J. Med. Chem. 2006, 49, 7887-7896
`
`7887
`
`Quinol-4-ones as Steroid A-Ring Mimetics in Nonsteroidal Dissociated Glucocorticoid Agonists
`
`John Regan,*,† Thomas W. Lee,† Rene´e M. Zindell,† Younes Bekkali,† Jo¨rg Bentzien,† Thomas Gilmore,†
`Abdelhakim Hammach,† Thomas M. Kirrane,† Alison J. Kukulka,† Daniel Kuzmich,† Richard M. Nelson,† John R. Proudfoot,†
`Mark Ralph,† Josephine Pelletier,‡ Donald Souza,‡ Lijiljana Zuvela-Jelaska,‡ Gerald Nabozny,‡ and David S. Thomson†
`
`Departments of Medicinal Chemistry and Immunology & Inflammation, Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury Road,
`Ridgefield, Connecticut 06877
`
`ReceiVed October 31, 2006
`
`We report on the nuclear receptor binding affinities, cellular activities of transrepression and transactivation,
`and anti-inflammatory properties of a quinol-4-one and other A-ring mimetic containing nonsteroidal class
`of glucocorticoid agonists.
`
`Introduction
`
`During the past 50 years the successful treatment of inflam-
`matory diseases has relied on the use of glucocorticoid (GC)a
`agonists such as dexamethasone (1a) and prednisolone (1b)
`(Chart 1). While effective in controlling asthma,1 rheumatoid
`arthritis,2 and other disorders, GC therapy is fraught with a
`number of severe side effects. The seriousness of the complica-
`tions often hampers high dose and chronic administration. In
`addition, cross reactivity with other steroid hormone receptors,
`such as progesterone receptor (PR) and mineral corticoid
`receptor (MR), of the commonly prescribed GCs can provoke
`off-target pharmacology. In direct response to the current state
`of GC therapy, finding drugs with good anti-inflammatory
`properties and reduced side effects remains an ongoing quest.
`The probability of identifying a GC agonist with a better
`safety profile compared to existing therapies has substantially
`increased with newer understandings of the molecular mecha-
`nism of action.3-8 After an agonist enters a target cell and binds
`to the glucocorticoid hormone receptor (GR),
`the ligand-
`activated complex (GRC) translocates into the nucleus where
`direct and indirect functional pathways can be accessed. Acting
`directly, the GRC serves as an endogenous transcription factor
`by binding to specific DNA sequences and coactivator proteins,
`thereby initiating transcription of metabolic and endocrine genes.
`GRC-mediated transactivation of these genes is believed to
`contribute to the side effect profile of GC therapy.9 Acting
`indirectly, the GRC adopts a conformation with an affinity for
`transcription factors (e.g., NF-kB and AP-1). Subsequent binding
`to these transcription factors results in the inhibition of
`expression of pro-inflammatory cytokines such as TNF-R and
`IL-6. This process, known as transrepression, is thought to
`contribute, in part, to the anti-inflammatory component of GCs.
`
`(203) 798-
`* To whom correspondence should be addressed. Phone:
`4768. Fax: (203) 791-6072. E-mail: jregan@rdg.boehringer-ingelheim.com.
`† Department of Medicinal Chemistry.
`‡ Department of Immunology & Inflammation.
`a Abbreviations: AP-1, activator protein-1; DBF, dihydrobenzofuran;
`DELFIA, dissociation-enhanced lanthanide fluorescent immunoassay; DMEM,
`Dulbecco’s modified Eagle’s medium; D-gal, D-galactose; CHAPS, 3-[(3-
`cholamidopropyl)dimethylammonio]-1-propanesulfonate; EDTA, ethylene-
`diamimetetraacetic acid; ELISA, enzyme-linked immunosorbent assay; FP,
`fluorescence polarization; GC, glucocorticoid; GRC, glucocorticoid recep-
`tor-ligand complex; GR, glucocorticoid receptor; GR-LBD, glucocorticoid
`receptor-ligand binding domain; HFF, human foreskin fibroblast; HSP,
`heat-shock protein; IL-6, interleukin-6; LPS, lipopolysaccharide; MMTV,
`mouse mammary tumor virus; MR, mineral corticoid receptor; NF, nuclear
`factor; PR, progesterone receptor; TES, N-[tris(hydroxymethyl)methyl]-2-
`aminoethanesulfonic acid; TNF-R, tumor necrosis factor-R.
