Anti-prion Butenolides and Diphenylpropanones from the Australian

5 hours ago - Compounds 1–6 were tested for anti-prion activity in a yeast-based ... on the ACS Publications website at DOI: 10.1021/acs.jnatprod.9b...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/jnp

Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX

Anti-prion Butenolides and Diphenylpropanones from the Australian Ascidian Polycarpa procera Laurence K. Jennings,†,‡ Luke P. Robertson,†,‡ Kathryn E. Rudolph,‡ Alan L. Munn,⊥ and Anthony R. Carroll*,†,‡,§

Downloaded via UNIV OF SASKATCHEWAN on August 22, 2019 at 19:41:09 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Environmental Futures Research Institute, Griffith University (Gold Coast campus), Parklands Drive, Southport, QLD 4222, Australia ‡ School of Environment and Science, Griffith University (Gold Coast campus), Parklands Drive, Southport, QLD 4222, Australia § Griffith Institute for Drug Discovery, Griffith University (Brisbane Innovation Park), Don Young Road, Nathan, QLD 4111, Australia ⊥ Menzies Health Institute Queensland, Griffith University (Gold Coast campus), Parklands Drive, Southport, QLD 4222, Australia S Supporting Information *

ABSTRACT: A library of 500 Australian marine invertebrate extracts was screened for anti-prion activity using a yeast-based assay, and this resulted in an extract from the ascidian Polycarpa procera showing potent activity. Purification of this extract led to the isolation of six new butenolide metabolites, the procerolides 1−4 and two related diphenylpropanones, the procerones 5 and 6, as the bioactive components. The structures of 1−6 were elucidated from the analysis of 1D/2D NMR and MS data, and their absolute configurations determined from comparison of experimental and computed ECD data. Compounds 1−6 were tested for anti-prion activity in a yeastbased assay, and 1 and 5 displayed potent bioactivity (EC50 of 23 and 29 μM, respectively) comparable to the potently active anti-prion compound guanabenz. The procerolides and procerones are the first anti-prion compounds to be reported from ascidians, indicating that ascidians may be an untapped source of new lead anti-prion compounds.

P

isolation, structure identification, and anti-prion bioactivity of a series of brominated butenolides and diphenylpropanones from P. procera named the proceolides and procerones, respectively.

rion diseases are fatal, neurodegenerative disorders caused by misfolded proteins. The human prion protein forms amyloid plaques that build up in the brain and cause cell damage leading to neurodegeneration.1 In the United Kingdom in the 1980s an outbreak of the prion disease variant Creutzfeldt-Jakob disease (vCJD) occurred. Patients contracted the disease from eating tainted meat obtained from cattle suffering from bovine spongiform encephalopathy (BSE). With no treatment options for this disease the outbreak prompted an extensive search for anti-prion therapeutics. However, to date, no effective treatment has been developed for this fatal disease.2 Over the last 60 years the marine environment has been a prolific source of bioactive compounds.3 Marine ascidians in particular have been shown to contain unique chemical diversity, and some ascidian-derived compounds have been developed as promising leads for the design and development of new drugs.4 However, neither ascidian extracts nor their natural products have been evaluated for anti-prion activity. We have developed a microplate high-throughput yeast-based anti-prion assay and used it to screen a library of 500 Australian subtropical marine macro-organism extracts. This led to the identification of an anti-prion active extract from the ascidian Polycarpa procera.5 Herein, we report the screening, © XXXX American Chemical Society and American Society of Pharmacognosy

