Abstract
Hexacyclic congeners 2 and 3 of palau’amine, which incorporate both guanidine functional groups and have the cis configuration of the azabicyclo[3.3.0]octane core, are prepared in 14 steps from cycloadduct 6. Synthetic access to these analogs allows the first direct comparison of NMR data for hexacyclic diguanidine structures having the originally proposed cis-azabicyclo[3.3.0]octane fragment with data for natural alkaloids of the palau’amine family. This comparison provides convincing evidence in favor of the recently proposed structural revision of these marine alkaloids, fully supporting the trans configuration of the central azabicyclo[3.3.0]octane ring system of palau’amine and congeners.
Introduction
The stunning degree of structural diversity in the pyrrole-imidazole family of alkaloids provides an excellent example of biosynthetic efficiency.1 This large family of marine natural products is believed to be based on one simple metabolite, oroidin, and includes complex molecular architectures such as the hexacyclic oroidin dimer palau’amine (1, Figure 1) and alkaloids such as stylissadines A and B containing four oroidin subunits. Since its structural elucidation by mass spectral and NMR investigations in 1993,2 palau’amine has received much attention from the synthetic community because of its promising immunomodulatory activity, challenging molecular architecture, and density of functionality.3 However, in spite of considerable efforts, no report of a successful total synthesis has emerged to date.1
Figure 1.
Originally proposed structure of palau’amine (1), revised structure of palau’amine (2), synthetic derivatives 3 and 4, and structure of tetrabromostyloguanidine (5). Stereogenic centers with revised relative configurations have been highlighted with gray circles.
Recently, the isolation and structural elucidation of additional palau’amine congeners has challenged the initial assignment of the relative configuration of palau’amine, leading to the proposal that stereogenic centers C12 and C17 should be inverted (2, Figure 1).4 Most notably, the revised structure 2 contains a highly strained trans-azabicyclo[3.3.0]octane fragment. Herein, we describe the synthesis of hexacyclic palau’amine derivatives 3 and 4 possessing the originally proposed cis-azabicyclo[ 3.3.0]octane central unit. Comparison of these compounds to palau’amine and tetrabromostyloguanidine (5)4c,5 using NMR spectroscopic and computational methods provides additional evidence for the revised relative configuration 2 of palau’amine.
Results and Discussion
In our approach toward the putative original structure 1 of palau’amine, an intramolecular azomethine imine 1,3-dipolar cycloaddition forms the cis-azabicyclo[3.3.0]octane subunit having the correct relative configuration at the fully substituted carbons 10 and 16.6 Starting from this intermediate, elaboration to diguanidine hexacycles 3/4 requires installation of the two spiroguanidine units and formation of the ketopiperazine ring. Building on precedent from Büchi’s pioneering synthesis of dibromophakellin,7 ketopiperazine ring closure was projected to occur by attack of the nitrogen of a tethered bromopyrrole on a 2-methylsulfanyl-4H-imidazole generated in situ (Scheme 1). The second (right-hand) spirocycle was envisaged to arise in turn from an S-methyl thiohydantoin intermediate or by guanylation of the amino functional group of a suitable precursor, followed by spirocyclization.
Scheme 1.
Plan for the conversion of intermediate 6 to palau’amine derivatives 3 and 4.
An optimized synthetic route for the conversion of triazatriquinane 6 into hexacyclic palau’amine deriviatives 3/4, which have the same relative configuration at their stereogenic centers as the originally proposed structure 1 of palau’amine, is outlined in Scheme 2. This sequence began with SmI2-mediated reduction of the N-N bond of 6, followed by selective S-methylation of the resulting thiohydantoin 7 to furnish the corresponding S-methyl isothiourea 8 (76% overall). Subsequent selective Teoc protection of the isothiourea provided intermediate 9 in quantitative yield. Many attempts to elaborate aminoester 9 into a hexacyclic diguanidine via the intermediacy of bis(S-methylthiohydantoin) spirocycles established the viability of the synthetic strategy for ketopiperazine ring closure, however, the diguanidine functional group array could not be formed from these intermediates. Alternative strategies involving delayed installation of the second spirocycle suffered from unproductive intramolecular rearrangements, originating from nucleophilic attack of the primary amine onto the adjacent isothiourea. Alternatively, direct guanylation of the α-amino ester functionality of intermediates 8 and 9 to afford carbamate-protected glycocyamidine intermediates proved to be capricious and low-yielding. Therefore, a two-step procedure for glycocyamidine formation was developed, wherein aminoester 9 was first converted in high yield to Cbz-protected thiourea 10 by reaction with benzyloxycarbonyl isothiocyanate (11).8 Incubation of 10 with EDCI and o-nitrobenzylamine led to efficient installation of the protected guanidine functionality, presumably via the intermediacy of a carbamoylcarbodiimide, followed by in situ spirocyclization to afford glycocyamidine 12 in 93% overall yield.9,10
Scheme 2. Conversion of intermediate 6 into diastereomeric palau’amine derivatives 3 and 4.a.