`
`Therefore, the search for GC agonists with a dissociated profile
`(greater transrepression than transactivation activity) has ac-
`celerated in recent years10,11 with an appreciation of the complex
`molecular pathways and the anticipation of an improved safety
`margin.12,13
`A primary objective in our GC agonist anti-inflammatory
`program is identifying nonsteroidal anti-inflammatory agents
`with the added benefit of exhibiting dissociation. Toward this
`end, we sought to discover pharmacophores capable of playing
`the hydrogen bond acceptor’s role of the C-3 carbonyl of
`dexamethasone that could be combined with a nonsteroidal
`scaffold. Dissociated GC agonists 2a and 2b14 represent a good
`starting point for such a venture as they contain a benzoxazinone
`group as a readily replaceable A-ring mimetic.15,16 Other
`important features include the alcohol as a steroidal C-11
`hydroxy surrogate, a trifluoromethyl substituent to maintain
`agonist activity,17 and a phenyl ring which is believed to reside
`adjacent to the area of steroid D-ring binding in the glucocor-
`ticoid receptor-ligand binding domain (GR-LBD). Expanding
`the scope of this effort to include modifications to the phenyl
`ring provides an opportunity to gauge whether changes to this
`region of the scaffold are a viable option to attenuate nuclear
`receptor potency and selectivity, as well as transrepression or
`transactivation pathways. We envision preparing the target
`molecules, represented by 5, by the opening of epoxide 3 (Chart
`1) with a nucleophilic amine embedded in a lipophilic ring
`containing a hydrogen bond acceptor functionality18 (i.e., atom
`A in 4). The syntheses and biological profiles of compounds
`such as 5 are the subject of this paper.
`
`Chemistry
`
`The preparations of a quinol-4-one-containing analogue (11)
`and a piperid-4-one (14) are shown in Scheme 1 as representa-
`tive examples of 5. The key step in this sequence is the opening
`of an epoxide (i.e., 9) with a nucleophilic amine. Briefly, the
`preparation of epoxide 9 begins with the copper-mediated 1,4-
`addition of Grignard reagent 6 to 1,1,1-trifluorobut-3-en-2-one
`(7)19 to provide ketone 8. Trimethylsulfoxonium iodide and NaH
`transform 8 to epoxide 9. Quinolone 11 is obtained by reacting
`9 with 4-hydroxyquinoline (10) and sodium ethoxide in etha-
`nol.20 Conversion of the methoxy group in 11 to a hydroxy in
`12 is accomplished with BBr3. Piperidin-4-one analogue 14 is
`prepared from the reaction of 9, cis-3,5-dimethyl-4-piperidone
`hydrochloride (13)21,22 and K2CO3 in DMF.
`
`10.1021/jm061273t CCC: $33.50 © 2006 American Chemical Society
`Published on Web 12/21/2006
`
`Downloaded via UNIV OF PENNSYLVANIA on January 29, 2021 at 03:36:57 (UTC).
`
`See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
`
`Liquidia's Exhibit 1023
`IPR2020-00770
`Page 1
`
`

`

`7888 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26
`
`Regan et al.
`
`Chart 1
`
`Scheme 1a
`
`a Reagents and conditions: (a) CuI (1.1 equiv)/THF, -20 (cid:176)C, 30 min, then 7 (1.1 equiv), room temperature overnight; (b) trimethylsulfoxonium iodide/
`NaH/DMSO; (c) 10/NaOEt/ethanol, heat; (d) BBr3 (10 equiv)/CH2Cl2, room temperature overnight; (e) 13 (2 equiv)/K2CO3 (5 equiv)/DMF, 100 (cid:176)C, 1.5 h.
`
`Results and Discussion
`The nuclear receptor binding potency of the target compounds
`is determined by their ability to compete for receptor binding
`with tetramethylrhodamine-labeled dexamethasone (GR, MR)
`or mifepristone (PR).23 The transrepression potential is measured
`in human foreskin fibroblast (HFF) cells. Inhibition of IL-1
`stimulated IL-6 production, and the percent efficacy is compared
`with that of dexamethasone.24,25 In an effort
`to establish
`transactivation activity, compounds are counterscreened for the
`induction of aromatase in HFF cells26 and the ability to activate
`the MMTV promoter in HeLa cells transfected with an MMTV
`luciferase contruct.27 Comparisons with dexamethasone establish
`a basis for identifying compounds with a dissociated profile.
`Table 1 shows the effects of A-ring replacements in scaffold
`5. Morpholine 15, pyridin-4-one 17, and imidazole 18 are devoid
`of activity, while piperidin-4-one 16 demonstrates weak GR
`binding affinity (IC50 ) 1400 nM). Introducing methyl groups
`on the ring is an opportunity to increase the lipophilicity and
`
`perhaps provide a water shield to the hydrogen bond network
`consisting of the A-ring mimetic acceptor and the GR-LBD
`donor. The outcome is a significant enhancement in GR binding
`affinity. For example, cis-dimethylpiperidin-4-one 14 exhibits
`a 140-fold gain in GR binding (IC50 ) 10 nM). Other methylated
`analogues include cis-dimethylmorpholine 20, dimethylpyridin-
`4-one 21, and dimethylimidazole 22 with IC50 values of 38,
`120, and 650 nM, respectively. Introducing methyl groups onto
`the A-ring replacements has no effect on PR binding affinity,
`whereas modest activity against MR is seen. In the transrepres-
`sion anti-inflammatory cellular assay, compounds 14 and 21
`inhibit IL-6 release from HFF with IC50 values of 75 and 53
`nM, respectively, and 87-89% efficacy compared to dexa-
`methasone.
