Materials and Reagents

All reagents and solvents were purchased from commercial vendors and used as received.

Phosphate-buffered saline (PBS; catalog no. 08J07A0022) was acquired from the Broad Institute Supply and Quality Management (SQM; Cambridge, MA). CellTiter Glo (catalog no. G7573 lot 268563) was purchased from Promega (Fitchburg, WI), and Alamar Blue (catalog no. DAL1025) was acquired from Biosource/Invitrogen (Grand Island, NY). White, sterile, TC-treated, 384-well plates (catalog no. 3570) were acquired from Corning.

2.1. Assays

A summary listing of completed assays and corresponding PubChem AID numbers is provided in Appendix A (Table A1). Refer to Appendix B for the detailed assay protocols.

2.2. Probe Chemical Characterization

The probe molecule (ML210, SID 103023280, BRD-K01877528-001-01-6) was synthesized in six steps from bis(4-chlorophenyl)methanone, 5-methylisoxazole-3-carboxylic acid and piperazine (Scheme 1).

Scheme 1

Synthesis of the probe.

General details. Nitration of the 5-methylisoxazole-3-carboxylic acid building block was accomplished using concentrated sulfuric acid and potassium nitrate in 70% yield. The carboxylic acid 1 was then converted to the corresponding acid chloride, 5-methyl-4-nitroisoxazole-3-carbonyl chloride 2 in quantitative yield. Bis(4-chlorophenyl)methanone was reduced using sodium borohydride to the corresponding alcohol 3 in 94% yield. Chlorination using oxalyl chloride was followed by treatment with an excess of piperazine in refluxing acetonitrile and afforded 1-(bis(4-chlorophenyl)methyl)piperazine 5 in 68% yield. Coupling of secondary amine 5 with the acid chloride 2 in dichloromethane afforded the probe compound 6 in 74% yield. Full experimental details and characterization are provided below.

The solubility of the probe was measured in water and in PBS at room temperature and was found to be below 1 μM in both cases. The stability of the probe (ML210) in PBS (0.1% DMSO) was measured over 48 hours, and the data is shown in Figure 5 (blue line). We suspected that poor solubility (and not instability) as the reason behind the dramatic drop in the amount of sample over time. To test this, we added acetonitrile to each well (final concentration 50%) and measured the amount of the probe. Amounts detected after addition of acetonitrile are shown in Figure 5 (red line). From these results it can be concluded that the probe is stable in PBS and 100% of the probe is still present after 48 hours of incubation in PBS.

Figure 5

PBS Stability Data for the Probe (ML210). Total ion count of the probe (ML210) over time (blue line). Total ion count of the probe upon addition of acetonitrile (red line).

The plasma protein binding was found to be 100% both in human and mouse. Plasma stability was 63.1% in human and 65.4% in mouse after a 5-hour incubation period. The probe and five analogs were submitted to the SMR collection [MLS003265661 (probe), MLS003265665, MLS003265666, MLS003265663, MLS003265664, MLS003265662].

2.3. Probe Preparation

General details. All reagents and solvents were purchased from commercial vendors and used as received. ¹H and ¹³C NMR spectra were recorded on a Bruker 300 MHz or Varian UNITY INOVA 500 MHz spectrometer as indicated. Proton and carbon chemical shifts are reported in ppm (δ) relative to tetramethylsilane (δ = 0 for both ¹H and ¹³C) or CDCl₃ solvent (¹H δ 7.26, ¹³C δ 77.0). NMR data are reported as follows: chemical shifts, multiplicity (obs. = obscured, br = broad, s = singlet, d = doublet, t = triplet, m = multiplet); coupling constant(s) in Hz; integration. Unless otherwise indicated NMR data were collected at 25 °C. Flash chromatography was performed using 40–60 μm Silica Gel (60 Å mesh) on a Teledyne Isco Combiflash Rf system. Tandem Liquid Chromatography/Mass Spectrometry (LC/MS) was performed on a Waters 2795 separations module and 3100 mass detector. Analytical thin layer chromatography (TLC) was performed on EM Reagent 0.25 mm silica gel 60-F plates. Visualization was accomplished with ultraviolet (UV) light and aqueous potassium permanganate (KMnO₄) stain followed by heating.