a(a) SmI2, THF/MeOH, 23 °C, 15 min, 79%; (b) MeI, i-PrEt2N, DMAP, CH2Cl2, 23 °C, 2 h, 96%; (c) Teoc-Cl, i-PrEt2N, CH2Cl2, 23 °C, 30 min, quant.; (d) benzyloxycarbonyl isothiocyanate (11), CH2Cl2, 40 °C, 1.5 h, 92%; (e) EDCI, 2-nitrobenzylamine hydrochloride, i-PrEt2N, CH2Cl2, 40 °C, 2.5 h, 93%; (f) TFA, CH2Cl2, 23 °C, 1 h, then sat. aq. Na2CO3, 94%; (g) Teoc-Cl, i-PrEt2N, CH2Cl2, 23 °C, 30 min, 90%; (h) NaBH4, MeOH/THF (2:1), 0 °C, 40 min; (i) Ac2O, pyridine, DMAP, CH2Cl2, 23 °C, 4 h, 51% (15), 29% (16) (2 steps); (j) TBAF, THF, 23 °C, 8 min, 95%; (k) IBX, DMSO, 23 °C, 4 h, 96%; (l) NaBH4, MeOH, 0 °C, 30 min, 94%; (m) m-CPBA, CH2Cl2, 0→23 °C, 40 min; (n) NH3, CH2Cl2,−78→23 °C, 14 h, 79% (2 steps); (o) hv, dioxane, 23 °C, 2.5 h; (p) H2, Pd/C, dioxane/H2O (0.1% TFA), 43% (3), 22% (4) (2 steps). DMAP = N,N-dimethylaminopyridine, Teoc-Cl = 2-trimethylsilylethyl chloroformate, EDCI = 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride, TFA = trifluoroacetic acid, TBAF = tetra(n-butyl)ammonium fluoride, IBX = o-iodoxybenzoic acid, m-CPBA = m-chloroperoxybenzoic acid.
Glycocyamidine 12 was elaborated to palau’amine derivatives 3/4 in the following way. Chemoselective removal of the SEM protecting group in the presence of the secondary TBS ether was first accomplished by reaction of 12 with 10% TFA–CH2Cl2 at room temperature, followed by basic aqueous work-up to provide the free pyrrole 13. Reintroduction of the Teoc group furnished bromopyrrole bis(spirocycle) 14 in 85% overall yield. Reduction of the carbonyl groups of the spirocyclic fragments was achieved by exposure of bis(spirocycle) 14 to an excess of NaBH4 in MeOH/THF, providing a mixture of two inseparable bis(hemiaminal) diastereoisomers. In this transformation, the Teoc and o-nitrobenzyl groups impart sufficient electronic activation to the spirocycles to allow for smooth reduction of the carbonyl groups under mild conditions. The hemiaminal hydroxyl groups were then acetylated under standard conditions to give diastereomeric diacetates 15 and 16 in 80% combined yield and a ratio of 1.8:1.11 Although these diastereoisomers could be separated by standard chromatography, short exposure of either diastereoisomer, or the mixture, to TBAF/THF promoted cyclization to form the ketopiperazine ring to give an inseparable mixture of hemiaminal epimers 17 (ratio 1.5:1) in excellent yield (95%).12,13 The configuration of the secondary alcohol was next inverted by a two-step sequence. Chemoselective oxidation of alcohol 17 with IBX provided ketone 18 in high yield (96%) as an inseparable mixture of hemiaminal epimers (1.8:1). As expected, exposure of this ketone to NaBH4/MeOH resulted in hydride delivery from the convex face to give alcohol 19, again as an inseparable mixture of epimers (1:1, 94%). Treatment of this mixture with three equiv of purified m-chloroperoxybenzoic acid,14 followed by incubation of the sulfone products with ammonia provided diguanidine 20 in good overall yield (79% after HPLC purification) as a 5:1 mixture of inseparable hemiaminal epimers. To complete the synthesis of 3 and 4, the o-nitrobenzyl and Cbz protecting groups were cleaved by photolysis and hydrogenolysis, respectively. In the course of the latter reaction, the bromine substituents of the pyrrole were also removed to provide a 2:1 mixture of hemiaminal epimers 3 and 4 in 65% combined overall yield. Although this mixture of hemiaminal epimers could be separated by HPLC, interconversion of the two diastereoisomers in H2O/MeCN/TFA (89:11:0.1) and D2O thwarted their isolation as single stereoisomers.