`Further increasing the lipophilicity of the A-ring mimetic from
`methyl substitutions to fusing an aromatic nucleus to the
`hydrogen bond acceptor ring provides analogues shown in Table
`2. Changes in GR binding avidity with this modification vary
`
`Liquidia's Exhibit 1023
`IPR2020-00770
`Page 2
`
`

`

`Quinol-4-ones as Steroid A-Ring Mimetics
`
`Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26 7889
`
`Table 1. Methyl-Substituted A-Ring Mimetics
`
`a IC50 values were determined from duplicate 11-point concentration-response curves. See ref 23. b IC50 values represent the mean of at least two independent
`determinations. See ref 25. c Not tested.
`
`Table 2. Phenyl-Fused A-Ring Mimetics
`
`a IC50 values were determined from duplicate 11-point concentration-response curves. See ref 23. b IC50 values represent the mean of at least two independent
`determinations. See ref 25. c Not tested.
`
`between 2- and 10-fold. For example, compare 14 with 24, 22
`with 26, and 21 with 11. However, of these compounds, only
`quinolone 11 potently inhibits cellular IL-6 production (IC50 )
`21 nM and 82% efficacy). These results reinforce the often
`observed lack of correlation between GR binding affinity and
`
`transrepression activity.10,28,29 A likely explanation is that the
`flexibility of the GR-LBD can accommodate these ligands but,
`in doing so, is rendered incapable of effectively interacting with
`anti-inflammatory transcription factors.30,31 Support for the
`critical role of the hydrogen bond accepting atom in the A-ring
`
`Liquidia's Exhibit 1023
`IPR2020-00770
`Page 3
`
`

`

`7890 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26
`
`Regan et al.
`
`Table 3. Phenyl Ring Modifications
`
`compd
`no.
`
`R2
`
`R4
`
`R5
`
`a
`
`GR IC50
`(nM)
`
`a
`
`PR IC50
`(nM)
`
`MR IC50
`a
`(nM)
`
`IL-6 inhib in
`b (nM)
`HFF IC50
`(efficacy, % vs
`dexamethasone)
`
`OMe H F
`OH
`H F
`H
`H H
`OH
`H H
`Me
`H H
`Me
`H F
`OH
`H Me
`Me
`H OMe
`Me
`H OH
`OH
`F
`H
`
`11
`27 (82)
`120
`475
`10
`12
`6 (92)
`580
`960
`5
`30
`78 (48)
`350
`106
`28
`31
`6 (90)
`450
`555
`6
`32
`61 (63)
`150
`240
`22
`33
`34 (66)
`120
`240
`10
`34
`5.6 (93)
`160
`160
`8
`35
`17 (86)
`91
`525
`11
`>2000
`>2000
`36
`ntc
`780
`37
`3 (88)
`880
`340
`3
`38d
`11 (86)
`98
`48
`4
`a IC50 values were determined from duplicate 11-point concentration-
`response curves. See ref 23. b IC50 values represent the mean of at least
`two independent determinations. See ref 25. c Not tested. d See Table 5 for
`the structure.
`
`replacements is highlighted with tetrahydroquinoline 25. A factor
`of 200 decrease in GR binding affinity, compared to that of
`dihydroquinolone 24, underscores the importance of achieving
`a hydrogen bond network in the GR-LBD with this series. The
`3-fold difference in GR binding affinity between benzo[1,4]-
`oxazine 23 and dihydroquinolone 24 may reflect a stronger
`hydrogen bond acceptor pharmacophore and an advantageous
`distance for 24 between the hydrogen bond donor of the protein
`and the acceptor group on the ligand. One consequence of fusing
`a phenyl ring to this scaffold is an erosion of the GR/PR ratio
`from greater than 75-fold for compounds without a fused
`aromatic nucleus (14 and 20) to less than 50-fold for those with
`a phenyl ring attached (11, 23, 24). The GR/MR binding ratio
`was nearly unchanged. Interestingly, benzimidazole 26 does not
`show a loss of nuclear receptor selectivity. Additional changes
`to the A-ring mimetic involve embedding a nitrogen atom into
`the aromatic nucleus as a goal to define polarity limits. Table
`2 describes the consequences of a quinazolinone (27) and
`naphthyridin-4-ones (28, 29) acting as A-ring replacements.
`Compared to 11, a significant diminution in GR binding is
`observed for 27 and less so with 28 and 29. Nevertheless, both
`28 and 29 retain satisfactory transrepression activity. These data
`suggest the lipophilic environment of the A-ring binding domain
`in GR can tolerate ligands with somewhat increased polarity
`whereas the PR-LBD does not readily accommodate these
`changes. Of the phenyl-fused A-ring replacements shown in
`Table 2, compounds 11, 23, 24, 26, 28, and 29 achieve the
`highest GR binding affinity, but only 11, 28, and 29 reach IC50
`values <100 nM in the transrepression assay.
`Molecular modeling suggests that, upon binding to the GR-
`LBD, the phenyl ring of 2b resides adjacent to the steroid D-ring
`domain.15,16 The contribution to GR binding and transrepression
`activity of simple groups attached to the phenyl ring in the
`quinolone-containing scaffold is summarized in Table 3.