5-methyl-4-nitroisoxazole-3-carboxylic acid (1): 5-methylisoxazole-3-carboxylic acid (1.5 g, 12.04 mmol) was added to a mixture of potassium nitrate (1.83 g, 18.06 mmol) and sulfuric acid (5 ml) at room temperature. After complete dissolution, the mixture was warmed to 50°C and stirred for 4 hours. The mixture was then cooled to 0°C, ice was added, and the solution was neutralized with sodium bicarbonate. The mixture was extracted with ethyl acetate (3 × 30 ml), dried over sodium sulfate, filtered and concentrated to give 1.45 g of a white solid (8.43 mmol, 70%). The product could be further recrystallized from dichloromethane. ¹H NMR (300 MHz, DMSO-d6): δ 6.72 (s, br, 1H), 2.28 (s, 3H); ¹³C NMR (75 MHz, DMSO-d6): δ 187.4, 124.2, 117.7, 107.3, 29.5.

5-methyl-4-nitroisoxazole-3-carbonyl chloride (2): 5-methyl-4-nitroisoxazole-3-carboxylic acid (200 mg, 1.16 mmol) was dissolved in dichloromethane (4 ml) under argon atmosphere. Oxalyl chloride (202 μL, 2.32 mmol) was added dropwise at room temperature, followed by addition of one drop of DMF. After stirring overnight, the mixture was concentrated and co-evaporated with chloroform to afford 220 mg of a slightly yellow oil (1.15 mmol, 99%), which was carried on to the next step without further purification.

Bis(4-chlorophenyl)methanol (3): Bis(4-chlorophenyl)methanone (3.0 g, 11.95 mmol) was dissolved in methanol/tetrahydrofuran 1:1 (100 ml) and cooled to 0°C. Sodium borohydride (452 mg, 11.95 mmol) was added in one portion, and the mixture was stirred for 30 minutes. After TLC showed complete conversion, the mixture was neutralized with acetic acid and concentrated. Dichloromethane was added, and the solution was washed with water (2 × 30 ml), dried on sodium sulfate and concentrated to afford 2.83 g of an off-white solid (11.18 mmol, 94%), which was carried to the next step without further purification. ¹H NMR (300 MHz, CDCl₃): δ 7.32 (d, J = 8.47 Hz, 4H), 7.28 (d, J = 7.28 Hz, 4H), 5.78 (s, 1H), 2.27 (s, 1H); LRMS (M + HCO₂)⁻: 298.95.

4,4′-(Chloromethylene)bis(chlorobenzene) (4): Bis(4-chlorophenyl)methanol (2.83 g, 11.18 mmol) was dissolved in dichloromethane (30 ml) at room temperature. Oxalyl chloride (975 μL, 11.18 mmol) was added followed by addition of a drop of dimethylformamide. After stirring overnight, the mixture was concentrated and co-evaporated with chloroform to afford 3.0 g of an off-white solid (11.05 mmol, 99%), which was carried to the next step without further purification. ¹H NMR (300 MHz, CDCl₃): δ 7.32 (m, 8H), 6.06 (s, 1H).

1-(bis(4-chlorophenyl)methyl)piperazine (5): 4,4′-(Chloromethylene)bis(chlorobenzene) (3.0 g, 11.05 mmol) was dissolved in acetonitrile (100 ml) at room temperature. Piperazine (8.6 g, 100 mmol) was added, and the mixture was refluxed overnight. Acetonitrile was evaporated. The mixture was dissolved in ethyl acetate and washed three times with water to remove the excess piperazine. The organic phase was dried over sodium sulfate, filtered, and concentrated. The crude mixture was purified on silica gel using a gradient of (7N NH₃ in methanol) in dichloromethane to afford 2.4 g of an off-white solid (7.51 mmol, 68%). ¹H NMR (300 MHz, CDCl₃): δ 7.32 (d, J = 8.50 Hz, 4H), 7.25 (d, J = 8.56 Hz, 4H), 4.18 (s, 1H), 2.87 (t, J = 4.79 Hz, 4H), 2.32 (br s, 4H), 1.55 (br s, 1H); LRMS (M + H)⁺: 322.04.