■

RESULTS AND DISCUSSION

A freeze-dried sample of P. procera was exhaustively extracted with MeOH. The LC-MS trace of the extract displayed a number of peaks possessing brominated ion clusters. The extract was separated by repeated preparative HPLC on C18 silica gel eluting with varying H2O to MeOH gradients. This yielded procerolides A, C, and D (1, 3, and 4) and procerone A (5). Extraction of another collection of P. procera and successive HPLC yielded procerolide B (2) and procerone B (6). Procerolide A (1), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 544.8243/546.8227/548.8209/ 550.8189 (1:3:3:1) in the (−) HRESIMS data, consistent with a molecular formula of C18H13Br3O5. The 1H NMR data (Table 1) contained resonances associated with five aromatic protons (δH 7.84−6.96), one oxygenated methine (δH 5.80), Received: June 16, 2019

A

DOI: 10.1021/acs.jnatprod.9b00551 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

correlations from δH 6.96 (H-5″) to δC 128.8 and 110.0 (C1″ and C-3″, respectively) and from δH 7.17 (H-2″) and 6.96 (H-6″) to δC 154.3 and 112.2 (C-4″ and C-5″, respectively). The HMBC correlation from δH 3.78 (C-4″ O-methyl) to δC 154.3 (C-4″) and attachment of a bromine atom to the nonprotonated carbon C-3″ allowed the assignment of the aromatic substituents. The methylene protons at δH 2.79/3.14 (δC 37.6) correlated to the furanone methine proton, δH 5.80 (H-5), in the COSY spectrum, while the aromatic protons at δH 7.17 and 6.96 (H-2″ and H-6″) showed an additional HMBC correlation to the aliphatic methylene carbon at δC 37.5 (C-6). This indicated that the 3-bromo-4-methoxybenzyl moiety (ring C) is attached to C-5 of the furanone ring. With the 2D structure of 1 established, the absolute configuration of the stereogenic center, C-5, was determined through comparison of experimental and predicted electronic circular dichroism (ECD) spectra. ECD spectra of the two possible enantiomers of 1 were calculated using time-dependent density functional theory (TDDFT) using the B3LYP/6-31G(d)// B3LYP/6-31G(d) functional/basis set combination. The calculated ECD spectrum of the (5S)-1 enantiomer matched with the experimental spectrum (Figure 1b). Therefore, procerolide A was assigned as (S)-(+)-4-(3,5-dibromo-4hydroxyphenyl)-5-(3-bromo-4-methoxybenzyl)-3-hydroxyfuran-2(5H)-one (1). Procerolide B (2), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 622.7337/624.7314/626.7308/ 628.7299/630.7302 (1:4:6:4:1) in the (−) HRESIMS data, consistent with a molecular formula of C18 H12 Br4 O5 . Comparison of 1H NMR and 13C NMR data between 1 and 2 indicated that ring C was tetrasubstituted in 2 instead of trisubstituted in 1. This was confirmed by an aromatic singlet at δH 7.42 that integrated for two protons, indicating symmetry in the aromatic ring. This, in addition to MS data, indicated an additional bromine was attached to C-5″ in ring C. The

one O-methyl (δH 3.78), and two geminal coupled aliphatic protons (δH 3.14, 2.79). The 13C NMR and HSQC data indicated the presence of 10 nonprotonated carbons (δC 168.5, 154.3, 151.0, 138.4, 128.8, 126.6, 125.1, 112.3 (2C), 110.0), five aromatic methines (δC 133.9, 131.0 (2C), 130.3, 112.2), an oxygenated methine (δC 77.4), an O-methyl (δC 56.2), and one aliphatic carbon (δC 37.6). HMBC correlations from the methine at δH 5.80 (H-5) to carbons at δC 168.5 (C-2), 138.4 (C-3), and 126.6 (C-4) allowed a 3,4,5-trisubstituted furan2(5H)-one to be assigned (ring A). The deshielded resonance at δC 138.4 (C-3) of the furanone ring indicated it was hydroxy substituted. A symmetrical 1,3,4,5-tetrasubstituted aromatic ring was identified from HMBC correlations from the 2H singlet at δH 7.84 (H-2′/6′). The deshielded resonance of C-4′ (δC 151.0) suggested that it was hydroxylated, and the chemical shift of the nonprotonated carbons, C-3′/5′ (δC 112.3), was consistent with bromine substitution.6 An HMBC correlation from δH 7.84 (H-2′/6′) to δC 126.6 (C4) indicated that the 3,5-dibromo-4-hydroxyphenyl moiety (ring B) is attached to C-4 of the furanone. A 1,3,4trisubstituted aromatic ring was identified from HMBC