`Conversion of the methoxy group at C-2 (11, R2 ) OMe) to a
`hydroxy (12, R2) OH) improves the transrepression activity.
`Removal of the hydroxy group and fluorine atom from 12
`(compound 30) weakens the GR binding affinity (IC50 ) 28
`nM) and decreases the agonist activity in the cellular assay (IC50
`) 78 nM and 48% efficacy). Also seen is a decline in the GR/
`PR ratio. Maintaining only a hydroxy group (31) restores the
`binding (GR IC50 ) 6 nM) and cellular activity (IC50 ) 6 nM
`
`and 90% efficacy). In addition, a good GR/PR binding ratio is
`re-established. Switching the hydroxy group at C-2 to a methyl
`(i.e., R2 ) Me, 32 and 33) maintains the GR binding affinity
`but lessens the GR/MR ratio. Reductions in IL-6 inhibition (IC50
`values of 30-60 nM) and lower efficacies are also observed
`(63-66%) for this change. Substituting an electronegative atom
`at C-5 for an electropositive atom (34) retains the GR binding
`and cellular activity with an increase in PR and MR binding
`affinities compared to those of 12. Further changing the
`electronic character at C-5 by introducing a methoxy improves
`the agonist activity compared to that of 33. However, placing
`a hydroxy group at C-5 (36) causes a significant loss of GR
`binding, implying a negative interaction of a hydrogen bond
`donor at this site. Phenyl ring analogues involving substitution
`at C-4 reveal a very narrow range of acceptable pharmacophores.
`Only fluorine at this site provides good GR binding affinity
`and cellular activity (37). Groups larger than fluorine (methyl,
`methoxy, hydroxy) generally lead to a significant loss of binding
`affinity (data not shown). Incorporating the oxygen atom at C-2
`into a dihydrobenzofuran (DBF) ring (38) significantly dimin-
`ishes the nuclear receptor selectivity. The transrepression activity
`is comparable to that of 11. The above survey of substituents
`on the phenyl ring of the quinolone-containing steroid mimetic
`generally demonstrates a range of nuclear receptor binding
`affinity smaller than a factor of 10. However, pharmacophores
`such as methoxy or hydroxy at C-2 (11, 12, 31, 34, and 37)
`and DHB (38) are required for achieving the best agonist profile.
`Another general trend (data not shown) for this series is a
`hydroxy group at C-2 displays better IL-6 inhibition and efficacy
`compared to a methoxy despite similar GR binding affinities
`(e.g., 11 vs 12). Also important are the substitution patterns at
`C-2 and C-5 (34 vs 35) and steric constraints at C-4. Taken
`together, these results highlight the important relationships of
`size and polarity required for favorable interactions adjacent to
`the steroid D-ring domain within the GR-LBD. In addition,
`similar observations are noted with phenyl ring substitutions
`for molecules containing A-ring mimetics other than quinol-4-
`one. See Table 5 for examples.
`A primary goal of this project is identifying GC agonists that
`can distinguish between transrepression and transactivation
`pathways. Two assays are used as a counterscreen to determine
`the extent of transactivation. Dexamethasone activates the
`MMTV promoter in HeLa cells transfected with an MMTV
`luciferase construct. In addition, dexamethasone induces expres-
`sion of aromatase in HFF.32 Several compounds were evaluated
`for their transactivation potential to activate MMTV and induce
`aromatase expression (EC50 and percent efficacy). A compound
`can be characterized as dissociated with either weak activity or
`low efficacy, or both, in the transactivation assays compared to
`the transrepression assay. The role that phenyl ring substitutions
`play in defining dissociation is seen in Table 4. In the MMTV
`assay, 2-hydroxyphenyl compounds 12 and 37 are partial
`agonists (EC50 ) 15-37 nM and 20-35% efficacy) while the
`other compounds are devoid of activity compared to dexa-
`methasone. In the aromatase induction assay, 12 and 37 exhibit
`pronounced activity with EC50 ) 15-17 nM and 65-74%
`efficacies. Alternatively, 2-methoxyphenyl, 2-methylphenyl, and
`DBF analogues (11, 32, and 38) display low potency or efficacy
`in the aromatase assay. Of the modifications at C-5 of the phenyl
`ring, methoxy 35 displays moderate activity against aromatase
`(EC50 ) 95 nM, 65% efficacy) while hydrogen or fluorine (32
`and 33) are weakly potent and efficacious derivatives. These
`data suggest groups on the phenyl ring have an important
`influence on the degree of dissociation and parallel the observa-
`
`Liquidia's Exhibit 1023
`IPR2020-00770
`Page 4
`
`

`

`Quinol-4-ones as Steroid A-Ring Mimetics
`
`Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26 7891
`
`Table 4. Dissociation Profiles
`
`Table 5. In Vivo Activities of Dissociated Agonists
`
`compd
`no.