(4-(bis(4-chlorophenyl)methyl)piperazin-1-yl)(5-methyl-4-nitroisoxazol-3-yl)methanone (6): 5-methyl-4-nitroisoxazole-3-carbonyl chloride (33 mg, 0.17 mmol) was dissolved in dichloromethane (1 ml) at room temperature and 1-(bis(4-chlorophenyl)methyl)piperazine (55 mg, 0.17 mmol) was added, along with one drop of triethylamine. After stirring for 2 hours, TLC shows complete conversion. The mixture was dissolved in dichloromethane and washed with a saturated aqueous solution of sodium bicarbonate (2 × 10 ml). The organic phase was dried over sodium sulfate, filtered, and concentrated. The crude mixture was purified on silica gel using a gradient of ethyl acetate in hexanes to afford the final product (61 mg) as a white solid (0.128 mmol, 74%). ¹H NMR (500 MHz, CDCl₃): δ 7.32 (d, J = 8.58 Hz, 4H), 7.26 (d, J = 8.56 Hz, 4H), 4.25 (s, 1H), 3.84 (t, J = 4.96 Hz, 2H), 3.36 (t, J = 4.99 Hz, 2H), 2.85 (s, 3H), 2.52 (t, J = 5.08 Hz, 2H), 2.37 (t, J = 4.97 Hz, 2H); ¹³C NMR (125 MHz, CDCl₃): δ 171.7, 156.4, 153.0, 139.9, 133.2, 129.0, 128.9, 94.7, 74.2, 51.5, 50.9, 46.9, 42.2, 13.4; HRMS (ESI): calculated mass for C₂₂H₂₀Cl₂N₄O₄ [M+H] 475.0934, found 475.0953.

The ¹H NMR and ¹³C spectra and UPLC chromatograms of the probe (ML210) and analogs are provided in Appendix C.

Probe Attributes

• Confirmed activity in the RAS-dependent strains <2.5 μM.
• At least 4-fold weaker IC₅₀ activity in the non-RAS strains.

The project included a primary high throughput screen of the entire MLSMR collection (greater than 300,000 substances). Approximately 0.4% of the compounds tested were selected for dose retest and selectivity. Of these, 14 compounds from the same chemical class were identified as having the desired selectivity. In addition to these commercially available compounds, multiple rounds of chemistry were performed to determine the SAR of several substituents on the scaffold, and the compound with the best combination of potency and selectivity was designated as the probe (ML210).

3.1. Summary of Screening Results

A high throughput screen of 303,344 substances (303,282 unique compounds, AID 1832) was performed in duplicate in a 7.5-μL reaction in 384-well plates seeded with 1000 cells per well in DMEM/FBS medium. Cells were grown overnight, then treated with compound at a final concentration of 7.5 μM for 48 hours, after which cell viability was measured through determination of ATP levels using Promega Cell-TiterGlo reagent. Compounds causing at least 50% reduction in ATP levels relative to DMSO-treated cells were considered active. This resulted in 516 active compounds, which were retested at dose along with several analogs of active families for a total of 1155 compounds (AID 1936). These 1155 compounds were counterscreened against BJeH-LT (AID 1935) and BJ-eH (AID 1933) and also confirmed in a secondary assay for mechanism of action in DRD cells (AID 1934.) The compounds that passed these four screens were analyzed by the Assay Provider using a related viability assay in which Alamar Blue was metabolized to a fluorescent product by living cells; however, the same selectivity profiles were not observed in these secondary assays, either due to differences in maintenance of cell lines or due to the different detection methods.

As a result of this discrepancy, the available most active compounds from the primary screen (435 compounds) were tested in BJeLR (AID 2610) and BJeH-LT (AID 2631) by the Assay Provider to determine selectivity. Of these, 73 of 435 compounds were then tested in the two additional cell lines: DRD (AID 2633) and BJeH (AID 2635). Of these compounds, 26 of 73 displayed the desired potency and selectivity in all four assays and were designated as probe candidates.

Two of these compounds were in the same chemical class and were commercially available, although only one confirmed in the four cell lines (AIDs 2607, 2608, 2609, 2611; Table 1 and Figure 6) This core was used for further SAR studies to improve the potency and selectivity. A novel compound was identified with selectivity similar to the initial hit but with an approximate 10-fold improved potency. This compound was selected as the probe ML210 ((4-(bis(4-chlorophenyl)methyl)piperazin-1-yl)(5-methyl-4-nitroisoxazol-3-yl)methanone).

Figure 6

Critical Path for Probe Development.

3.2. Dose Response Curves for Probe

The inhibition of cell viability curves for the probe (ML210) in four different cell lines are displayed in Figure 7.

Figure 7

Concentration-dependent Activities by the Probe (ML210) in HRAS^V12 Expressing and Non-expressing Cell Lines. Activity curves for probe (ML210): BJeLR (AID 493053) (a); DRD (AID 493050)(b); BJeH-LT (AID 493093)(c);BJeH (AID 493051)(d).

3.3. Scaffold/Moiety Chemical Liabilities

A search of PubChem for the hit compound (CID 15945539, Figure 8) shows that it has been screened in 356 different assays and was identified as active in only two other assays, AID 485294: qHTS inhibitors of AmpC Beta-Lactamase (confirmatory) and AID 485297: qHTS Assay for Rab9 Promoter Activators (confirmatory). This shows that the probe molecule is very unlikely to be promiscuous, considering that the only difference between the hit compound and the probe molecule is the presence of a para-chloro substituent on both phenyl rings (Figure 8).

