During the synthesis of deuterated 18-hydroxycortisol, two of the synthetic intermediates have been found to exist in tautomeric forms as the acyclic 18-hydroxy 20-ketone and the cyclic 18,20-hemiketal corresponding to the previously identified less polar (L) and more polar (M) forms of C-18 hydroxylated steroids, respectively. Specifically, p-chloranil oxidation of 18-hydroxycortisol-17,21-acetonide afforded two isomers of the 6,7-dehydro analogue; separate catalytic reduction of each isomer under deuterium gave a single isomer of acetonide-protected 18-hydroxycortisol-1,6,7-d3 for each, with the more polar isomer giving a more polar product and the less polar isomer giving a less polar product. The more polar product (corresponding to M) was characterized as 18,20-hemiketal; 18-hydroxycortisol-17,21-acetonide-18,20-hemiketal-1,6,7-d3: in the deuterochloroform solution, it was found to slowly convert to a substance consistent with the hydroxy ketone structure with features resembling those of the isolated less polar isomer (corresponding to L). Deacetonidization of each gave 18-hydroxycortisol as a single product, which was characterized as the 18,20-hemiketal. The issues associated with the existence of 18-hydroxysteroids as hydroxy ketones and hemiketals, both in solution and as isolable solids, are discussed.

GPR88 is an orphan G protein-coupled receptor which has been implicated in a number of striatal-associated disorders. Herein we describe the synthesis and pharmacological characterization of the first GPR88 radio-ligand, [3H]RTI-33, derived from a synthetic agonist RTI-13951-33. [3H]RTI-33 has a specific activity of 83.4 Ci/ mmol and showed one-site, saturable binding (KD of 85 nM) in membranes prepared from stable PPLS-HA-hGPR88-CHO cells. A competition binding assay was developed to determine binding affinities of several known GPR88 agonists. This radioligand represents a powerful tool for future mechanistic and cell-based ligand-receptor interaction studies of GPR88.

Human trace amine-associated receptor subtype 1 (hTAAR1) is a G protein-coupled receptor that has therapeutic potential for multiple diseases, including schizophrenia, drug addiction, and Parkinson's disease (PD). Although several potent agonists have been identified and have shown positive results in various clinical trials for schizophrenia, the discovery of potent hTAAR1 antagonists remains elusive. Herein, we report the results of structure-activity relationship studies that have led to the discovery of a potent hTAAR1 antagonist (RTI-7470-44, 34). RTI-7470-44 exhibited an IC50 of 8.4 nM in an in vitro cAMP functional assay, a Ki of 0.3 nM in a radioligand binding assay, and showed species selectivity for hTAAR1 over the rat and mouse orthologues. RTI-7470-44 displayed good blood-brain barrier permeability, moderate metabolic stability, and a favorable preliminary off-target profile. Finally, RTI-7470-44 increased the spontaneous firing rate of mouse VTA dopaminergic neurons and blocked the effects of the known TAAR1 agonist RO5166017. Collectively, this work provides a promising hTAAR1 antagonist probe that can be used to study TAAR1 pharmacology and the potential therapeutic role in hypodopaminergic diseases such as PD.

The availability of highly pure animal antibodies is critical in the production of diagnostic tools and biosensors. The peptoid PL16, previously isolated from an ensemble of peptoid variants of the IgG-binding peptide HWRGWV, was utilized in this work as affinity ligand on WorkBeads resin for the purification of immunoglobulin G (IgG) from a variety of mammalian sources and chicken immunoglobulin Y (IgY). The chromatographic protocol initially optimized for murine serum and ascites was subsequently employed for processing rabbit, goat and sheep, donkey, llama, and chicken sera. The PL16-WorkBeads resin proved able to recover all antibody targets with values of yield between 50 and 90%, and purity consistently above 90%. Notably, PL16 not only binds a broader spectrum of animal immunoglobulins than the reference ligands Protein A and G, but it also binds equally well with all their subclasses. Unlike the protein ligands, in fact, PL16 afforded excellent values of yield and purity of mammalian polyclonal IgG, namely murine (47 and 94%), rabbit (66.5 and 91.7%), caprine IgG (63 and 91-95%), donkey, and llama (93 and 97%), as well as chicken IgY (42 and 92%). Of notice, it is also the ability of PL16 to target monomeric IgG without binding aggregated IgG; when challenged with a mixture of monomeric and aggregated murine IgG, PL16 eluted <3% of fed aggregates, against 11-13% eluted by Protein A and G. Collectively, these results prove the potential of the proposed peptoid ligand for large-scale purification of animal immunoglobulins.

