mmol) was added to it in two portions of about 140mg each. After addition of sodium nitrite, the reaction mixture turned green. The reaction mixture was stirred overnight at room temperature followed by addition of ice-cold water, which resulted in precipitation of a yellow solid. The aqueous solution was filtered under vacuum to yield 0.55g (96% yield) of 2-cyclohexyl-5-methyl-4-nitrosophenol (2) as a yellow solid. ¹H NMR (500 MHz, CDCl₃) δ 8.69 (s, 1H), 7.47 (s, 1H), 6.30 (d, J = 1.2 Hz, 1H), 2.74 (t, J = 12.0 Hz, 1H), 2.15 (d, J = 1.1 Hz, 3H), 1.88 – 1.65 (m, 5H), 1.49 – 1.31 (m, 2H), 1.29 – 1.07 (m, 3H).

Step 2: 2-cyclohexyl-5-methyl-4-nitrosophenol (50mg, 0.23mmol) was dissolved in 2ml pyridine followed by addition of catalytic amount of DMAP (N, N′-Dimethylaminopyridine). Benzenesulfonyl chloride (40mg, 0.23mmol) was added to the reaction mixture and the reaction mixture was stirred for 4–6h. The reaction mixture was partitioned with approximately 20 ml of 1:1 ethyl acetate and water and the organic layer was washed with 10% HCl (2×20ml) and brine (1×20ml). The organic layer was concentrated in vacuo and the crude product was purified by column chromatography to yield 60mg (73%) of 4-methyl-5-(((phenylsulfonyl)oxy)imino)-[1,1′-bi(cyclohexane)]-3,6-dien-2-one as a yellow solid. ¹H NMR (500 MHz, CDCl₃) δ 8.12 – 7.99 (m, 2H), 7.73 (t, J = 7.5 Hz, 1H), 7.62 (t, J = 7.8 Hz, 2H), 7.28 (d, J = 0.8 Hz, 1H), 6.33 (d, J = 1.3 Hz, 1H), 2.75 (dt, J = 21.2, 7.4 Hz, 1H), 2.10 (d, J = 1.3 Hz, 3H), 1.92 – 1.69 (m, 5H), 1.47 – 1.31 (m, 2H), 1.32 – 1.12 (m, 3H). ¹³C NMR (125 MHz, CDCl₃) δ 185.66, 154.32, 151.39, 144.11, 134.74, 134.60, 131.50, 129.23, 129.11, 118.61, 36.57, 32.29, 26.37, 26.01, 16.77.

3.1. Summary of Screening Results

The flowchart on the right summarizes the screening results (Figure 3). Following primary HTS of approximately 330,000 Molecular Libraries Small Molecules Repository (MLSMR) compounds at 20 μM against the APJ receptor (AID2520), 347 initial actives (~0.1% hit rate) were obtained at a 30% activity cut-off which corresponded to a Z-score of 6.0. For a cell-based assay, the screen was robust with an average Z′ of 0.56.

Figure 3

Flowchart Summary of Screening Results.

Fresh stock solution “cherry picks” of the 347 initial hits were requested from the MLSMR and 311 were received (89.6%) and tested (AID2764) and activities (EC₅₀) were confirmed in full 10-point dose-response titrations (AID488748). In parallel, the selectivity against the angiotensin receptor 1 (ATR1) was also obtained (AID488865).

Only one tractable and reasonable compound/scaffold (quinone-oxime sulfonate) was obtained, that appeared selective against the AT1 receptor. However the potency (11.8 μM EC₅₀ from liquid reorder) did not meet the desired 5 μM potency, and the level of agonism of the most potent compounds were in the 30 – 60% efficacy range. Interestingly, this compound was listed in PubChem with undefined stereochemistry. Upon reorder of the dry powder, the compound was found to be a single isomer CID5601088 (MLS-0437377), although reordered CID1909670 (MLS-0304385) had the same spectral properties and biological activity.

Over 40 additional analogs were available for purchase, which helped to guide the SAR and synthesis plan as detailed in the SAR sections below.

3.2. Dose Response Curves for Probe

Figure 4Dose response curves for ML233

3.3. Scaffold/Moiety Chemical Liabilities

Unlike the related quinone imine sulfonate scaffold, which is generally reactive and appears as a frequent hitter in bioassays, the quinone oxime sulfonate is stable and does not appear as a promiscuous compound from a Scifinder® search performed March 22, 2011.

To address concerns of potential hydrolytic stability or reactivity of ML233, an aliquot of the compound was prepared as a solution in 50% aqueous acetonitrile and was analyzed by LC/MS. A comparison at time zero, six, fifteen and thirty hours indicates the oxime and sulfonamide moiety is stable in aqueous solution (see Supplementary Information). In addition, an aliquot of probe ML233 was treated with a solution of glutathione to evaluate potential reactivity as an electrophile. The results indicate ML233 to be unreactive to glutathione over a seven-hour period (see Supplementary Information).