Gratifyingly, most of the NMR resonances of diastereoisomers 3 and 4 in D2O were well separated and allowed for full assignment of 1H and 13C NMR spectra (Table 1). Comparison of 1H and 13C chemical shifts of 3 and 4 with those published for palau’amine2b reveals a surprisingly close overall match considering the dissimilarities in substituents at C17 and C18. An important difference, however, is the value of the 1H,1H coupling constant for the angular protons of the azabicyclo[3.3.0]octane moiety, which are 12.0 and 10.7 Hz for 3 and 4, respectively. This vicinal coupling constant is greater than 14 Hz for palau’amine and all palau’amine congeners, with the large value of this coupling constant for the natural alkaloids being an important piece of evidence in support of the revised trans fusion of the azabicyclo[3.3.0]octane unit.4c In addition, qualitative interpretation of the NOESY spectra of 3 and 4 revealed a strong correlation between the two protons at the ring junction (H11/H12), and the absence or weak correlations for the 1,3-related pairs of protons H11/H18 and H11/H13β, respectively. This latter data again stands in sharp contrast to data published for palau’amine and its congeners, in which a NOE correlation between H11 and H12 was often not mentioned,[15] and correlations for H11/H13β and H11/H18 were reported. As an additional point of divergence, correlations between H13α and H18 were observed for 3 and 4, whereas H18 was correlated to H13β in the natural products. No correlation between H13β and H18 was observed for 3 or 4.
Table 1.
1H and 13C NMR data for palau’amine,2b 3, and 4 in D2O.
| palau’amine | 3 | 4 | ||||
|---|---|---|---|---|---|---|
| carbon | 13C, ppm | 1H, ppm (mult, J [Hz]) | 13C, ppm | 1H, ppm (mult, J [Hz]) | 13C, ppm | 1H, ppm (mult, J [Hz]) |
| 2 | 122.5 | 120.5 | 120.3 | |||
| 3 | 115.6 | 6.85, (dd, J = 3.9, 1.5) | 115.0 | 6.80, (dd, J = 4.0, 1.3) | 115.0 | 6.79, (dd, J = 4.0, 1.3) |
| 4 | 113.8 | 6.35, (dd, J = 3.9, 2.8) | 113.0 | 6.30, (dd, J = 4.0, 2.7) | 113.2 | 6.29, (dd, J = 4.0, 2.7) |
| 5 | 125.2 | 6.99, (dd, J = 2.8, 1.5) | 124.7 | 6.96, (dd, J = 2.7, 1.3) | 124.5 | 6.93, (dd, J = 2.7, 1.3) |
| 6 | 69.0 | 6.33, (s) | 68.1 | 6.19, (s) | 68.7 | 6.07, (s) |
| 8 | 157.8 | 156.9 | 156.7 | |||
| 10 | 80.8 | 82.0 | 81.8 | |||
| 11 | 56.3 | 3.08, (d, J = 14.1) | 53.1 | 3.65, (d, J = 12.0) | 59.9 | 3.18, (d, J = 10.7) |
| 12 | 41.8 | 2.52, (dddd) | 45.3 | 2.90, (dddd, J = 12.0, 10.1, 6.1, 4.1) | 44.2 | 2.98, (dddd, J = 10.7, 9.9, 5.1, 4.6) |
| 13 | 46.1 | 3.96, (dd, J = 10.4, 7.3) | 42.2 | 3.99 (β) (dd, J = 12.3, 10.1) | 42.7 | 3.90 (β) (dd, J = 12.3, 9.9) |
| 3.28, (dd, J = 10.3, 10.4) | 3.48 (α) (dd, J = 12.3, 6.1) | 3.55 (α) (dd, J = 12.3, 4.6) | ||||
| 15 | 159.5 | 157.4 | 157.4 | |||
| 16 | 72.1 | 71.9 | 70.7 | |||
| 17 | 74.0 | 4.35, (d, J = 7.9) | 49.9 | 2.18 (β) (dd, J = 14.3, 3.7) | 41.2 | 2.62 (β) (dd, J = 15.4, 5.1) |
| 1.99 (α) (d, J = 14.3) | 1.91 (α) (dd, J = 15.4, 2.5) | |||||
| 18 | 48.6 | 2.47, (dddd) | 70.1 | 4.17, (m) | 69.2 | 4.17, (m) |
| 19 | 41.9 | 3.32, (dd, J = 13.2, 7.0) | - | - | - | - |
| 3.24, (dd, J = 13.2, 7.0) | ||||||
| 20 | 83.7 | 5.96, (s) | 85.8 | 5.17, (s) | 87.9 | 5.02, (s) |
| 22 | 157.9 | 158.7 | 158.7 | |||
For further corroboration of the qualitative interpretation of NOE correlations, a quantitative NOESY analysis was carried out to experimentally determine important interproton distances.16 Data gathered from these experiments were evaluated by comparison to the corresponding distances obtained from in silico geometry optimization, and to values published for tetrabromostyloguanidine (5).4c Geometry optimizations performed using the molecular modeling packages Maestro 5.0.019 with the AMBER* forcefield including solvation (H2O) and Spartan 04 (B3LYP 6-31G*) returned consistent results.17
A selected set of diagnostic interproton distances obtained from NOESY experiments and the corresponding calculated values are given in Table 2 (for further details see the Supporting Information).