Opioid agonists have been used in the field of pain management for centuries, but are considered a major public health concern, especially in the US. One of the biggest challenges consists in understanding the ability to target specifically the different subtypes of opioid receptors and to trigger a specific functional response correlating with an in vivo analgesic response devoid of side‐effects.These opioid receptors are part of the GPCRs family, which is known to have common features across different subfamilies. One important feature of these receptors is the presence of allosteric binding pockets, which is progressively gaining momentum as a novel target for current therapeutic strategies. We hypothesized that targeting an allosteric site could potentially allow for receptor functional selectivity, helping to reduce the overall side effects profile. One site, in particular, known as the Na+ binding pocket, is known to play a key role in GPCR function and is thought to be the center of the functional mechanism, with several bias switches characterized in the pocket residues (Katritch et al., Trends Biochem Sci. 2014).Here we report structure‐based rational drug design of bitopic ligands targeting the Na+ pocket of mu (MOR) and kappa opioid receptors (KOR). Our designed bitopics show high affinity, subtype selectivity, G‐protein bias and analgesic actions in mice. Mutations in Na+ binding pocket diminished functional potency validating allosteric site targeting. Together, results suggest the possibility of achieving subtype selectivity and identification of an additional subpocket in both MOR and KOR to reduce βarrestin‐2 recruitment by targeting an allosteric pocket in the opioid receptors. The highly conserved constitution of the Na+ binding pocket opens the opportunity for rational discovery of new GPCR modulators with desired functional, and potentially therapeutic profile for opioids as well as other class A GPCRs.Support or Funding InformationWe gratefully acknowledge the financial support from Faculty Research Incentive Fund, NIDA, NIAAA and US department of defense. We also would like to thank the Philippe Foundation for their financial support.

A great number of protein-binding peptides are known and utilized as drugs, diagnostic reagents, and affinity ligands. Recently, however, peptide mimetics have been proposed as valuable alternative to peptides by virtue of their excellent biorecognition activity and higher biochemical stability. This poses the need to develop a strategy for translating known protein-binding peptides into peptoid analogues with comparable or better affinity. This work proposes a route for translation utilizing the IgG-binding peptide HWRGWV as reference sequence. An ensemble of peptoid analogues of HWRGWV were produced by adjusting the number and sequence arrangement of residues containing functional groups that resemble both natural and non-natural amino acids. The variants were initially screened via IgG binding tests in non-competitive mode to select candidate ligands. A set of selected peptoids were studied in silico by docking onto putative binding sites identified on the crystal structures of human IgG(1), IgG(2), IgG(3), and IgG(4) subclasses, returning values of predicted binding energy that aligned well with the binding data. Selected peptoids PL-16 and PL-22 were further characterized by binding isotherm analysis to determine maximum capacity (Q(max) (4) over tilde8-57 mg of IgG per mL of adsorbent) and binding strength on solid phase (K-D (5) over tilde .4-7.8 10(-7) M). Adsorbents PL-16-Workbeads and PL-22-Workbeads were used for purifying human IgG from a cell culture supernatant added with bovine serum, affording high values of IgG recovery (up to 85%) and purity (up to 98%) under optimized binding and elution conditions. Both peptoid ligands also proved to be stable against proteolytic enzymes and strong alkaline agents. Collectively, these studies form a method guiding the design of peptoid variants of cognate peptide ligands, and help addressing the challenges that, despite the structural similarity, the peptide-to-peptoid translation presents. (C) 2019 Elsevier B.V. All rights reserved.