Table 1. NMR Spectroscopic Data for Procerolides A−D (1−4) in DMSO-d6a procerolide A (1) position

δC, type

δH (J in Hz)

procerolide B (2) δC, type

2 3 3-OR2 4 5 6

168.5, C 138.4, C

168.4, C 138.3, C

126.6, C 77.4, CH 37.6, CH2

126.9, C 77.2, CH 38.0, CH2

1′ 2′ 3′ 4′ 5′ 6′ 1″ 2″ 3″ 4″ 4″-OMe 5″ 6″

125.1, C 131.0, CH 112.3, C 151.0, C 112.3, C 131.0 CH 128.8, C 133.9, CH 110.0, C 154.3, C 56.2, CH3 112.2, CH 130.3, CH

a1

5.80, dd (5.9, 4.0) 2.79, dd (14.7, 6.0) 3.14, dd (14.7, 4.0) 7.84, s

7.84, s 7.17, d (1.3)

3.78, s 6.96, s 6.96, d (1.3)

125.0, C 131.0, CH 112.2, C 150.9, C 112.2, C 131.0, CH 135.4, C 133.9, CH 116.8, C 152.1, C 60.4, CH3 116.8, C 133.9, CH

procerolide C (3)

δH (J in Hz)

5.84, dd (6.1, 4.0) 2.78, dd (14.7, 6.1) 3.15, dd (14.7, 4.0)

δC, type 167.0, C 139.9, C 58.4, CH3 135.8, C 77.0, CH 37.4, CH2 123.7, C 131.5, CH 112.1, C 151.9, C 112.1, C 131.5, CH 128.5, C 133.8, CH 109.9, C 154.2, C 56.2, CH3 112.1, CH 130.2, CH

7.86, s

7.86, s 7.42, s

3.77, s 7.42, s

δH (J in Hz)

3.84, s 5.88, dd (6.3, 3.9) 2.79, dd (14.6, 6.3) 3.11, dd (14.6, 3.9) 7.80, s

7.80, s 7.18, d (1.7)

3.80, s 6.99, s 6.98, d (1.6)

procerolide D (4) δC, type 167.1, C 140.0, C 58.7, CH3 136.3, C 77.0, CH 37.9, CH2 123.7, C 131.8, CH 112.2, C 152.1, C 112.2, C 131.8, CH 134.9, C 134.1, CH 117.0, C 152.3, C 60.6, CH3 117.0, C 134.1, CH

δH (J in Hz)

3.77, s 5.91, dd (7.3, 4.4) 2.84, dd (14.4, 7.3) 3.09, dd (14.4, 4.4) 7.80, s

7.80, s 7.40, s

3.89, s 7.40, s

H NMR was measured at 500 MHz for 1, 3, and 4 and 800 MHz for 2. 13C NMR was measured at 125 MHz for 1, 3, and 4 and at 200 MHz for

2. B

DOI: 10.1021/acs.jnatprod.9b00551 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