Table 2.
Selected experimental and calculated interproton distances.a
| Protons | 3[a] | 3[b] | 4[a] | 4[b] | 1[b] | trans-3[b],[c] | trans-4[b],[c] | 5[d] | 5[b] | 2[b] |
|---|---|---|---|---|---|---|---|---|---|---|
| 11/12 | 221 | 221 | 228 | 224 | 226 | 304 | 304 | - | 304 | 303 |
| 11/13β | 335 | 342 | 334 | 343 | 407 | 268 | 267 | 250 | 270 | 263 |
| 11/18 | n.o. | 405 | n.o. | 407 | 327 | 253 | 258 | 233 | 253 | 277 |
| 11/20 | 270 | 317 | 208 | 221 | 311 | 322 | 264 | 271 | 332 | 334 |
| 12/18 | 249 | 240 | 227 | 240 | 221 | 305 | 305 | - | 303 | 303 |
| 13α/18 | 313 | 299 | 292 | 297 | 381 | 346 | 345 | - | 343 | 334 |
| 13β/18 | n.o. | 349 | n.o. | 348 | 359 | 251 | 249 | 260 | 247 | 252 |
For a more comprehensive dataset see the Supporting Information. Distances given in pm.
Average distances from NOESY spectra recorded at 100, 150 and 200 ms mixing times.
Distances obtained from molecular modeling using the software package Maestro 5.0.019 with the AMBER* forcefield including solvation (H2O). Global minima identified by conformational searches were evaluated.
trans-3 and trans-4 denote structures corresponding to 3 and 4 having trans configured azabicyclo[3.3.0]octane ring systems.
Data from Ref. 4c.
n.o. = not observed.
Overall, experimentally determined interproton distances were in good agreement with results obtained from molecular modeling. The similar distances between H11 and H20 for 3 and tetrabromostyloguanidine (5) (270 and 271 pm) and the respective shorter distance seen in hemiaminal epimer 4 (208 pm), support a relative configuration at C20 of palau’amine identical to that of analog 3 and tetrabromostyloguanidine (5). As expected for the cis-fusion of the five-membered rings, short distances were obtained between the angular protons H11 and H12 (221 and 228 pm for 3 and 4, respectively). The corresponding distance in the trans-fused ring system, which is predicted by molecular modeling to be considerably larger (304 pm), unfortunately could not be quantified in the investigation of tetrabromostyloguanidine (5).4c More diagnostic information was obtained from the examination of 1,3-interrelated proton pairs H11/H13β, H11/H18, and H13β/18, situated on the same side of the molecular plane. In accordance with the molecular model of cis-configured structure 1, all of these distances were found to be large (≥ 334 pm) for 3 and 4, whereas significantly shorter interproton distances (≤ 260 pm) were reported for tetrabromostyloguanidine (5), and calculated for 2 and hypothetical trans-configured diastereoisomers trans-3 and trans-4. In particular, the distance between H11 and H13β, which could be experimentally quantified in all cases, was found to be 85 pm larger for cis-fused 3 and 4 (335 and 334 pm) than for tetrabromostyloguanidine (5) (250 pm). No NOE correlations were observed for proton pairs H11/H18 and H13β/18 in the case of 3 and 4, whereas the corresponding distances for 5 were experimentally determined to be short (233 and 266 pm, respectively). In addition, the close proximity of H12 and H18 (249 and 227 pm) and a NOE correlation between H13α and H18 (313 and 292 pm), observed in the case of 3 and 4, were not mentioned in the quantitative analysis of tetrabromostyloguanidine (5). As highlighted by Grube and Köck,4c this discrepancy is best explained by assuming a trans fusion of the central azabicyclic fragment of palau’amine and congeners.