`
`GR IC50
`a
`(nM)
`
`IL-6 inhib in
`b (nM)
`HFF IC50
`(efficacy, % vs
`dexamethasone)
`
`MMTV activation in
`HeLa EC50
`c (nM)
`(efficacy, % vs
`dexamethasone)
`
`aromatase inhib in
`HFF EC50
`d (nM)
`(efficacy, % vs
`dexamethasone)
`
`1a
`1.9 (100)
`17 (100)
`0.5 (100)
`3.4
`1b
`19 (92)
`16 (91)
`6.6 (96)
`15
`>2000 (14)
`2a
`91 (56)
`59 (80)
`3
`>2000 (10)
`11
`220 (67)
`27 (82)
`10
`12
`14 (74)
`37 (20)
`6 (92)
`5
`14
`240 (93)
`nte
`75 (89)
`10
`>2000 (3)
`21
`170 (77)
`53 (87)
`120
`28
`ipf (41)
`nt
`56 (78)
`77
`>2000 (1)
`29
`ip (39)
`35 (72)
`34
`>2000 (29)
`>2000 (1)
`32
`61 (63)
`22
`>2000 (2)
`33
`ip (22)
`34 (66)
`10
`>2000 (6)
`35
`95 (65)
`17 (86)
`11
`37
`17 (82)
`15 (35)
`3 (88)
`3
`38g
`ip (60)
`nt
`11 (86)
`4
`39g
`ip (26)
`nt
`39 (77)
`34
`>2000 (14)
`40g
`nt
`47 (59)
`48
`>2000 (20)
`41g
`nt
`70 (71)
`125
`a IC50 values were determined from duplicate 11-point concentration-
`response curves. See ref 23. b IC50 values represent the mean of at least
`two independent determinations. See ref 25. c EC50 values represent the
`mean of at least two independent determinations. See ref 27. d EC50 values
`represent the mean of at least two independent determinations. See ref 26.
`e Not tested. f Due to the flat shape of the dose-response curve, an EC50
`value beyond 2000 nM could not be calculated. g See Table 5 for the
`structures.
`
`tion that some steroids containing modified D-rings show
`dissociated activities.33 The contribution of the A-ring replace-
`ment to the dissociation profile is seen in Table 4. Comparing
`compounds with the 2-methoxy-5-fluorophenyl ring, dimeth-
`ylpiperidin-4-one 14, dimethylpyridin-4-one 21, and naphthy-
`ridin-4-one 28 demonstrate a somewhat weaker transactivation
`response compared to quinol-4-one 12. However, the 2-methyl-
`5-fluorophenyl and DBF rings impart a much more dissociated
`profile for dimethylpyridin-4-one (39 and 41), naphthyridin-4-
`one (40), and quinol-4-one (33 and 38) A-ring mimetics. Thus,
`a distinction between potent transrepression activity and dimin-
`ished transactivation is achieved with the proper choice of A-ring
`replacements and phenyl substitution with this class of com-
`pounds.
`Selected dissociated compounds were evaluated for their anti-
`inflammatory properties in an LPS-stimulated mouse model of
`TNF-R production. Test compounds were administered orally
`in Cremophor 60 min prior to LPS challenge. TNF-R levels
`were measured 60 min later. Table 5 summarizes the GR binding
`and cellular and in vivo transrepression profiles of various
`combinations of A-ring (dimethylpyridin-4-one, quinol-4-one,
`and naphthyridin-4-one) and phenyl ring analogues. When dosed
`at 10 mg/kg, the nondissociated analogue 12 and dissociated
`compounds 33 and 38-41 inhibit TNF-R production (48-58%).
`
`Conclusion
`Several new types of pharmacophores can serve as A-ring
`mimetics with good binding affinity for GR for this scaffold.
`These include dimethylpiperidin-4-one (14), quinolone (11), cis-
`dimethylmorpholine (20), benzo[1,4]oxazine (23), dihydroqui-
`nolone (24), benzimidazole (26), and naphthyridin-4-one (29).
`In addition, good nuclear receptor selectivity is achieved. Of
`these, 11, 29, and dimethylpyridin-4-one (21) exhibit good
`transrepression activity in a cellular assay of IL-6 production.
`The substitution patterns on the phenyl ring play an important
`role in defining GR binding affinity and the degree of trans-
`repression. Compounds with a hydroxy group at C-2 obtain the
`highest efficacy levels of transrepression, while a C-5 hydroxy
`is deleterious toward activity. Methyl and methoxy substituents
`
`compd
`no.
`
`A
`
`phenyl
`ring
`
`IL-6 inhib in
`b (nM)
`HFF IC50
`(efficacy, % vs
`dexamethasone)
`
`a
`
`GR IC50
`(nM)
`
`inhib of
`TNF-R in micec
`(%) (n ) 8)
`(10 mg/kg)
`84 ( 8
`1bd
`7 (96)
`15
`see Chart 1
`55 ( 6
`12
`6 (92)
`5
`see Table 3
`36 ( 10
`32
`61 (63)
`22
`II
`II
`50 ( 17
`33
`34 (66)
`10
`I
`II
`49 ( 10
`38
`11 (86)
`4
`III
`II
`49 ( 6
`39
`39 (77)
`34
`III
`I
`58 ( 12
`40
`47 (59)
`48
`I
`III
`48 ( 11
`41
`70 (71)
`125
`I
`I
`a IC50 values were determined from duplicate 11-point concentration-
`response curves. See ref 23. b IC50 values represent the mean of at least
`two independent determinations. c See the Experimental Section. d 3 mg/
`kg, po.