Figure 8

Hit Compound and Probe Molecule Structures.

The benzhydrylpiperazine moiety is present in cetirizine (Zyrtec®, Pfizer, Figure 9), a second generation antihistamine used in the treatment of allergies. Isoxazoles are also present in drugs, such as in valdecoxib (Bextra, G. D. Searle & Company), a cyclooxygenase-2 selective inhibitor that was used as a nonsteroidal anti-inflammatory drug until 2005, and risperidone (Risperdal®, Johnson & Johnson, Figure 9), an atypical antipsychotic that is still one of the best-selling drugs for the treatment of schizoaffective disorders and bipolar disorders. Nitro groups are usually undesirable in drug discovery because they can be metabolized to toxic reactive intermediates such as nitroso or hydroxylamines. However, when present on an electron-rich, five-membered heterocyclic ring, they are often more tolerable. Several nitro-aromatic-containing drugs have been developed. For example, nifedipine (Procardia®, Bayer) is a calcium channel blocker used as an anti-anginal and antihypertensive agent, and clonazepam (Klonopin®, Roche, Figure 9) is used as an anticonvulsant, muscle relaxant, and anxiolytic.

Figure 9

Drugs Containing Benzhydrilpiperazine, Isoxazole and Nitro Groups.

3.4. SAR Tables

Even though the nitro group is present in several FDA-approved drugs, it can be a liability in many instances. Thus, we first investigated the possibility of replacing the nitro group with other functional groups (Table 1).

Replacement of the nitro group by a primary amine, acetamide, sulfonamide, bromine, methyl, and hydrogen (entries 2–7, respectively, Table 1) led to completely inactive compounds, showing the importance of the nitro group for activity. We also tried to substitute the nitro group by a cyano group. Although cyanation of a variety of aromatic and heteroaromatic halides is known (25), there are no reports of a direct conversion of 4-haloisoxazoles to the corresponding 4-cyanoisoxazoles. We investigated the possibility to synthesize the corresponding nitrile analog starting from a bromoisoxazole model compound using a wide variety of reported cyanation conditions (Scheme 2, Table 2).

Scheme 2

Cyanation of a Bromoisoxazole Model Compound.

Table 2

Cyanation of a Bromoisoxazole Model Compound.

Rosenmund-von Braun conditions (entry 1, Table 2) using an excess of copper (I) cyanide in refluxing N-methylpyrrolidone led to decomposition of the starting material. We then turned to palladium catalyzed cyanation reactions (entries 2–5, Table 2), using copper (I) cyanide or zinc (II) cyanide and different phosphine ligands; but, in all cases, we either observed decomposition of the starting material or no reaction.

We also performed the cyanation reactions directly on the full probe that includes the substituted piperazine moiety (Scheme 3), and these results are summarized in Table 3.

Scheme 3

Cyanation of the Bromoisoxazole Analog.

Table 3

Cyanation of the Bromoisoxazole Analog of the Probe (CID 3689413/ML210).

Unfortunately, no desired product could be detected in these experiments. The only product we could observe was the reductive opening of the isoxazole ring.

Next, we investigated the influence of the nitroisoxazole ring on the activity of the compound (Table 4). Replacing the methyl substituent on the 3-position of the isoxazole by an isopropyl group (entry 2, Table 4) led to an inactive compound. Removing the nitro group at the 4-position led to a complete loss of activity, irrespective of the substituents on the 3-position of the isoxazole ring (entries 3–6, Table 4). Benzoisoxazole (entry 7, Table 4) and 2,4,5-trimethylfuran (entry 8, Table 4) analogs were also found to be inactive, further confirming the necessity of the nitro substituent for activity. Replacing the nitroisoxazole ring with other nitroaromatics such as 4-methyl-2-nitrophenyl (entry 9, Table 4) or 4-nitropyrazole (entry 10, Table 4) led to completely inactive compounds. Hence, the presence of a nitro group is necessary for activity but not sufficient. Thus, the nitroisoxazole moiety of the molecule is important for activity and was conserved intact for further SAR studies. Interestingly, replacing the nitroisoxazole ring with a 2-nitrofuran ring (entry 11, Table 4) led to a reverse selectivity, and the compound was selectively toxic to the HRAS wild-type cell line BJeH-LT versus the HRAS^V12 cell line BJeLR.