Conclusions
In summary, the first synthesis of hexacyclic palau’amine congeners that incorporate both guanidine functional groups has been accomplished. These palau’amine analogs, 3 and 4, have the cis configuration of their azabicyclo[3.3.0]octane cores. Synthetic access to these analogs allowed the first direct comparison of NMR data for hexacyclic diguanidine structures having the originally proposed cis-azabicyclo[ 3.3.0]octane fragment with data for natural alkaloids of the palau’amine family. This comparison provides additional strong evidence in favor of the recently proposed structural revision of these compounds, fully supporting the trans configuration of the central azabicyclo[3.3.0]octane ring system of palau’amine and congeners.
Experimental Section18
Intermediate 17
n-Tetrabutylammonium fluoride (155 µL of a 1 m solution in THF, 0.155 mmol) was added at room temperature to a solution of bis(spirocycle) 15 (37 mg, 0.031 mmol) in THF (2 mL), and the yellowish solution was stirred at room temperature for 8 min., then partitioned between saturated aqueous NH4Cl solution and EtOAc. The organic phase was dried over Na2SO4 and concentrated. The residue was purified by column chromatography (SiO2; EtOAc/MeOH 14:1) to afford 24.5 mg (95%) of an inseparable mixture (1.5:1) of hemiaminal diastereoisomers 17 as a colorless, amorphous solid. 1H NMR (500 MHz, CD3OD) δ 8.04 (d, J = 7.9 Hz, 1 H), 7.90 (d, J = 7.8 Hz, 0.66 H), 7.66 (t, J = 7.9 Hz, 1 H), 7.53-7.45 (m, 3.98 H), 7.36-7.23 (m, 8.3 H), 6.95 (s, 0.66 H), 6.94 (s, 1 H), 6.22 (s, 1 H), 5.81 (s, 0.66 H), 5.06-5.01 (m, 5.64 H), 4.78 (d, J = 17.4 Hz, 1 H), 4.63 (m, 1.66 H), 4.44-4.37 (m, 2.66 H), 3.46 (m, 1.66 H), 3.37 (m, 0.66 H), 2.75-2.59 (m, 3.32 H), 2.52-2.45 (m, 5.64 H), 2.02 (bd, J = 10.3 Hz, 1 H); 13C NMR (125 MHz, CD3OD) δ 169.9, 149.4, 165.3, 165.2, 163.7, 162.8, 155.3, 150.3, 149.9, 138.8, 138.6, 134.8, 134.5, 134.2, 133.5, 131.8, 130.4, 130.0, 129.61, 129.59, 129.5, 129.14, 129.07, 129.0, 126.2, 126.1, 124.7, 124.4, 117.3, 117.2, 109.3, 109.0, 93.0, 92.5, 89.3, 88.5, 76.8, 76.5, 71.8, 70.5, 69.9, 68.5, 68.3, 60.4, 56.1, 50.7, 50.4, 44.1, 43.3, 40.8, 30.9, 14.6, 13.9; FTIR (film) 3371, 3263, 2933, 1636, 1590, 1524, 1497, 1443, 1426, 1383, 1337, 1287, 1266, 1119, 1063, 982, 799, 735, 700 cm−1; HRMS (ESI) m/z calcd. for C32H30Br2N8NaO7S (M+Na): 851.0223, found: 851.0219.