`
`at C-2 and C-5 are well tolerated. The dissociated profile of
`this series in either the MMTV or aromatase assays was
`dependent upon the choice of the A-ring mimetic and the
`substitution on the phenyl ring. As A-ring replacements, fused
`rings 11, 28, and 29 exhibit a superior dissociated profile due
`to differences in either potency or efficacy in the transactivation
`assays compared to monocyclic rings 14 and 21. In general,
`methyl and methoxy groups at C-2 of the phenyl ring demon-
`strate weaker transactivation activities than hydroxy at this site.
`In an animal model of inflammation, compounds 12, 33, and
`38-41 potently inhibit TNF-R production. Thus, the nuclear
`receptor binding affinities, the cellular levels of transrepression
`and transactivation, and the anti-inflammatory properties have
`been demonstrated with the quinol-4-one-containing class of
`GR agonists. Further work in this series will be reported in due
`course.
`
`Experimental Section
`All solvents and reagents were obtained from commercial sources
`and used without further purification. Flash chromatography was
`performed according to the procedure of Still et al. (EM Science
`Kieselgel 60, 70-230 mesh). Melting points were obtained from a
`Mel-temp 3.0 or Fisher-Johns melting point apparatus and are
`uncorrected. The 1H NMR spectra were recorded on a Bruker DPX
`400 spectrometer. Chemical shifts are reported in parts per million
`((cid:228)) from the tetramethylsilane resonance in the indicated solvent.
`Infrared spectra were recorded without solvent on a Nicolet Impact
`410. Mass spectra were obtained from a Finnigan-SSQ7000
`spectrometer. Samples were generally introduced by a particle beam
`and ionized with NH4Cl. The analytical HPLC purity was estab-
`lished with the following protocols. Method A: Gilson chromato-
`graphic system, Waters SureFire reversed-phase C18 analytical
`column (4.6 (cid:2) 50 mm, 5 (cid:237)m, part number 186002557), flow rate
`5 mL/min, detection with a Gilson UV/VIS-155 at 254 nm, sample
`
`Liquidia's Exhibit 1023
`IPR2020-00770
`Page 5
`
`

`

`7892 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 26
`
`Regan et al.
`
`size 25 (cid:237)L, gradient 5% acetonitrile in water with 0.1% TFA to
`95% acetonitrile in water with 0.1% TFA over 6 min and held.
`Method B: Rainin chromatography system, model SD-200, Tosoh
`Bioscience TSK Super ODS column (4.6 mm (cid:2) 10 cm, 2 (cid:237)M,
`part number G0030-90F), flow rate 1.2 mL/min, detection with
`Dynamax absorbance detector at 254 nM, gradient 5% acetonitrile
`in water with 0.05% TFA to 95% acetonitrile in water with 0.05%
`TFA over 10 min and held. Method C: Rainin chromatography
`system, model SD-200, Whatman PartiSphere C-8 column (4.6 (cid:2)
`125 mm, 5 (cid:237)m, part number 4621-0503), detection with Dynamax
`absorbance detector at 254 nM, gradient 5% acetonitrile in water
`with 0.1% TFA to 95% acetonitrile in water with 0.1% TFA.
`1,1,1-Trifluoro-4-methylpent-3-en-2-one (7). To a 0-5 (cid:176)C
`mixture of N,O-dimethylhydroxylamine hydrochloride (15.8 g) in
`CH2Cl2 (400 mL) was added dropwise trifluoroacetic anhydride
`(21.7 mL). Pyridine (37 mL) was added to the mixture dropwise.
`The resulting mixture was stirred at 0-5 (cid:176)C for 0.5 h and diluted
`with water. The organic layer was washed with water, 1 N aqueous
`HCl, water, and brine and dried (MgSO4). Removal of the volatiles
`in vacuo for 5 min provided 2,2,2-trifluoro-N-methoxy-N-methyl-
`acetamide as a colorless oil.
`To a 0-5 (cid:176)C mixture of 2,2,2-trifluoro-N-methoxy-N-methyl-
`acetamide (3 g) in 30 mL of anhydrous ether was added (2-
`methylpropenyl)magnesium bromide in THF (42 mL of a 0.5 M
`solution). The reaction was stirred at 0-5 (cid:176)C for 0.5 h, warmed to
`room temperature, and stirred overnight. The reaction was quenched
`with saturated aqueous NH4Cl and extracted with ether three times.
`The organic layers were combined, washed with water and brine,
`and dried (MgSO4). Most of the volatiles were removed at
`atmospheric pressure, providing 7 as an approximately 75% pure
`solution containing ether and THF that was used without further
`purification.