indicated the O-Me was attached to C-3 in ring A. Similar to the previous compounds the positive optical rotation and ECD matching of that observed for 1 (Figures S4−7) allowed the assignment of a 5S configuration. Therefore, procerolide D was assigned as (S)-(+)-4-(3,5-dibromo-4-hydroxyphenyl)-5-(3,5dibromo-4-methoxybenzyl)-3-methoxyfuran-2(5H)-one (4). The procerolides are butenolides related to the previously reported ascidian metabolites rubrolides,7 cadiolides,8 and prunolides.6 The procerolides differ from the previous compounds in that they do not have an additional degree of unsaturation at C-5 in the furanone ring. The procerolides also differ in the substitution of C-3 in the furanone ring. The substitution at C-3 of a hydroxy or O-methyl has not been identified in related compounds from ascidians. Ascidianderived butenolides have been found to possess a wide range of bioactivities including antibacterial,9,10 antiviral,11 and antiinflammatory activity,12 photosynthesis inhibition,13 and cancer cell cytotoxicity.14 Similar butenolide metabolites with the additional level of saturation at C-5 and hydroxy substitution at C-3 have also been identified from the fungus Aspergillus terreus.15 These compounds were found to inhibit cyclin-dependent kinases, which have important functions in a number of diseases including cancer, Alzheimer’s, and Parkinson’s diseases.16 Procerone A (5), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 504.8260/506.8259/508.8240/ 510.8198 (1:3:3:1) in the (−) HRESIMS data, consistent with a molecular formula of C16H13Br3O4. The 1H NMR data for 1 and 5 (Table 2) were similar; however, comparison of 13C NMR and HSQC data for 1 and 5 showed that 5 has two fewer nonprotonated olefinic carbon resonances and the ester carbonyl (δC 168.3) is replaced with a ketone carbonyl (δC 197.2). Comparison of 2D NMR data between 1 and 5

Figure 1. (a) Key COSY and HMBC correlations for 1. (b) Comparison of the experimental ECD spectrum of 1 and the calculated spectrum of the (S)- and (R)-enantiomers of 1 at the B3LYP/6-31G(d) level.

absolute configuration of the C-5 stereogenic center was assigned through the comparison of optical rotation (OR) between the four procerolides. As all specific rotations were positive and 2 shares the same biosynthetic pathway, the C-5 stereogenic center was assigned in an S configuration. Therefore, procerolide B was assigned as (S)-(+)-4-(3,5dibromo-4-hydroxyphenyl)-5-(3,5-dibromo-4-methoxybenzyl)-3-hydroxyfuran-2(5H)-one (2). Procerolide C (3), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 558.8386/560.5372/562.8354/ 564.8328 (1:3:3:1) in the (−) HRESIMS data, consistent with a molecular formula of C19H15Br3O5. The 1H NMR and 13C NMR data for 3 contained signals for an additional O-Me group (δH 3.84, δC 58.4) compared to 1. An HMBC correlation between δH 3.84 and δC 139.9 (C-3) indicated the O-Me was attached to C-3 in ring A. The absolute configuration of the C-5 stereogenic center was assigned through the comparison of OR and ECD data between 1 and 3. A positive rotation and a similar ECD spectrum to 1 (Figures S3−7) indicated that the molecule had a 5S configuration. Therefore, procerolide C was assigned as (S)(+)-4-(3,5-dibromo-4-hydroxyphenyl)-5-(3-bromo-4-methoxybenzyl)-3-methoxyfuran-2(5H)-one (3). Procerolide D (4), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 636.7508/638.7500/640.7496/ 642.7462/644.7434 (1:4:6:4:1) in the (−) HRESIMS data, consistent with a molecular formula of C19H14Br4 O 5. Comparison of 1H NMR and 13C NMR data between 2 and 4 indicated an additional O-Me group (δH 3.77, δC 58.7). An HMBC correlation between δH 3.77 and δC 140.0 (C-3)

Table 2. NMR Spectroscopic Data for Procerones A and B (5 and 6) in DMSO-d6a procerone A (5)

procerone B (6)

δC, type

δH (J in Hz)

δC, type

1 2

197.2, C 73.2, CH

5.01, dd (8.2, 4.9)