Intermediate 20
A solution of m-chloroperoxybenzoic acid (washed with saturated aqueous NaHCO3 solution; 5.4 mg, 0.031 mmol) in CH2Cl2 (1 mL) was added at 0 °C to a solution of hemiaminal diastereoisomers 19 (8.6 mg, 0.01 mmol) in CH2Cl2 (2 mL) in a sealable high-pressure reaction tube. The resulting clear solution was stirred at 0 °C for 10 min., then allowed to warm to room temperature and stirred for another 30 min. This solution was cooled to −78 °C, and NH3 (ca. 3 mL) was condensed into the solution. The reaction tube was sealed, and the resulting cloudy solution was allowed to warm to room temperature (clear solution after ca. 10 min), and kept at this temperature for 14 h. The solution was then re-cooled to −78 °C, the tube was unsealed, and excess NH3 was allowed to evaporate with warming to room temperature. The resulting suspension was concentrated, and the residue was purified by RP-HPLC (Phenomenex C18, 5 µm, 250 × 21.2 mm, H2O (with 0.1% TFA) / MeOH linear gradient 50:50 → 10:90 in 20 min, flow rate 10 mL/min, tR = 14.8 min) to provide 7.5 mg (79%) of an inseparable mixture (5:1) of bis(guanidine) diastereoisomers 20 (as the trifluoroacetate salts) as a colorless, glass-like solid. 1H NMR (500 MHz, CD3OD) δ 8.17 (d, J = 8.1 Hz, 1 H), 8.09 (d, J = 8.1 Hz, 0.2 H), 7.78 (t, J = 7.6 Hz, 1 H), 7.71 (t, J = 7.7 Hz, 0.2 H), 7.61 (t, J = 7.7 Hz, 1.2 H), 7.55 (d, J = 7.7 Hz, 1 H), 7.41-7.31 (m, 6.2 H), 7.07 (s, 0.2 H), 7.06 (s, 1 H), 6.57 (s, 1 H), 6.25 (s, 0.2 H), 5.31 (d, J = 17.4 Hz, 1 H), 5.29 (d, J = 12.2 Hz, 0.2 H), 5.26 (d, J = 12.2 Hz, 0.2 H), 5.23 (s, 2 H), 4.97 (d, J = 17.4 Hz, 1 H), 4.84 (d, J = 16.6 Hz, 0.2 H), 4.31 (m, 1.2 H), 4.24 (dd, J = 9.9, 12.4 Hz, 1 H), 4.05 (dd, J = 8.9, 12.4 Hz, 0.2 H), 3.86 (d, J = 12.2 Hz, 1 H), 3.77 (dd, J = 3.8, 12.4 Hz, 0.2 H), 3.67 (dd, J = 6.5, 12.4 Hz, 1 H), 3.11 (dd, J = 6.4, 15.5 Hz, 0.2 H), 3.04 (m, 1.2 H), 2.32 (d, J = 13.8 Hz, 1 H), 2.28 (d, J = 13.8 Hz, 1 H), 2.16 (dd, J = 3.1, 15.5 Hz, 0.2 H); 13C NMR (125 MHz, CD3OD) δ 158.95, 158.93, 158.88, 158.6, 155.8, 155.4, 150.1, 149.7, 137.1, 136.9, 135.4, 135.3, 134.0, 132.2, 132.1, 131.4, 131.3, 130.9, 130.7, 130.5, 130.3, 129.8, 129.6, 129.4, 129.2, 126.8, 126.7, 124.7, 124.5, 118.6, 118.2, 109.3, 109.0, 104.9, 104.5, 90.8, 89.3, 83.5, 83.2, 72.2, 71.8, 70.4, 70.1, 70.0, 69.9, 69.4, 69.0, 54.8, 52.8, 47.2, 45.4, 45.1, 44.8, 44.1, 43.0, 42.5; FTIR (film) 2244, 3261, 2916, 2873, 2831, 1766, 1702, 1671, 1594, 1528, 1439, 1345, 1272, 1202, 1135, 801, 724 cm−1; HRMS (ESI) m/z calcd. for C31H29Br2N9O7 (M+H): 798.0635, found: 798.0668.