3.4. SAR Tables

The general SAR strategy we pursued around the quinone oxime sulfonate scaffold is depicted in Figure 5. The structure represented by MLS-0304385 (as noted above, upon reorder of the dry powder, the compound was found to be a single isomer CID5601088 (MLS-0437377)) emerged as the only scaffold from the HTS to demonstrate selectivity over the AT-1 receptor, which was a key element of the desired probe criteria. Once hydrolytic stability and lack of reactivity towards GSH were confirmed (see section 3.3), analog synthesis was initiated to develop SAR with a primary focus on aliphatic substitutions on the core quinone moiety (in green) and a range of substituted aryl sulfonates (in blue). The results are summarized in the table below (Table 4).

Figure 5

General SAR strategy around quinone-oxime sulfonate scaffold.

Table 4

SAR Analysis of APJ Agonists: Quinone-oxime sulfonates.

For the core quinone portion it was found that the optimal substitution consisted of a single larger aliphatic (for example isopropyl, t-butyl or cyclohexyl) flanking the oxygen (R₁) combined with a smaller aliphatic group (such as methyl) flanking the imine (R₃). For example, one of the most potent full agonists was achieved with R₁= cyclohexyl and R₃= methyl (Entry 6, ML233, CID46905036). Other combinations with R₃=methyl that were potent were R₁=isopropyl (Entries 5, 8, 10, 11, 13, 15 28–30) and R₁=t-butyl (Entry 7). Interestingly, the reverse substitution with R₃= isopropyl and R₁= methyl gave less potent (Entry 18) or partial agonists (Entries 16–17).

In general, singly substituted analogs were inactive (Entries 21–27 and 40–42). Other combinations gave compounds that were either less active or showed only partial agonism. For example, with R₁ and R₂= methyl, only partial agonists were obtained (Entries 1–4), as with combinations of R₁ and R₃= methyl (Entries 9 and 12).

A range of substitution was tolerated on the pendant phenyl ring, including chloro, bromo, cyano, methyl and methoxy groups (Entries 32–39). Interestingly, a 3-pyridyl was found to be a suitable replacement for phenyl (Entry 31, compared to entry 11), providing opportunities to further improve the solubility of this series.

3.5. Cellular Activity

All of the primary and selectivity assays are cell based. Specific measures of permeability and toxicity are discussed in section 3.6 below.

3.6. Profiling Assays

The nominated probe was evaluated in a detailed in vitro pharmacology screen as shown in Table 5.

Table 5

Summary of in vitro ADME Properties of APJ Antagonist probe ML233.

ML233 is poorly soluble in aqueous media at pH 5.0/6.2/7.4.

The PAMPA (Parallel Artificial Membrane Permeability Assay) assay is used as an in vitro model of passive, transcellular permeability. An artificial membrane immobilized on a filter is placed between a donor and acceptor compartment. At the start of the test, drug is introduced in the donor compartment. Following the permeation period, the concentration of drug in the donor and acceptor compartments is measured using UV spectroscopy. Consistent with the predicted LogP (see Table 3), ML233 is a highly permeable compound in this assay. This data suggests that, like many therapeutic molecules, the poor aqueous solubility of ML233 will have little impact on its activity in cells and tissues.

Plasma Protein Binding is a measure of a drug’s efficiency to bind to the proteins within blood plasma. The less bound a drug is, the more efficiently it can traverse cell membranes or diffuse. Highly plasma protein bound drugs are confined to the vascular space, thereby having a relatively low volume of distribution. In contrast, drugs that remain largely unbound in plasma are generally available for distribution to other organs and tissues. ML233 is highly bound to plasma proteins in mouse plasma (99% bound). Similarly, although to a lesser extent, ML233 is highly protein bound in human plasma (98% bound). This interaction may confer stability to the molecule in plasma, protecting it from metabolizing enzymes.

Plasma Stability is a measure of the stability of small molecules and peptides in plasma and is an important parameter, which can strongly influence the in vivo efficacy of a test compound. Drug candidates are exposed in plasma to enzymatic processes (proteinases, esterases), and they can undergo intramolecular rearrangement or bind irreversibly (covalently) to proteins. Although ML233 is extensively bound to plasma proteins, the compound exhibits very poor stability in human plasma (18.68% remaining) after 3 hr. In contrast, ML233 is very stable in mouse plasma (100% remaining). This may be explained by the relative activities of metabolizing enzymes in the differing plasmas.

Alternatively, these disparate results may be explained by the subtle differences in plasma protein binding in mouse versus human plasma. Although the difference in percent compound bound in mouse versus human plasma appears minor (99% vs. 98%), the plasma protein-binding assay is not performed in a way that is suitable for assessing the kinetics of ML233 binding to plasma proteins. Therefore, we cannot exclude the possibility that as ML233 dissociates from human plasma proteins, it is rapidly metabolized.

The microsomal stability assay is commonly used to rank compounds according to their metabolic stability. This assay addresses the pharmacologic question of how long the parent compound will remain circulating in plasma within the body. ML233 shows poor stability (<1.0% remaining after 60 minutes) in both human and mouse liver homogenates. Ultimately this limits the utility of this probe to in vitro studies or apelin receptor or in vivo studies using acute intravenous doses to avoid hepatic metabolism.