alpha-PVP (alpha-pyrrolidinovalerophenone) and MDPV (3,4-methylenedioxypyrovalerone) are potent abused stimulants that are members of the synthetic cathinone class of drugs. Although these drugs are taken with recreational intent, high doses can lead to unintended adverse effects including agitation, cardiovascular effects, sympathomimetic syndromes, hallucinations, and psychoses. One possible treatment is the use of a vaccine to block or attenuate adverse medical effects. These studies report the preparation of a vaccine that generates high affinity antibodies specific for both drugs and the pharmacological testing of this vaccine in male rats. Alkylation of a hydroxy-alpha-PVP analog with an appropriate thiol-bearing linker afforded the hapten. When hapten-conjugated carrier protein was mixed with adjuvant, the resulting vaccine stimulated production of antibodies in male Sprague Dawley rats that were found to significantly reduce alpha-PVP- and MDPV-induced hyperlocomotion as well as to significantly reduce the concentrations of MDPV drugs in critical organs. The novel vaccine produced high affinity antibodies against MDPV, (R)-MDPV, (S)-MDPV, and alpha-PVP. Cross-reactivity testing against nine structurally similar cathinones showed very limited binding, and no binding to off-target endogenous and exogenous compounds. Antibodies generated by this bi-specific vaccine also significantly shortened the duration of locomotor activity induced by both drugs up to a dose of 5.6 mg/kg in male rats. (C) 2019 Elsevier Ltd. All rights reserved.

Alpha‐pyrrolidinopentiophenone (alpha‐PVP) is a psychoactive synthetic cathinone that elicits pharmacological effects similar to methamphetamine and cocaine and at high doses can produce psychosis, cardiovascular toxicity, and death. The goal for these studies was to characterize the pharmacokinetics (PK) of alpha‐PVP and test an active vaccine for treating alpha‐PVP‐induced effects. The hapten‐protein conjugate vaccine was developed from an alpha‐PVP/MDPV‐like hapten structure capable of generating high affinity antibodies that significantly cross react with both alpha‐PVP and 3,4‐methylenedioxypyrovalerone (MDPV). Previous studies showed the bi‐specific vaccine could significantly reduce locomotor effects of MDPV. For the current studies, we hypothesized that treatment with this vaccine could stimulate production of high affinity antibodies against alpha‐PVP capable of altering the PK disposition and attenuating the cardiovascular and locomotor effects of alpha‐PVP. The hapten was conjugated to immunocyanin monomers of keyhole limpet hemocyanin, mixed with Sigma Adjuvant System, and used to immunize male Sprague Dawley rats (n=6/group) at 0, 3, and 9 weeks. Pharmacological testing occurred between 12–26 weeks from the start of immunizations. For the PK studies, three doses of alpha‐PVP (0.56, 1, and 3 mg/kg) were administered subcutaneously (sc) to control and vaccinated rats (n=6/group/dose); and tail vein blood samples were collected over time. PK values of alpha‐PVP showed linear PK with a half‐life of 1.7 h in controls and a significant (p<0.05) treatment‐dependent difference in the area under the alpha‐PVP concentration time curve between control and vaccinated rats. In a separate group of rats (n=5–6/group), cardiovascular parameters recorded by radiotelemetry showed a significantly lower area under the heart rate and blood pressure time curves after 0.56, 1 and 3 mg/kg alpha‐PVP in vaccinated rats compared to controls. Finally, locomotor studies of control and vaccinated rats (n=6/group) were conducted using a series of increasing doses of alpha‐PVP (0.3–5.6 mg/kg, sc, 2 days apart). The immunized rats displayed significantly lower alpha‐PVP‐induced locomotor activity and a significantly shorter duration of action compared to controls. In conclusion, alpha‐PVP has linear PK in control rats, and active vaccination with a bi‐specific conjugate vaccine significantly increased the binding of alpha‐PVP in the serum and reduced alpha‐PVP‐induced cardiovascular and locomotor effects compared to controls.Support or Funding InformationFunded by NIDA grants DA039195 and F31 DA046121, and T32 GM106999This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