Products 3 and 4
A solution of bis(guanidine) diastereoisomers 20 (25 mg, 0.027 mmol) in dioxane (5 mL) was irradiated at room temperature for 2.5 h using a sunlamp (275 W). Pd/C (10%, 10 mg) was added to the resulting yellow solution, and the suspension was stirred under a H2 atmosphere for 4 h. The suspension was filtered through Celite, and the Celite was washed with dioxane (1 mL) and H2O/0.1%TFA (1 mL). Pd/C (10%, 20 mg) was added to the yellowish filtrate, and the suspension was stirred under a H2 atmosphere for 24 h, then filtered (45 µm nylon filter). The filtrate was concentrated, and the residue was redissolved in H2O/0.1%TFA and purified by RP-HPLC (Phenomenex C18, 5 µm, 250 × 21.2 mm, H2O (with 0.1% TFA) / CH3CN linear gradient 95:5 → 80:20 in 20 min, flow rate 10 mL/min) to afford 8 mg (43%) and 4.1 mg (22%) of bis(guanidine) hemiaminal diastereoisomers 3 and 4 respectively, (enriched, as the bis(trifluoroacetate) salts), as colorless, glass-like solids.19 Data for 3: tR = 14.9 min; 1H NMR (500 MHz, D2O) δ 7.13 (d, J = 1.3, 2.7 Hz, 1 H), 6.97 (dd, J = 1.3, 4.0 Hz, 1 H), 6.47 (dd, J = 2.7, 4.0 Hz, 1 H), 6.36 (s, 1 H), 5.34 (s, 1 H), 4.34 (m, 1 H), 4.16 (dd, J = 10.1, 12.3 Hz, 1 H), 3.82 (d, J = 12.0 Hz, 1 H), 3.65 (dd, J = 6.1, 12.3 Hz, 1 H), 3.07 (m, 1 H), 2.35 (dd, J = 3.7, 14.3 Hz, 1 H), 2.16 (d, J = 14.3 Hz, 1 H); 13C NMR (125 MHz, D2O) δ 158.7, 157.4, 156.9, 124.7, 120.5, 115.0, 113.0, 85.8, 82.0, 71.9, 70.1, 68.1, 53.1, 49.9, 45.3, 42.2; FTIR (film) 3188, 2836, 1671, 1578, 1557, 1472, 1432, 1389, 1326, 1191, 1135, 1092, 1038, 843, 801, 724 cm−1; HRMS (ESI) m/z calcd. for C16H20N8O3 (M+H): 373.1737, found: 373.1736. Data for 4: tR = 13.9 min; 1H NMR (500 MHz, D2O) δ 7.10 (d, J = 1.3, 2.7 Hz, 1 H), 6.97 (dd, J = 1.3, 4.0 Hz, 1 H), 6.47 (dd, J = 2.7, 4.0 Hz, 1 H), 6.24 (s, 1 H), 5.19 (s, 1 H), 4.34 (m, 1 H), 4.07 (dd, J = 9.9, 12.3 Hz, 1 H), 3.72 (dd, J = 4.6, 12.3 Hz, 1 H), 3.35 (d, J = 10.7 Hz, 1 H), 3.15 (m, 1 H), 2.79 (dd, J = 5.1, 15.4 Hz, 1 H), 2.08 (dd, J = 2.5, 15.4 Hz, 1 H); 13C NMR (125 MHz, D2O) δ 158.7, 157.4, 156.7, 124.5, 120.3, 115.0, 113.2, 87.9, 81.8, 70.7, 69.2, 68.7, 59.9, 44.2, 42.7, 41.2.
Supplementary Material
Experimental procedures and characterization data for new compounds, comprehensive set of experimentally determined and calculated interproton distances for 3 and 4, and molecular models of computed lowest-energy conformers. This material is available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment
Financial support from the National Heart, Lung, and Blood Institute (HL-25854), NIH NRSA postdoctoral fellowship support to B.A.L. (GM-71132), and Deutsche Akademie der Naturforscher Leopoldina (BMBF-LPD 9901/8-142) postdoctoral fellowship support for R.P. are gratefully acknowledged. We thank Prof. J.S. Nowick for helpful discussions.
References
- 1.For recent reviews on the structural diversity of pyrrole-imidazole alkaloids and synthetic approaches toward some of the family members, see: Hoffmann H, Lindel T. Synthesis. 2003:1753–1783.Jacquot DEN, Lindel T. Curr. Org. Chem. 2005;9:1551–1565.
- 2.Kinnel RB, Gehrken H-P, Scheuer PJ. J. Am. Chem. Soc. 1993;115:3376–3377.Kinnel RB, Gehrken H-P, Swali R, Skoropowski G, Scheuer PJ. J. Org. Chem. 1998;63:3281–3286.. Additional congeners have been isolated and characterized: Kato T, Shizuri Y, Izumida H, Yokoyama A, Endo M. Tetrahedron Lett. 1995;36:2133–2136.Kobayashi J, Suzuki M, Tsuda M. Tetrahedron. 1997;53:15681–15684.
- 3.Publications not covered in Refs. 1a and 1b include: Dransfield PJ, Dilley AS, Wang S, Romo D. Tetrahedron. 2006;62:5223–5247.Wang S, Dilley AS, Poullennec KG, Romo D. Tetrahedron. 2006;62:7155–7161.Nakadai M, Harran PG. Tetrahedron Lett. 2006;47:3933–3935.Tan X, Chen C. Angew. Chem. Int. Ed. 2006;45:4345–4348. doi: 10.1002/anie.200601208.