`1,1,1-Trifluoro-4-(5-fluoro-2-methoxyphenyl)-4-methylpentan-
`2-one (8). To a 0-5 (cid:176)C solution of 7 (24 g, approximately 75%
`pure solution containing ether and THF) and CuI (30.1 g) in ether
`(300 mL) was added (5-fluoro-2-methylphenyl)magnesium bromide
`(315 mL, 157 mmol, 0.5 M solution in THF). The mixture was
`warmed to room temperature, stirred overnight, and diluted with
`saturated aqueous NH4Cl (15 mL) and ethyl acetate. The organic
`layer was washed with water and brine and dried (MgSO4).
`Removal of the volatiles in vacuo provided a residue which was
`purified by flash silica gel chromatography using 0-5% ethyl
`acetate in hexanes as the eluent. The product-rich fractions were
`collected and the volatiles removed in vacuo to yield 20.3 g (46%)
`of 8 as an oil.
`2-[2-(5-Fluoro-2-methoxyphenyl)-2-methylpropyl]-2-(trifluo-
`romethyl)oxirane (9). To a suspension of trimethylsulfoxonium
`iodide (4.76 g, 21.5 mmol) in DMSO (25 mL) was added NaH
`(0.863 g of a 60% dispersion in oil, 21.5 mmol) in four portions.
`After being stirred for 0.5 h, the solution was added to 8 (5.00 g,
`17.9 mmol) in DMSO (20 mL). After 2 h the mixture was diluted
`with water (150 mL) and extracted with ether (2 (cid:2) 125 mL). The
`combined organic layers were washed with water and brine and
`dried (MgSO4). Removal of the volatiles in vacuo provided 9 as
`an oil which was used without further purification.
`1-[4-(5-Fluoro-2-methoxyphenyl)-2-hydroxy-4-methyl-2-(tri-
`fluoromethyl)pentyl]-1H-quinolin-4-one (11). A mixture of 9
`(0.226, 0.773 mmol), 4-hydroxyquinoline (10; 0.112 g, 0.773
`mmol), sodium ethoxide (0.28 mL of a 21 wt % solution in ethanol,
`0.77 mmol), and anhydrous ethanol (4 mL) was heated at 85 (cid:176)C
`for 6 h, cooled to room temperature, diluted with ethyl acetate and
`acetic acid, washed with water and brine, and dried (MgSO4).
`Removal of the volatiles in vacuo provided a residue which was
`purified with flash silica gel chromatography using ethyl acetate
`as the eluent. The product-rich fractions were collected and the
`volatiles removed in vacuo to yield 0.13 g (38%) of 11. Mp: 198-
`200 (cid:176)C.
`1H NMR (400 MHz, CDCl3): (cid:228) 8.21 (d, 1H), 7.42 (m,
`2H), 7.25 (s, 1H), 7.1 (m, 1H), 6.92 (m, 2H), 6.80 (m, 1H), 4.15
`(dd, 2H), 3.80 (s, 3H), 2.55 (dd, 2H), 1.5 (s, 3H), 1.33 (s, 3H).
`FTIR neat (cm-1): 1624, 1610. MS (ESI): m/e 438 (MH+). HPLC
`
`purity: method A 99%, retention time (tR) 3.53 min; method B
`98%, tR 7.98 min.
`1-[4-(5-Fluoro-2-hydroxyphenyl)-2-hydroxy-4-methyl-2-(tri-
`fluoromethyl)pentyl]-1H-quinolin-4-one (12). To a solution of 11
`(0.082 g, 0.19 mmol) in CH2Cl2 (3 mL) at -20 (cid:176) C was added
`dropwise BBr3 (1.9 mL of a 1 Msolution in CH 2Cl2, 1.9 mmol).
`The mixture was stirred for 1 h and quenched with saturated
`NaHCO3 and ethyl acetate. The organic layer was washed with
`saturated NaHCO3 and brine and dried (MgSO4). Removal of the
`volatiles in vacuo left a residue which was purified by trituration
`in ether and ethyl acetate and provided 0.041 g of product. Mp:
`222-223 (cid:176)C.
`1H NMR (400 MHz, DMSO-d6): (cid:228) 9.98 (s, 1H),
`8.30 (d, 1H), 7.87 (d, 1H), 7.75 (t, 1H), 7.52 (t, 1H), 7.33 (m, 2H),
`7.13 (m, 1H), 7.05 (m, 1H), 7.48 (s, 1H), 6.20 (d, 1H), 4.21 (m,
`2H), 3.24 (d, 1H), 2.15 (d, 1H), 1.75 (s, 3H), 1.55 (s, 3H). MS
`(ESI): m/e 424 (MH+). HPLC purity: method A 98%, tR 3.11 min;
`method B 95%, tR 7.16 min.
`cis-1-[4-(5-Fluoro-2-methoxyphenyl)-2-hydroxy-4-methyl-2-
`(trifluoromethyl)pentyl]-3,5-dimethylpiperidin-4-one Hydro-
`chloride (14). A mixture of 9 (0.092 g, 0.32 mmol), cis-3,5-
`dimethylpiperidin-4-one hydrochloride (13; 0.103 g, 0.64 mmol),
`and powdered K2CO3 (0.218 g, 1.58 mmol) in DMF (3 mL) was
`heated at 100 (cid:176)C for 1.5 h, cooled to room temperature, diluted
`with ether, washed thoroughly with water and brine, and dried
`(MgSO4). Removal of the volatiles in vacuo provided a residue
`which was treated with ethereal HCl. The mixture was filtered and
`the solid washed with ether and dried in vacuo to provide 0.015 g
`of product. Mp: 77-80 (cid:176)C.