Synthetic cathinones like 3,4‐methylenedioxypyrovalerone ((R,S)‐MDPV) are medically dangerous psychoactive drugs which at high doses produce psychosis, cardiovascular toxicity, and death. The goal for these studies was to develop and test a vaccine for treating ((R,S)‐MDPV) substance use disorders. A racemic MDPV conjugate vaccine was developed from a hapten structure capable of generating antibodies that bind with high affinity to (R,S)‐MDPV and other potent synthetic cathinones. We hypothesized that treatment with an optimized MDPV vaccine could produce high affinity antibodies against (R,S)‐MDPV capable of mitigating the pharmacological effects of (R,S)‐MDPV. An (R,S)‐MDPV‐like hapten was conjugated to different carrier proteins, mixed with Sigma Adjuvant System, and used to immunize male Sprague Dawley rats (n=5/group). The immunochemical characteristics of the rats' immune response was then evaluated. The KD value as determined by radioimmunoassay for (R,S)‐MDPV binding was 2.5 ± 1.9 nM, which demonstrated production of high affinity antibodies. The KD value for the individual (R)‐ and (S)‐MDPV binding was 0.71 ± 0.8 nM and 2.8 ± 2.3 nM, respectively. These similar KD values for the MDPV enantiomers indicates antibodies for both enantiomers contribute to the high binding of (R,S)‐MDPV. The rat antisera also showed similar high affinity binding of (R,S)‐α‐pyrrolidinovalerophenone (α‐PVP). No significant cross reactivity was detected with over‐the‐counter medications like (+)‐pseudoephedrine, neurotransmitters (dopamine, norepinephrine, epinephrine, and serotonin), or non‐cathinone drugs of abuse (cocaine, morphine, phencyclidine, (+)‐methamphetamine, (+)‐MDMA, and (−)‐MDMA). Moreover, the vaccination protocol produced high serum antibody titers, determined by rapid equilibrium dialysis, from 5 weeks after the start of immunizations until termination of the rats 3 months later. Finally, locomotor studies of control and vaccinated rats were conducted to test the vaccine's efficacy using a series of increasing doses of (R,S)‐MDPV (0.3–5.6 mg/kg, sc, 2 days apart). The immunized rats displayed significantly (p<0.05) lower MDPV‐induced locomotor activity and a significantly shorter duration of action compared to controls. In conclusion, active vaccination with an (R,S)‐MDPV conjugate vaccine produced very high affinity antibodies against (R,S)‐MDPV; and the vaccinated rats showed significantly reduced (R,S)‐MDPV‐induced locomotor effects compared to controls.Support or Funding InformationFunded by NIDA grants DA039195 and T32 GM106999This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

Past studies have shown that it has been difficult to discover and develop potent and selective κ opioid receptor antagonists, particularly compounds having potential for clinical development. In this study, we present a structure-activity relationship (SAR) study of a recently discovered new class of tetrahydroisoquinoline κ opioid receptor antagonists which led to (3 R)-7-hydroxy- N-{(1 S)-2-methyl-1-[(-4-methylpiperidine-1-yl)methyl]propyl}-1,2,3,4-tetrahydroisoquinoline-3-carboxamide (12) (4-Me-PDTic). Compound 12 had a Ke = 0.37 nM in a [35S]GTPγS binding assay and was 645- and >8100-fold selective for the κ relative to the μ and δ opioid receptors, respectively. Calculated log BB and CNS (central nervous system) multiparameter optimization (MPO) and low molecular weight values all predict that 12 will penetrate the brain, and pharmacokinetic studies in rats show that 12 does indeed penetrate the brain.

Animal pharmacological studies suggest that potent and selective κ opioid receptor antagonists have potential as pharmacotherapies targeting depression, anxiety, and substance abuse (opiates, alcohol, nicotine, cocaine). We recently reported lead compound 1 as a new class of κ opioid receptor antagonists with only one basic amine group. Analogues were synthesized and evaluated for their in vitro opioid receptor antagonist properties using a [35S]GTPγS binding assay. All analogues were pure opioid receptor antagonists with no agonist activity. Compounds 1, 8, 9, 13, and 14 ( Ke values 0.058-0.64 nM) are highly potent and highly selective for the κ relative to the μ and δ opioid receptors. Favorable calculated physiochemical properties were confirmed in rat PK studies, demonstrating brain penetration for selected compounds 1, 9, and 13. High κ opioid receptor potency and selectivity and highly favorable calculated physiochemical and PK properties for brain penetration suggest these compounds should be considered for further development.

