Article - Organic Chemistry Laboratory, UW-Platteville - Example of a Literature Article

Technical Papers

An Example

DR. SUNDIN

Organic Chemistry Laboratory

Give below is a technical paper that appeared in the "Journal of the American Chemical Society". You may want to check the journal itself to see how it actually looks in print. Note that the tables and drawing have been removed. While you certainly aren't expected to understand the paper, you should be able to read enough of it to see that:

The Abstract summerizes in as few sentences as possible, what was done in the work being considered.

The Discussion discusses the signifance of the work, the interpretations of the experimental work, the conclusions that are drawn, etc. This will include opinions (which later may be found to be wrong).

The Experimental section just gives the facts, what was done, how it was done, what was observed or measured. Note that the "Experimental" is not a lab manual in that it does not give directions. It reports what was actually done! Nonetheless, the experimental should have enough detail so that others could repeat the experiments or modify them for different compounds or materials.

J. Am. Chem. Soc., 118 (38), 9017 -9023, 1996. ja950054m S0002-7863(95)00054-0

Copyright © 1996 by the American Chemical Society

Ecteinascidins: Putative Biosynthetic Precursors and Absolute Stereochemistry

Ryuichi Sakai, Elizabeth A. Jares-Erijman, Ignacio Manzanares, Maria V. Silva Elipe, and Kenneth L. Rinehart*

Contribution from the Roger Adams Laboratory, University of Illinois, Urbana, Illinois 61801, and PharmaMar, S.A., Tres Cantos, Spain

Received: January 6, 1995

New bioactive ecteinascidins (Et's) 597 (1), 583 (2), 594 (3), and 596 (4)-putative biosynthetic precursors of previously described Et's [e.g., Et 743 (5)]-were isolated from the Caribbean tunicate Ecteinascidia turbinata. Structures assigned to these compounds based on spectroscopic data represent a novel series of Et's with L-cysteine or its -oxo analog as unit C. The absolute configuration of the L-Cys unit of 1 was assigned by chiral GC, while a 2D ROESY (rotating-frame Overhauser enhancement spectroscopy) spectrum of its acetyl derivative 1a completed the assignment of the stereochemistry of 1 as 1R,2R,3R,4R,11R,13S,21S,1'R.

In the present paper we report new ecteinascidins (Et's) 1-4 (Et's 597, 583, 594, and 596, respectively; Chart 1). Et's are exceedingly potent antitumor agents first isolated by Holt1a-c from the marine tunicate Ecteinascidia turbinata, following earlier decade-long efforts.1d-f Previously reported Et's 743 (5), 729 (6), 736 (7), and 722 (8) show promising efficacy in vivo, including activity against P388 murine leukemia, B16 melanoma, M5076 ovarian sarcoma, Lewis lung carcinoma, and several human tumor xenograft models in nude mice.1g With advanced stage MX-1 mammary tumor xenografts, 9 of 10 mice were tumor free on day 23 following treatment with 5, and with early stage MX-1 xenografts, 10 of 10 were tumor free (Table 1). The most abundant analogue, 5, is now in phase I clinical trials as an anticancer agent in three European countries and in the United States.2

The structures of Et's 743 (5), 729 (6), 745 (9), 759 B (10), and 770 (11) were assigned primarily by spectroscopic, especially FABMS and NMR, studies. Pseudomolecular formulas (M + H - H2O)+ of Et's were assigned by HRFABMS and ESCA (electron spectroscopy for chemical analysis), which identified the sulfur,1a and molecular formulas (M - H)- assigned by negative ion HRFABMS, which required an extra water molecule.3a The three tetrahydroisoquinoline units and their aromatic ring (A, B, C) substituents were identified by Holt1a and these were combined into partial formulas,1c the substituent patterns of which at C-5 to C-8 and C-6' to C-8' were later corrected, first by Wright et al.3b and also by us,3a and incorporated into consensus structures based on the correct formulas.3a,b The relative stereochemistry of Et's for most of the stereocenters in the A-B unit was assigned by NMR data,3a,b,d and crystal structures of some Et analogues completed the unambiguous assignment of relative stereochemistry, especially at the difficult quaternary C-1'.1g,3c Although the X-ray diffraction data did not assign the absolute stereochemistry, it was assumed to be the same as in safracin A, R at C-1'.3a,b The present study confirms the absolute stereochemistry suggested.