`1H NMR (400 MHz, CDCl3): (cid:228) 7.02
`(dd, 1H), 6.78 (m, 1H), 6.68 (m, 1H), 3.75 (s, 3H), 1.52 (s, 3H),
`1.28 (s, 3H), 0.78 (d, 6H). MS (ESI): m/e 420 (MH+). HPLC
`purity: method C 92%, tR 10.25 min.
`1,1,1-Trifluoro-4-(5-fluoro-2-methoxyphenyl)-4-methyl-2-(mor-
`pholin-4-ylmethyl)pentan-2-ol Hydrochloride (15). Mp: 195-
`200 (cid:176)C.
`1H NMR (400 MHz, CDCl3): (cid:228) 6.97 (dd, 1H), 6.78 (m,
`1H), 6.68 (m, 1H), 4.55 (s, 1H), 3.75 (s, 3H), 3.46 (br s, 4H), 2.33
`(m, 4H), 2.14 (d, 2H), 1.97 (d, 2H), 1.52 (s, 3H), 1.32 (s, 3H). MS
`(ESI): m/e 380 (MH+). HPLC purity: method A 99%, tR 2.41 min;
`method B 99%, tR 6.35 min.
`1-[4-(5-Fluoro-2-methoxyphenyl)-2-hydroxy-4-methyl-2-(tri-
`fluoromethyl)pentyl]piperidin-4-one Hydrochloride (16). Mp:
`134-136 (cid:176) C. 1H NMR (400 MHz, DMSO-d6): (cid:228) 6.8 (m, 3H), 4.55
`(br s, 1H), 3.62 (s, 3H), 2.7-2.0 (m, 12H), 1.36 (s, 3H), 1.18 (s,
`3H). MS (ESI): m/e 392 (MH+). HPLC purity: method C 94%, tR
`7.42 min.
`1-[4-(5-Fluoro-2-methoxyphenyl)-2-hydroxy-4-methyl-2-(tri-
`fluoromethyl)pentyl]-1H-pyridin-4-one (17). Mp: 205-206 (cid:176)C.
`1H NMR (400 MHz, DMSO-d6): (cid:228) 7.48 (d, 2H), 7.12 (m, 3H),
`6.47 (s, 1H), 6.07 (d, 2H), 3.92 (s, 3H), 3.81 (d, 1H), 3.60 (d, 1H),
`2.92 (d, 1H), 2.18 (d, 1H), 1.65 (s, 3H), 1.45 (s, 3H). IR (cm-1):
`1639. MS (ESI): m/e 388 (MH+). HPLC purity: method A 99%,
`tR 2.72 min; method B 99%, tR 6.84 min.
`1,1,1-Trifluoro-4-(5-fluoro-2-methoxyphenyl)-2-(imidazol-1-
`ylmethyl)-4-methylpentan-2-ol (18). Mp: 166-167 (cid:176) C. 1H NMR
`(400 MHz, CDCl3): (cid:228) 8.1 (br s, 1H), 7.2-6.7 (m, 6H), 3.94 (dd,
`2H), 3.78 (s, 3H), 2.45 (s, 2H), 1.44 (s, 3H), 1.33 (s, 3H). MS
`(ESI): m/e 361 (MH+). HPLC purity: method A 99%, tR 2.56 min;
`method B 98%, tR 6.59 min; method C 96%, tR 8.67 min.
`2-[(trans-2,6-Dimethylmorpholin-4-yl)methyl]-1,1,1-trifluoro-
`4-(5-fluoro-2-methoxyphenyl)-4-methylpentan-2-ol Hydrochlo-
`ride (19). Mp: 50-55 (cid:176)C. Free base
`1H NMR (400 MHz,
`CDCl3): (cid:228) 7.02 (m, 1H), 6.81 (m, 1H), 6.69 (m, 1H), 4.68 (br d,
`1H), 3.76 (m, 5H), 2.26 (m, 4H), 2.1-1.85 (m, 4H), 1.51 (s, 3H),
`1.30 (s, 3H), 1.03 (br s, 6H). MS (ESI): m/e 408 (MH+). HPLC
`purity: method A 99%, tR 3.16 and 3.61 min; method B 99%, tR
`8.19 and 9.23 min.
`2-[(cis-2,6-Dimethylmorpholin-4-yl)methyl]-1,1,1-trifluoro-4-
`(5-fluoro-2-methoxyphenyl)-4-methylpentan-2-ol Hydrochloride
`(20). Mp: 133-135 (cid:176)C.
`1H NMR (400 MHz, DMSO-d6): (cid:228) 6.8
`(m, 3H), 4.13 (br s, 1H), 3.74 (br s, 2H), 3.58