Table 4

Summary of SAR on the Nitroisoxazole Ring.

To improve the solubility of the probe, we then investigated the influence of the benzhydryl-piperazine portion of the molecule (Table 5). Introduction of a solubilizing agent such as morpholine (entry 1, Table 5) improved solubility up to >500 μM, but led to an inactive compound. A shorter amide like the p-fluorobenzyl analog (entry 2, Table 5) was also inactive.

Table 5

Summary of SAR on the Benzhydryl-Piperazine Portion.

Keeping the piperazine ring in place and only replacing the benzhydryl portion with an ethylcarbamate (entry 3, Table 5) or a o-methoxyphenyl ring (entry 4, Table 5) also led to inactive compounds. Quite surprisingly, even removing one of the two phenyl rings (entry 5, Table 5) led to a completely inactive compound, showing that the benzhydryl portion is necessary for activity.

Realizing that both phenyl groups are required for activity, we next investigated the influence of the piperazine linker (Table 6). Replacing the piperazine linker with homopiperazine (entry 2, Table 6) led to a small decrease in activity, showing that homopiperazine is tolerated at that position. The ethylene diamine analog (entry 3, Table 6) was almost completely inactive, showing that some rigidity as well as the absence of hydrogen bond donors is preferred for activity.

Table 6

Summary of SAR on the Piperazine Ring.

To improve activity and solubility of the probe, the substitution on both phenyl rings of the benzhydryl group was investigated (Table 7). Replacement of both phenyl group by 2-pyridyl (entry 2, Table 7) or 4-pyridyl groups (entry 3, Table 7) dramatically increased the PBS solubility up to 500 μM and reduced plasma protein binding down to 40–60%. Unfortunately, both of these compounds were found to be inactive.

Table 7

Summary of SAR on the Benzhydryl Portion of the Molecule.

Introducing p-fluoro substituents on both phenyl rings (entry 4, Table 7) led to a 4-fold increase in activity. It had no effect on solubility and plasma protein binding but slightly improved the plasma stability. Introduction of p-methoxy substituents on both phenyl groups (entry 5, Table 7) led to a decrease in activity although it improved plasma stability. The introduction of p-chloro substituents on both phenyl rings (entry 6, Table 7) led to a 10-fold increase in activity at the cost of a slight decrease in selectivity. However, this analog was chosen as the probe. Taking clues from the structure of the drug Cetirizine (see Figure 9), we made an analog that contains an unsymmetrical benzhydryl unit where one phenyl ring is unsubstituted while the other phenyl ring has a 4-chloro substituent (entry 7, Table 7). Unfortunately, this modification led to a decrease in activity.

Overall, studying the effect of structural changes of the benzhydryl portion of the molecule on activity turned out to be productive, and we identified a synthetic compound (entry 6, Table 7) with the 4-chlorophenyl substituent as the probe. The probe compound was also found to have an IC₅₀ of 179 ± 72 nM in the DRD cell line, which represents a 15-fold increase in activity compared to the hit compound. It also shows an IC₅₀ of 21.4 ± 20.8 μM in the BJeH cell line.

During the course of our SAR investigation, we found that a 2-nitrofuran analog had a selective toxicity for the HRAS wild-type cell line BJeH-LT. Thus, we synthesized and screened several 2-nitrofuran analogs having different linker nature and different substitution pattern on the benzhydryl portions (Table 8). The selective toxicity toward the BJeH-LT cell line versus the BJeLR cell line was conserved when both phenyl group of the benzhydryl portion were substituted with a p-fluoro (entry 1, Table 8) or a p-methoxy group (entry 2, Table 8). The same tendency was observed when the piperazine linker was replaced by a homopiperazine linker (entry 3, Table 8) or an ethylene diamine linker (entry 4, Table 8). When the amide portion was reduced to an amine (entry 5, Table 8), the compound was found to be inactive in both cell lines. This switch in the selectivity profile for nitrofuran analogs versus nitroisoxazole analogs is somewhat interesting and probably implies a different mode of action.

Table 8

Summary of SAR on Several Nitrofuran Analogs.

Figure 10 presents a visual, high-level summary of the SAR performed on the nitroisoxazole scaffold. All positions of the hit compounds were investigated: the nitroisoxazole portion, which proved to be required for activity; the linker nature, where piperazine was found to be preferred and homopiperazine tolerated; and the benzhydryl portion, where both phenyl groups are necessary for activity and p-chloro substitution is preferred.

Figure 10

Summary of SAR performed on the Hit Compound (24S/11P).