ML233 shows some toxicity (LC₅₀ = 25.8 μM) toward human hepatocyctes.

Profiling against other GPCRs. The probe, ML233 (CID46905036), was submitted to the Psychoactive Drug Screening Program (PDSP) at the University of North Carolina (PDSP, Bryan Roth, PI) and the data against a GPCR binding assay panel is shown in Figure 6. ML233 shows potentially significant binding against several other receptors, including the 5-HT_1A, α_2C adrenergic, and benzylpiperazine receptors (55%I, 51%I and 65%I at 10 μM, respectively) and norepinephrine transporter (57%I at 10 μM). It is not known whether these activities in binding assays are translated into functional modification of the activities of these receptors. In addition, these potential cross-reactivities should not confound any in vitro data obtained with ML233.

Figure 6

GPCR profiling panel for ML233.

As a follow up experiment, ML233 was submitted for further testing by DiscoveRx (as a CRO) in a panel of functional assays of selected GPCRs (see Table 6 below). Those GPCRs tested were chosen based on their logical association with apelin/APJ due to their effects on cardiovascular function. Table 6 shows that ML233 has little to no agonist activity at 14 related receptors when assayed at 10μM. Those receptors tested include the adrenergic receptor family, the endothelin and bradykinin receptor families, as well as the vasopressin receptors. In this panel of GPCRs, the AT1 receptor was also profiled (AGTR1, line 7). The data returned from this externally executed panel is shown below in Table 6.

Table 6

ML233 Functional profiling data against the DiscoveRx GPCR Panel tested at a single 10 μM concentration.

4.1. Comparison to existing art and how the new probe is an improvement

There has been only one report of a “small molecule” APJ agonists in the open or patent literature (12), but, as discussed in the introduction, this is a peptide derived compound with mw>1000. Thus, ML233 represents the first true small molecule (mw= 359) APJ agonist that will be available to the scientific community.

4.2. Mechanism of Action Studies

Dr. Smith’s laboratory has performed a series of exploratory studies to investigate the mechanism of action of this probe. The mechanism of action of ML233 is that of a classic full agonist of APJ in the cell-based assays employed in this project. Because APJ is capable of both G-protein-dependent and β-arrestin-dependent signaling, and the primary assay utilized in the project measured the activity of ML233 as a function of β-arrestin recruitment to APJ, we assessed the activity of ML233 in an assay of G-protein signaling by APJ. This MOA assay uses the High Sensitivity Lance TR-FRET cAMP assay (Perkin Elmer). Consistent with its role as an APJ agonist, ML233 reduced forskolin stimulated increases in intracellular cAMP (Figure 7A). Interestingly, the observed potency of ML233 to reduce cAMP was significantly higher than the maximal concentration studied (100μM). In addition, ML233 induced APJ internalization (Figure 7B) taken together these probe characterization studies suggest that, like most GPCRs, the apelin receptor is capable of pleiotropic signaling, and that ML233 is a useful tool probe β-arrestin-mediated signaling and receptor internalization of APJ. Those investigators specifically interested in these aspects of APJ pharmacology and function will be well served by using ML233. However, the utility of ML233 in probing G-protein signaling of APJ is yet to be fully explored. Investigators using ML233 for this purpose should further explore the effects of ML233 in their assay of choice.

Figure 7

ML233 reduces forskolin stimulated intracellular cAMP and stimulates APJ internalization. (A) CHO-K1 cells heterologously expressing the human apelin receptor APJ were exposed to forskolin (15 μM) and a range of concentration of ML233. ML233 weakly (more...)

4.3. Planned Future Studies

As described above, studies to support the proposed mechanism of action are underway in Dr. Smith’s laboratory using the probe molecule described in this report. Future studies also planned will involve testing ML233 in genetically tractable model organisms including zebra fish and mice. Additional medicinal chemistry efforts to improve the drug-like properties of ML233 are under way. Specific efforts will be directed at improving the metabolic stability and especially solubility. Improvements in these properties will improve the in vivo utility of ML233 in vivo.

Future in vitro studies to be performed by Dr. Smith will determine if ML233 is a direct competitor of apelin binding using standard competition binding studies and radiolabeled apelin-13. Consistent with our Chemical Probe Development Plan, we will also evaluate the affinity of ML233 to the angiotensin II type 1 receptor, the receptor most closely related to APJ. Additional studies will utilize existing cell based assays to determine if ML233 is an allosteric modulator of APJ

6.1. Assay Details: APJ Beta-Arrestin 1536-Well Agonist Assay Protocol

6.2. Assay Details: AT1 Beta-Arrestin 1536-Well Agonist Assay Protocol

6.3. Supplementary Information for Probe Characterization

¹H & ¹³C NMR Spectra and LC/MS Data for Structural Determination of ML233

LC/MS Data for Structural Determination of ML233

Stability study of ML233 in 50% aqueous acetonitrile

LC/MS of ML233 with glutathione after 0, 30 min, 2 and 7 hours