197.0, C 72.5, CH

2-OH 3

38.3, CH2

position

1′ 2′ 3′ 4′ 4′−OH 5′ 6′ 1″ 2″ 3″ 4″ 4″-OMe 5″ 6″

128.0, C 133.2, CH 112.3, C 155.0, C

5.68, bs 2.77, dd (14.1, 8.2) 2.93, dd (14.1, 4.9) 8.11, s

38.0, CH2

129.3, 133.2, 111.5, 155.1,

C CH C C

δH (J in Hz) 5.07, dd (8.6, 4.6) 2.77, dd (14.0, 8.6) 2.93, dd (14.0, 4.6) 8.10, s

10.98, bs 112.3, C 133.2, CH 131.7, C 133.7, CH 110.1, C 153.9, C 56.2, CH3 111.5, CH 130.0, CH

8.11, s 7.45, d (2.2)

3.81, s 7.00, d (8.4) 7.18, dd (8.3, 2.2)

111.5, C 133.2, CH 137.8, C 133.9, CH 116.9, C 151.7, C 60.4, CH3 116.9, C 133.9, CH

8.10, s 7.52, s

3.75, s 7.52, s

a1 H NMR was measured at 500 MHz for 5 and 800 MHz for 6. 13C NMR was measured at 125 MHz for 5 and 200 MHz for 6.

C

DOI: 10.1021/acs.jnatprod.9b00551 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

bromo-4-methoxyphenyl)-1-(3,5-dibromo-4-hydroxyphenyl)2-hydroxypropanone. Procerone B (6), isolated as a pale yellow solid, had an ion cluster for [M − H]− at m/z 582.7405/584.7393/586.7378/ 588.7359/590.7337 (1:4:6:4:1) in the (−) HRESIMS data, consistent with a molecular formula of C16H12Br4O4. The 1H NMR data for 5 were almost identical to those reported for 6 except that a 2H aromatic singlet at δH 7.52 replaced the three ring C protons in 6. Ring C was therefore 1,3,4,5tetrasubstituted, and since the MS data indicated that 6 contains an additional bromine, C-5″ must be brominated. Therefore, 6 was assigned as (S)-(+)-1-(3,5-dibromo-4hydroxyphenyl)-3-(3,5-dibromo-4-methoxyphenyl)-2-hydroxypropanone. We hypothesize that the procerones are likely derived from the procerolides. The procerones contain aromatic rings B and C linked via a 2-hydroxypropanone moiety instead of the furanone moiety in the procerolides. The procerolides are most likely amino acid derived, through the enzymatic condensation of two alpha-keto acids derived from tyrosine.18 The decarboxylation and reduction at C-5 would then generate the procerolides (Figure 3). We then surmise that the

indicated that 5 also contained 3-bromo-4-methoxyphenyl (ring B) and 3,5-dibromo-4-hydroxyphenyl (ring C) moieties. COSY correlations between an oxygenated methine at δH 5.01 (H-2) and methylene protons at δH 2.77/2.93 (H2-3) indicated they are attached to vicinal carbons. HMBC correlations from δH 5.01 and 2.77/2.93 (H-2 and H2-3) to the carbonyl resonance at δC 197.2 (C-1) as well as the further deshielding of H-2 (δH 5.01) compared to typical oxygenated methine protons indicated the presence of the 2-hydroxypropanone moiety. HMBC correlations from δH 8.11 (H-2′ and H-6′) to δC 197.2 (C-1) and from δH 7.45 (H-2″) and δH 7.18 (H-6″) to δC 38.3 (C-3) indicated the attachment of rings B and C to the 2-hydroxypropanone. With the 2D structure of 5 established, the absolute configuration of the stereogenic center, C-2, was determined through comparison of experimental and predicted ECD spectra as described previously. The calculated ECD spectrum of the (2S)-5 enantiomer had the closest match to the experimental spectrum (Figure 2c).