- 4.Kobayashi H, Kitamura K, Nagai K, Nakao Y, Fusetani N, van Soest RWM, Matsunaga S. Tetrahedron Lett. 2007;48:2127–2129.Buchanan MS, Carroll AR, Addepalli R, Avery VM, Hooper JNA, Quinn RJ. J. Org. Chem. 2007;72:2309–2317. doi: 10.1021/jo062007q.Grube A, Köck M. Angew. Chem. Int. Ed. 2007;46:2320–2324. doi: 10.1002/anie.200604076.. The relative configuration of the C20 stereogenic center was assigned as shown for 1 in the initial isolation report (Ref. 2a), then corrected in the following publication (Ref. 2b). All congeners of palau’amine were assigned the initial relative configuration at C20 (Refs. 2c, 2d, 4a–c).
- 5.This natural product was also isolated from another marine sponge and named carteramine A, see Ref. 4a.
- 6.(a) Overman LE, Rogers BN, Tellew JE, Trenkle WC. J. Am. Chem. Soc. 1997;119:7159–7160. [Google Scholar]; (b) Bélanger G, Hong F-T, Overman LE, Rogers BN, Tellew JE, Trenkle WC. J. Org. Chem. 2002;67:7880–7883. doi: 10.1021/jo026282y. [DOI] [PubMed] [Google Scholar]; (c) Katz JD, Overman LE. Tetrahedron. 2004;60:9559–9568. [Google Scholar]
- 7.Foley LH, Büchi G. J. Am. Chem. Soc. 1982;104:1776–1777. [Google Scholar]
- 8.Prepared from potassium thiocyanate and benzyl chloroformate according to: Wang SS, Magliocco LG American Cyanamid Company. 5194673. US Patent. 1993 (see the Supporting Information)
- 9.Treatment of 10 with EDCI and an excess of hexamethyldisilazane (HMDS) under similar conditions provided the corresponding (monoprotected) guanidine, which could be converted to the glycocyamidine spirocycle by incubation with diisopropylethylamine in toluene at 80 °C for 6 h (see Ref. 10d). However, carbamate protection of the endocyclic glycocyamidine nitrogen provided derivatives that were labile, and, therefore, inferior to intermediates masked with alkyl groups.
- 10.(a) Linton BR, Carr AJ, Orner BP, Hamilton AD. J. Org. Chem. 2000;65:1566–1568. doi: 10.1021/jo991458q. [DOI] [PubMed] [Google Scholar]; (b) Manimala JC, Anslyn EV. Tetrahedron Lett. 2002;43:565–567. [Google Scholar]; (c) Guisado O, Martinez S, Pastor J. Tetrahedron Lett. 2002;43:7105–7109. [Google Scholar]; (d) Shinada T, Umezawa T, Ando T, Kozuma H, Ohfune Y. Tetrahedron Lett. 2006;47:1945–1947. [Google Scholar]
- 11.The configuration of the hemiaminal stereocenters was elucidated by 2D NMR experiments (see the Supporting Information).
- 12.The inability to separate the hemiaminal epimers of this product and later intermediates likely results from ready interconversion of the diastereomers on SiO2 and in protic solvents.
- 13.The simultaneous removal of the TBS ether is undoubtedly facilitated by the hemiaminal hydroxyl group, because desilylation does not occur in cyclizations of related substrates in which C21 is at the carbonyl oxidation state.
- 14.Commercial m-chloroperoxybenzoic acid was purified by washing a CH2Cl2 solution of the peroxy acid with saturated aqueous NaHCO3 solution, drying the organic phase over Na2SO4, and carefully concentrating the dried solution in vacuo.
- 15.Among the publications on palau’amine and related structures, a H11/H12 NOE correlation was only mentioned for konbu’acidin A and tetrabromostyloguanidine (Refs. 2d and 4c).
- 16.Interproton distances were obtained from NOESY spectra with different mixing times (100, 150, and 200 ms). Intensity data from volume integration of NOESY crosspeaks were calibrated using the geminal proton pair at C13 (178 pm). Each NOESY spectrum was analyzed separately (linear approximation, verified by a linear relationship between volume integrals for different mixing times).
- 17.In all cases conformational searches were carried out to determine global minima. Only minima lacking intramolecular H-bonds were considered for calculations using Spartan 04. These results are given in the Supporting Information.
- 18.General experimental details are provided in the Supporting Information.
- 19.Hemiaminal diastereoisomers 3 and 4 interconvert at room temperature in H2O and therefore could not be obtained in diastereomerically pure form (i.e., diastereoisomer 3 could only be enriched to ca. 90% (1H-NMR)).
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Experimental procedures and characterization data for new compounds, comprehensive set of experimentally determined and calculated interproton distances for 3 and 4, and molecular models of computed lowest-energy conformers. This material is available free of charge via the Internet at http://pubs.acs.org.