3.5. Cellular Activity

All assays used in this project were cellular cytotoxicity assays, with the probe selected on the basis of selective cellular growth inhibition. Therefore, no additional testing of cell toxicity or cell permeability was necessary.

3.6. Profiling Assays

The probe compound (ML210) is a novel compound and has not been reported previously in PubChem. As described above, a search of PubChem for the initial hit compound (CID 15945539), which has similar selectivity to the final probe, shows that it has been screened in 356 different assays and was identified as active in only two other assays (AID 485294 and AID 485297). Given the similarity of the probe to the initial hit, it is, therefore, unlikely that the probe is a promiscuous compound. We plan to perform further off-target testing through a standard commercial panel (CEREP/Ricerca).

The compound will be submitted to further cancer cell-line profiling. A previously reported HRAS selective probe developed as part of this project, ML162, showed significant selectivity when tested in the NCI60 cell-line panel and in cancer cell-lines as part of another initiative underway at the Broad Institute (26). This new probe, ML210, has similarly been added to the compound set for profiling across hundreds of cancer cell-lines.

To date, ML210 has been tested in a small subset (approximately 20) of these cell lines to determine its’ potency and selectivity. It displays differential toxicity comparable to probe ML162, with an IC₅₀ difference of over two log units between certain cancer lineages and tumor types. Further analysis will determine how similar the pattern of selectivity of ML210is to ML162 and to other known compounds and will provide insight into the potential uniqueness of the molecular mechanism of this probe.

4.1. Comparison to Existing Art and How the New Probe is an Improvement

Probe ML210 is an improvement over the existing art. The two compounds previously identified using the same screening methods are: erastin and RSL-3. Erastin effectively inhibits the engineered HRAS^V12-expressing cell lines with a low micromolar IC₅₀ and 4-fold selectivity over isogenic non-HRAS^V12-expressing cells. A molecular target, voltage-dependent anion channel (VDAC) has also been determined for erastin. RSL-3 is reported to be more potent (IC₅₀ ~100 nM) and somewhat more selective, although the molecular target is not known. In our hands, erastin was found to have an IC₅₀ of 1.1 μM in BJeLR and 3.2 μM in BJeHLT, which gives a 2.9-fold selectivity for BJeLR. RSL-3 was found to have an IC₅₀ of 0.75 μM in BJeLR and 1.5 μM in BJeHLT, which gives a 2-fold selectivity for BJeLR. Probe ML210 is five times more potent than erastin and 3.7 times more potent than RSL-3 in BJeLR and its selectivity (2.6-fold) is superior to RSL-3 and comparable to erastin. Probe ML210 displays an improved potency over one existing probe, erastin, while maintaining similar selectivity. It is also a novel chemotype compared to existing probe RSL-3 and our other probe ML162, lacking the potential chemical liability of an electrophilic alpha-chloro amide. Although day to day variability in the response of engineered cell lines was observed due to some instability of the constructs, ML210 consistently displayed equal or better potency and selectivity than the prior art probes erastin and RSL3 when all molecules were run in parallel experiments.

Furthermore, ML210 appears to cause a phenotype similar to the other probe arising from this screening campaign, ML162, while belonging to an entirely different structural class. Preliminary profiling experiments in the NCI60 cell line panel show that while ML210 is less potent than ML162, as expected from the observed screening results, it shows a similar pattern of activity in the NCI60 cell lines, with the inter-replicate variability of ML162 (Figure 11A) being similar to the inter-compound variability of ML162 and ML210 (Figure 11B and 11C). The existence of two probes of completely different structure that have similar patterns of biological activity is very useful for identifying potential targets for further study and will advance the field of cancer and specifically Ras biology.

Figure 11

ML210 vs ML162 activity in the NCI60 cell lines. Log GI₅₀ values of ML210 against the NCI60 panel of cell lines. Replicate comparison of ML162 (panel A). ML210 compared to two replicate tests of ML162 (panel B and C). GI₅₀s weaker than 100 μM (more...)

Chemically, probe ML210 is in a novel structural class not previously reported in the field of RAS selective compounds. Further profiling activity in nonengineered cancer cells will demonstrate the importance of multiple probes with different activity patterns and may lead to identification of novel pathways important for small-molecule susceptibility of cancers.

Investigation into relevant prior art entailed searching the following databases: SciFinder, Reaxys, PubChem, PubMed, US Patent and Trademark Office (USPTO), PatFT, AppFT, and World Intellectual Property Organization (WIPO). The search terms applied and hit statistics are provided in Table 9. Abstracts were obtained for all references returned and were analyzed for relevance to the current project. The searches were performed on and are current as of January 26, 2011.