Figure 3. Proposed biosynthetic pathway of the procerolides (1−4) from tyrosine derivatives and then to the procerones (5 and 6).

procerones are biosynthesized through the oxidative cleavage of the double bond19 in the furanone, resulting in a ketone and oxalate unit. These procerolides and procerones may then be precursors to a number of other larger butenolide and related metabolites. The procerolides and procerones exhibited significant spectroscopic differences dependent on the basicity of solutions. Under basic conditions the 1H NMR spectrum of procerolide A showed both the shielding and broadening of the resonances associated with H-2′/6′ and H-5 compared to that under acidic conditions (Figure 4a). Additionally, the 13C NMR resonances associated with C-2, C-3, C-4, C-1′, C-3′/5′, and C-4′ were not visible, and those associated with C-5, C-6, C-2′/6′, and C-1″ had significantly reduced intensities. In the UV spectrum of procerolide A a bathochromic shift (λmax 288 → 353) was observed under basic conditions (Figure 3c). Similarly, in the 1H NMR spectrum of procerone B the shielding and broadening of the resonances associated with H2′/6′ and H-5 (Figure 4b) along with a bathochromic shift

Figure 2. (a) Key COSY and HMBC correlations for 5. (b) Δδ values [Δδ (in ppm) = δS − δR] obtained for (S)- and (R)-MTPA esters of 5. (c) Comparison of the experimental ECD spectrum of 5 and the calculated spectrum of the (S)- and (R)-enantiomers of 5 at the B3LYP/6-31G(d) level.

This configuration was further supported from results obtained using the Mosher’s ester method.17 The esterification of the C2 hydroxy with the (R)- and (S)-MTPA acid chloride afforded the (S)- and (R)-MTPA esters of 5, 5S and 5R, respectively. Analysis of the 1H NMR and HSQC data for these MTPA esters of 5 indicated the deshielding of H-2′/6′ and the shielding of H-6″, H-5″, and H2-3 in 5S compared to 5R (Figure 2b). This indicated the position of the phenyl moiety of the MTPA in both 5S and 5R, allowing the assignment of a 2S configuration. Therefore, 5 was assigned as (S)-(+)-3-(3D

DOI: 10.1021/acs.jnatprod.9b00551 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

similar arrangement to that present in the procerolides and procerones. These include anle138b21 and N-benzylidenebenzohydrazide.22 Procerolides C and D (3 and 4) have significantly weaker anti-prion activity compared to 1, 2, 5, and 6. Procerolides C and D (3 and 4) are O-methylated at C-3 and therefore lack a hydrogen bond donor moiety between rings B and C. This hydrogen bond donor may be important for anti-prion activity. It is also important to note that previous studies, with similar butenolide metabolites, have suggested that the differences in antimicrobial activity are likely to reflect differences in a compound’s ability to permeate cell membranes.23 Since activity in the in vivo yeast-based assay used in this study is dependent on the ability of the compounds to pass through the cell membrane, the membrane permeability of the compounds can influence the potency of their anti-prion activity. The calculated LogP of the procerolides and procerones (Table 3) indicated that their permeability may influence their EC50’s. The results obtained in this study support the view that the procerolides and the procerones may help in the future design and development of novel lead compounds for neurodegenerative diseases.

■

Figure 4. Proposed keto−enol tautomerism within the procerolides (a) and the procerones (b) under basic conditions with Δδ values of protons [Δδ (in ppm) = δacidic − δbasic]. (c) Observed bathochromic shift that accompanied the tautomerism of both 1 and 5.