Potent and selective κ opioid receptor antagonists have been derived from the N-substituted trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidine class of pure opioid receptor antagonists. In order to determine if the 3-hydroxyphenyl and/or the piperidine amino groups are required for obtaining the pure opioid antagonists, (3R)-7-hydroxy-N-[(1S)-2-methyl-1-(piperidine-1-ylmethyl)propyl]-1,2,3,4-tetrahydroiosquinoline-3-carboxamide (1), which does not have a 4-(3-hydroxyphenyl) group, and (3R)-N-(1R)-1-(cyclohexylmethyl)-2-methylpropyl]-7-hydroxy-1,2,3,4-tetrahydroisoquinoline-3-carboxamide (2), which does not have a 4-hydroxylphenyl or a piperidine amino group, were synthesized and evaluated for their [35S]GTPγS binding properties at the μ, δ, and κ opioid receptors. Surprisingly compound 1 remained a pure opioid antagonist with a Ke = 6.80 nM at the κ opioid receptor and is 21- and 441-fold selective for the κ receptor relative to the μ and δ opioid receptors, respectively. Even more unexpected and novel is the finding that 2 has a Ke = 0.14 nM at κ and is 1730- and 4570-fold selective for κ relative to the μ and δ opioid receptors, respectively.

The potent and selective KOR antagonist JDTic was derived from the N-substituted trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidine class of pure opioid antagonists. In previous studies we reported that compounds that did not have a hydroxyl on the 3-hydroxyphenyl group and did not have methyl groups at the 3- and 4-position of the piperidine ring were still potent and selective KOR antagonists. In this study we report JDTic analogs 2, 3a-b, 4a-b, and 5, where the 3-hydroxyphenyl ring has been replaced by a 2-, 3-, or 4-pyridyl or 3-thienyl group and do not have the 3-methyl or 3,4-dimethyl groups, remain potent and selective KOR antagonists. Of these, (3R)-7-hydroxy-N-(1S)-2-methyl-[4-methyl-4-pyridine-3-yl-carboxamide (3b) had the best overall binding potency and selectivity in a [(35)S]GTPγS functional assay, with a Ke=0.18nM at the KOR and 273- and 16,700-fold selectivity for the KOR relative to the MOR and DOR, respectively. Calculated physiochemical properties for 3b suggest that it will cross the blood-brain barrier.

In order to gain additional information concerning the active conformation of the N-substituted trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidine (1) class of opioid receptor antagonists, procedures were developed for the synthesis of structurally rigid N-substituted-6-(3-hydroxyphenyl)3-azabicyclo[3.1.0]hexane and 3-methyl-4-(3-hydroxyphenyl)-4-azabicyclo[4.1.0]heptanes. Evaluation of the conformationally constrained series in a [35S]GTPγS assay showed that structural rigid compounds having the 3-hydroxyphenyl group locked in the piperidine equatorial orientation had potencies equal to or better than similar compounds having more flexible structures similar to 1. The studies of the rigid compounds also suggested that the 3-methyl group present in compound 1 type antagonists may not be necessary for their pure opioid antagonist properties.

The design and discovery of JDTic as a potent and selective kappa opioid receptor antagonist used the N-substituted trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidine pharmacophore as the lead structure. In order to determine if the 3-methyl or 4-methyl groups were necessary in JDTic and JDTic analogs for antagonistic activity, compounds 4a-c, and 4d-f which have either the 3-methyl or both the 3- and 4-methyl groups removed, respectively, from JDTic and analogs were synthesized and evaluated for their in vitro opioid receptor antagonist activities using a [(35)S]GTPγS binding assay. Other ADME properties were also assessed for selected compounds. These studies demonstrated that neither the 3-methyl or 3,4-dimethyl groups present in JDTic and analogs are required to produce potent and selective κ opioid receptor antagonists.

N-substituted trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidines (2a,b) are opioid receptor antagonists where the antagonist properties are not due to the type of N-substituent. In order to gain a better understanding of the contribution that the 3- and 4-methyl groups make to the pure antagonist properties of 2a,b, we synthesized analogues of 2a,b that lacked the 4-methyl (5a,b), 3-methyl (6a,b), and both the 3- and 4-methyl group (7a,b) and compared their opioid receptor properties. We found that (1) all N-methyl and N-phenylpropyl substituted compounds were nonselective opioid antagonists (2) all N-phenylpropyl analogues were more potent than their N-methyl counterparts, and (3) compounds 2a,b which have both a 3- and 4-methyl substituent, were more potent antagonists than analogues 5a,b, 6a,b, and 7a,b. We also found that the removal of 3-methyl substituent of N-methyl and N-phenylpropyl 3-methyl-4-(3-hydroxyphenyl)piperazines (8a,b) gives (4a,b), which are opioid antagonists.