Table 9

Search Strings and Databases Employed in the Prior Art Search.

The literature and patent searches summarized in Table 9 uncovered three small molecules that are known to be selectively lethal to cell lines harboring an HRAS mutation (HRAS^V12), namely erastin, RSL-3, and RSL-5 (see Figure 2).

4.2. Mechanism of Action Studies

The putative mechanism of action, based on the design of the primary screen and counterscreens, involves a target related to expression of oncogenic HRAS. As is demonstrated by the prior art, this does not mean the target is HRAS or on the RAS pathway; rather, the direct molecular target may be involved in processes that are merely more sensitive to inhibition when oncogenic HRAS is expressed. In the case of erastin, such a target has been shown to be VDACs.

In regard to the generality of the probe’s activity, we have shown that MEK inhibitor U0126 suppresses the activity of ML210 over 30 fold in two cell lines, BJeLR (figure 12A) and human lung carcinoma Calu-1 (Figure 12B), which carry mutant HRAS and KRAS, respectively. This result indicates that ML210 likely targets common biological characteristics induced by mutations in different isoforms of Ras. Although we have not yet tested whether previously reported Ras synthetic lethal genes are involved in the mechanism of action of ML210, many of these genes were identified in cancer type and Ras isoform specific contexts (9–12), and may not be active generically. Therefore, further genetic suppressor screens using this new probe in multiple cell lines with Ras mutations may discover potential drug targets in common among different Ras isoform mutations.

Figure 12

Activity of ML210 in BJeLR (HRAS mutant) and Calu-1 (KRAS mutant) in the presence and absence of a MEK inhibitor (U0126).

A genetic method to identify a target is the use of RNAi in conjunction with the probe compound. An analysis of the target of probe ML210 may result in a similar activity profile in various cancer cell lines and will synergize with lower doses of the probe. The Broad Institute has created an RNAi platform dedicated to genome-scale experiments for the systematic application of RNA knockdown methods. Both historical data mining and future experiments can be used to identify candidate genes and pathways that are responsible for the selective phenotype of this molecule.

We are also using profile experiments to determine additional correlations with sensitivity to the probe. The compound was recently submitted to the NCI for testing in the NCI60 cell panel. Upon completion of NCI60 panel profiling, correlations will be attempted to the known genetic characteristics of these cell lines. We have also added this probe to a small set of approximately 300 compounds that are being profiled against a larger (1000) panel of cell lines as part of another initiative underway at the Broad Institute (26). This profiling analysis will correlate effects of the probe on cell viability and other cellular markers with gene expression data. Additional hypotheses can then be generated as to the molecular pathway or target of probe ML210.

4.3. Planned Future Studies

Probe ML210 was identified through a series of phenotypic screens. Therefore, one of the key uses of this probe will be the identification of a molecular target or pathway, as described above, which confers selective toxicity in cancer cells with certain genetic expression profiles. This can be done through biochemical and proteomic studies in an attempt to capture and identify the direct binding partner.

Another approach involves additional profiling in a sufficient number of well characterized cell lines to correlate sensitivity to the probe with genetic features. This is the goal of the 1000-cell line profiling already underway using probe ML210.

As a designated probe compound, ML210 will by default be included in several planned studies at the Broad Institute that will characterize the MLSMR probe set. The Broad Institute Center Driven Research Project (CDRP) will profile probes, analogs, and a subset of less characterized compounds from the MLSMR using both multi-parametric, image-based analysis and 1000-plex Luminex gene expression arrays in a well-defined cancer cell line, U2OS. In addition, probe compounds will be further characterized in 1000-plex gene expression arrays in up to 20 additional cell lines. These and other profiling approaches will continue to enhance the understanding of the mechanism of action and applicability of probes such as ML210.

We envision that as this probe continues to be developed, it will become a common tool for testing susceptibility of cancers to specific inhibition and differentiating between genetic features. Additional characterization of probe ML210 through more widespread use will further define its phenotypic properties, continuing to enhance its utility as a research tool for cancer biology.