General Experimental Procedures. Optical rotations were measured using a JASCO P-1020 polarimeter. UV spectra were measured using a Shimadzu UV-1800 UV spectrophotometer. ECD spectra were measured using a JASCO J-715 spectropolarimeter. IR spectra were measured using a Thermo Scientific Nicolet iS5 iD5 ATR spectrometer. NMR spectra were recorded at 25 °C on a Bruker Avance III 500 MHz spectrometer (BBFO Smartprobe, 5 mm 31P109Ag) and a Bruker BioSpin GmbH 800 MHz spectrometer with a triple (TCl) resonance 5 mm cryoprobe. NMR solvent DMSO-d6 peak was referenced to δH 2.50 and δC 39.52. High-resolution ESITOF data were recorded on an Agilent 6530-accurate mass Q-TOF LC/MS mass spectrometer with a 1200 Series autosampler and 1290 Infinity HPLC. Altech Davisil 35−76 μm 60 Å C18 silica was used to adsorb samples prior to HPLC separation. A Merck Hitachi L7100 pump and an L7455 PDA detector were used for HPLC. HPLC columns used were a Thermo Betasil C18 5 μm, 100 Å, 150 mm × 21.2 mm and a Phenomenex EVO C18 5 μm, 100 Å, 150 mm × 21.2 mm. All solvents used for chromatography and MS were HPLC grade, and the H2O was Millipore Milli-Q PF filtered. Trifluoroacetic acid (TFA) was spectroscopy grade from Alfa Aesar. Sodium chloride (NaCl) was analytical grade from Merck. Analytical grade compounds guandine hydrochloride (GuHCl) and guanabenz were purchased from Sigma. The S. cerevisiae STRg6 strain was used to screen against the [PSI+] prion (MATa ade1-14 trp1-289 his3Δ200 ura3-52 leu23,112 erg6::TRP1 [PSI+]). The S. cerevisiae SB34 strain was used to screen against the [URE3] prion (MATa ade2-1 trp1-1 leu2-3,112 his3-11,15 ura2::HIS3 dal5::ADE2 [URE3]). Greiner sterile 96-well plates were used for screening. Animal Material. Samples of the ascidian Polycarpa procera were collected on June 21, 2011, by hand using scuba in the shallow waters off the coast of Coffs Harbour, NSW, Australia. This sample was freeze-dried and stored at room temperature. A voucher specimen (ACENV0263) is located at Griffith University, Gold Coast, Queensland, Australia. The ascidian was taxonomically identified by A.R.C. by morphological analysis. Extraction and Isolation. The freeze-dried sample of the ascidian P. procera (5.41 g) was exhaustively extracted with MeOH to yield an extract (0.2930 g). This extract was adsorbed onto the C18 silica gel at a 1:1 ratio and packed into the HPLC refillable cartridge (10 mm × 20 mm) that was connected in series with a Betasil C18 bonded silica HPLC column (21 mm × 150 mm). The columns were eluted with a gradient from 100% H2O containing 0.1% TFA to 100% MeOH containing 0.1% TFA over 60 min at a flow rate of 9 mL/min. The column was further eluted with 100% MeOH containing 0.1% TFA

correlated with a change in the basicity. These spectroscopic data indicate that the procerolides and the procerones undergo keto−enol tautomerism under basic conditions. We propose that this tautomerism in the procerolides is the interconversion of the phenol-hydroxyfuranone moiety with a dienonedihydroxyfuranylidene moiety (Figure 4a) and in the procerones the interconversion of the phenol-hydroxypropanone moiety with a dienone-dihydroxypropylidene moiety (Figure 4b). The procerolides and procerones were screened for antiprion activity using yeast containing either the [PSI+] or the [URE3] prions (Table 3).5 Guanabenz, a potent anti-prion Table 3. Anti-prion Activities of the Procerolides (1−4) and the Procerones (5 and 6) compound

ClogP

procerolide A (1) procerolide B (2) procerolide C (3) procerolide D (4) procerone A (5) procerone B (6) guanabenz

5.12 5.76 5.75 6.21 4.61 5.20 1.69

[PSI+] prion curing

[URE3] prion curing

EC50 ± SE (μM)

curing at 100 μM

± ± ± ± ± ± ±

yes yes yes yes yes yes yes

23 34 65 67 29 33 26

9 3 4 5 5 7 5

EXPERIMENTAL SECTION

compound, was used as a positive control for a comparison of activity. Guanabenz has been shown to cure yeast prions at concentrations of