JDTic analogues 4-15 which have the hydroxyl groups replaced with other groups were synthesized and their in vitro efficacy at the μ, δ, and κ opioid receptors determined and compared to JDTic using [(35)S]GTPγS assays. Compounds 4, 5, 6, 13, 14, and 15 had Ke = 0.024, 0.01, 0.039, 0.02, 0.11, and 0.041 nM compared to the Ke = 0.02 nM for JDTic at the κ receptor and were highly selective for the κ receptor relative to the μ and δ opioid receptors. Unexpectedly, replacement of the 3-hydroxyl substituent of the 4-(3-hydroxyphenyl) group of JDTic with a H, F, or Cl substituent leads to potent and selective KOR antagonists. In vitro studies to determine various ADME properties combined with calculated TPSA, clogP, and logBB values suggests that the potent and selective κ opioid receptors 4, 5, 13, and 14 deserve consideration for further development toward potential drugs for CNS disorders.

There is continuing interest in the discovery and development of new κ opioid receptor antagonists. We recently reported that N-substituted 3-methyl-4-(3-hydroxyphenyl)piperazines were a new class of opioid receptor antagonists. In this study, we report the syntheses of two piperazine JDTic-like analogues. Evaluation of the two compounds in an in vitro [(35)S]GTPγS binding assay showed that neither compound showed the high potency and κ opioid receptor selectivity of JDTic. A library of compounds using the core scaffold 21 was synthesized and tested for their ability to inhibit [(35)S]GTPγS binding stimulated by the selective κ opioid agonist U69,593. These studies led to N-[(1S)-1-{[(3S)-4-(3-hydroxyphenyl)-3-methylpiperazin-1-yl]methyl}-2-methylpropyl]-4-phenoxybenzamide (11a), a compound that showed good κ opioid receptor antagonist properties. An SAR study based on 11a provided 28 novel analogues. Evaluation of these 28 compounds in the [(35)S]GTPγS binding assay showed that several of the analogues were potent and selective κ opioid receptor antagonists.

A compound represented by the formula: ** ** Formula wherein W, X, Y and Z are each independently hydrogen, C1-6 alkyl, C2-6 alkenyl, C2-6 alkynyl, OC 1-6 alkyl, OH, F, Cl, Br, CN, CF3, NO2, N3, SO2CH3, SO2CF3, SO2NH2, NR31R32, NHCOR33, NHCO2R34, CONR35R36, CH2 (CH2) NY2, CH2O2-C1-8 alkyl, C1-8 or C CO2-alkyl ( > = NH) NR37R38; R31, R32, R33, R34, R35, R36, R37 and R38 are independently hydrogen, C1-6 alkyl, C2-6 alkenyl, C2-6 alkynyl, CH2-aryl substituted with one more substituents OH, Br, Cl, F, CN, CF3, NO2, N3, SO2CH3, SO2CF3, SO2NH2, C1-6 alkyl or CH2 (CH2) NY2 'alkyl; each Y2 is independently hydrogen, CF3, CO2R9, C1-8 alkyl, NR10R11, NHCOR12, NHCO2R12, CONR13R14, CH2OH, CH2OR8, COCH2R9, ** ** Formula each n is independently 0, 1, 2 or 3; each R8, R9, R10, R11, R12, R13 and R14 is independently hydrogen, C1-8 alkyl, CH2-aryl wherein the aryl group is substituted with one or more substituents OH, Br, Cl, F, CN, CF3 , NO2, N3, SO2CH3, SO2CF3, SO2NH2, C1-6 alkyl or CH2 (CH2) NY2 '; each Y1 is independently hydrogen, OH, Br, Cl, F, CN, CF3, NO2, N3, SO2CH3, SO2CF3, SO2NH2, OR8, CO2R9, C1-6 alkyl, NR10R11, NHCOR12, NHCO2R12, CONR13R14, or CH2 (CH2 ) NY2, or two adjacent groups form a -O-Y1-O- or -O CH2-CH2CH2-O- group; each Y2 'is independently hydrogen, CF3 or C1-6 alkyl; or a pharmaceutically acceptable salt thereof.