Primary Screen for BJeLR Cell Viability

1. Maintain cells in 35 ml of growth medium (per 1 Liter: 730 ml DMEM with 4 mM L-glutamine, 210 ml M199, 150 ml heat-inactivated fetal bovine serum (FBS) in a T175 cell culture flask.
2. Incubate in a TC incubator at 95% humidity, 5% CO₂, 37°C.
3. To passage, harvest cells by first aspirating the media and rinsing the flask with 10 ml sterile PBS.
4. Aspirate PBS and add 5 ml trypsin to the flask. Incubate for 5 minutes at 22°C.
5. Add 8 ml of growth medium to the trypsin to quench the reaction.
6. Resuspend the cells, count, and transfer 4.5 million cells in approximately 2 ml to 33 ml fresh growth medium in a new T175 flask. Carry the lines for no more than 20 passages.
7. For screening, harvest the cells, and adjust the concentration to 33,000/ml. While gently stirring, disperse the cells with a Combi multidrop by adding 30 μL of suspension per well to white, sterile, TC-treated, 384-well plates for a total of 1000 cells per well. Incubate the plates overnight in an automated TC incubator at 95% humidity, 5% CO₂, 37°C.
8. For compound screening, add 50 nL or 100 nL of compound, depending on the desired final compound concentration, using slotted steel pins on a pin tool.
9. Return the plates to the incubator for 48 hours. To read viability, remove the cells from the incubator and cool to room temperature for 30 minutes. Remove the lids, and add 30 μL of diluted CellTiter Glo (1:3 dilution with PBS) to each with a Combi Multidrop. Incubate the plates for 10 minutes, and detect luminescence on an Envision (Perkin-Elmer) multimode reader (0.1 seconds per well).

Secondary Screen for BJeLR Cell Viability (Alamar Blue)

1. Dilute the compounds into growth medium by adding 2 μL of DMSO compound solution to 148 μL of medium and mix thoroughly. Make dilutions in 384-well stock plates (Greiner, catalog no. 781270).
2. Further dilute these plates in series by adding 75 μL of solution to 75 μL of fresh growth medium, proceeding across the 384-well plate, and generate concentration-response curves.
3. Next, add 36 μL of cell suspension at 28,000 cells per well to the assay plates (1000 cells/well), and add 4 μL of the medium containing the dilution series of compound to the cells.
4. Incubate the cells for 48 hours in a TC incubator at 95% humidity, 5% CO₂, 37°C.
5. To measure viability, add 10 μL Alamar Blue solution (50% in growth medium) to each well.
6. Incubate the cells for 16 hours, and read fluorescence intensity (544 nM excitation, 590 nM emission.)

Spectroscopic Data for SAR Analogs

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 15945539

UPLC Chromatogram of Analog CID 15945539

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766535

LCMS Chromatogram of Analog CID 49766535

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766526

UPLC Chromatogram of Analog CID 49766526

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766512

UPLC Chromatogram of Analog CID 49766512

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766538

UPLC Chromatogram of Analog CID 49766538

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766527

UPLC Chromatogram of Analog CID 49766527

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 3244419

UPLC Chromatogram of Analog CID 3244419

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 15945537

UPLC Chromatogram of Analog CID 15945537

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 3242450

UPLC Chromatogram of Analog CID 3242450

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 5307213

UPLC Chromatogram of Analog CID 5307213

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 9550629

UPLC Chromatogram of Analog CID 9550629

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 5307265

UPLC Chromatogram of Analog CID 5307265

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766531

UPLC Chromatogram of Analog CID 49766531

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 18271554

UPLC Chromatogram of Analog CID 18271554

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766550

UPLC Chromatogram of Analog CID 49766550

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49786173

UPLC Chromatogram of Analog CID 49786173

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 9337176

UPLC Chromatogram of Analog CID 9337176

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 16188083

LCMS Chromatogram of Analog CID 16188083

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 16189308

LCMS Chromatogram of Analog CID 16189308

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 46255915

LCMS Chromatogram of Analog CID 46255915

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 20864743

LCMS Chromatogram of Analog CID 20864743

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 15945452

UPLC Chromatogram of Analog CID 15945452

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766520

UPLC Chromatogram of Analog CID 49766520

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766541

UPLC Chromatogram of Analog CID 49766541

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766542

UPLC Chromatogram of Analog CID 49766542

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766547

UPLC Chromatogram of Analog CID 49766547

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766518

UPLC Chromatogram of Analog CID 49766518

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766549

UPLC Chromatogram of Analog CID 49766549

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766530

UPLC Chromatogram of Analog CID 49766530

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49766534

UPLC Chromatogram of Analog CID49766534

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 29218174

UPLC Chromatogram of Analog CID 29218174

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49786174

UPLC Chromatogram of Analog CID 49786174

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49786176

UPLC Chromatogram of Analog CID 49786176

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49786172

UPLC Chromatogram of Analog CID 49786172

¹HNMR Spectrum (300 MHz, CDCl₃) of Analog CID 49786175

UPLC Chromatogram of Analog CID 49786175