The previously reported ecteinascidins are composed either of three tetrahydroisoquinoline subunits A-C or two tetrahydroisoquinoline units (A,B) plus one tetrahydro--carboline unit (C) (Chart 1). The carbon and nitrogen framework of units A-B of the Et's is the same as that of the saframycins and safracins, antitumor agents first isolated from cultured Streptomyces species.4 Compounds related to the saframycins which have the same bis(tetrahydroisoquinoline) framework have been found from other microorganisms5a and also from a marine sponge.5b Et's seem to share the biosynthetic origin of their A-B unit with those saframycin-class compounds, especially with safracin B (12)4b for unit A. However, the third tetrahydroisoquinoline or the tetrahydro--carboline unit C attached to unit B by the 10-membered sulfide-containing lactone is a quite distinctive feature of the Et molecules, both structurally and biosynthetically. The potent anticancer activity of Et's may, at least in part, be attributed to the unit C, since the related saframycin A (13), for example, which lacks the C unit, has lower efficacy than Et 729 in comparable tumor models.5a,6 Thus, it is likely that unit C in Et molecules has not only chemical but also biological importance:

We earlier proposed a biosynthetic scheme for the formation of the A-B and C units of Et's 729 and 7431c,3a which was later modified to include the new series of Et's 722 (7) and 736 (8) with a tetrahydro--carboline instead of a tetrahydroisoquinoline as unit C.1g Tryptamine was also isolated from the same source.1g This generalized our previously proposed biosynthetic scheme for the formation of unit C in which it would be formed by a Pictet-Spengler condensation of the cysteine-derived -mercaptopyruvic acid and an aromatic amine to give a tetrahydroisoquinoline or a tetrahydro--carboline C unit which then adds to the A-B unit.

In the present study we report the isolation, structure assignments, and biological activities of the new ecteinascidins 1-4 (Et's 597, 583, 594, and 596), putative biosynthetic precursors of the Et's. Isolation of these compounds allows us to derive a more detailed scheme for the biogenesis of unit C which differs from that originally proposed.1c,3a Moreover, the absolute configuration of 1 has been determined unambiguously, allowing us to assign that for the medicinally important Et 743 (5).

Ecteinascidins 597 and 583 (1 and 2). New Et's 1-4 were isolated as described in the Experimental Section,3d and their molecular formulas are based on positive ion HRFABMS data on dehydrated "molecular" ions (M + H-H2O).1c,g;3a,b,d;8 NMR spectra of 1, interpreted by COSY(correlation spectroscopy), HMBC (1H-detected multiple bond heteronuclear multiple-quantum coherence), and HMQC (1H-detected heteronuclear multiple-quantum coherence) data and those of 2 and 3, assigned by COSY, HMQC, and analogy to other previously reported Et's1c,g;3a,b,d (Table 2), revealed a lack of aromatic signals for unit C, i.e., neither a third tetrahydroisoquinoline unit nor a tetrahydro--carboline unit. On the other hand, resonances assigned for the A-B unit in 1 and 2 were very similar to those of 5 except for lack of a methylenedioxy group (13C NMR ca. 103 ppm; 1H NMR ca. 6.1 ppm, AB quartet, J = ca. 1 Hz)1g in 1 and 2. We have previously assigned the fragmentation patterns for Et's in FAB mass spectrometry using FABMS/CID/MS and high-resolution FAB mass spectra.1c,g;3a,b This detailed assignment of FABMS fragments for Et's is useful in identifying related biosynthetic precursors. FABMS/CID/MS data on M + H - H2O of 1 (m/z 598) gave product ions a and b common to the A unit of 5 at m/z 204 and 218, but product ions f and g for unit B, at m/z 262 and 248, and c and d for the subunit A-B in 1, at m/z 465 and 495, were shifted 2 Da higher from those of 5 (Scheme 1, Table 3).1c These MS observations and the NMR data suggested that unit A of 1 is the same as that of 5 and 7, while unit B of 1 contains two more hydrogens, in agreement with replacement of the methylenedioxy group in unit B of 5 and 7 by a methoxyl (1H NMR 3.71, 13C NMR 61.1 ppm) plus a hydroxyl group in 1 instead. The position of the 7-methoxyl group was confirmed by ROESY data (supporting information, Figure 5S) for N1'-acetyl-Et 597 (1a), which showed NOE cross peaks, indicating that aromatic methyl and methoxyl groups are adjacent to each other in both units A and B (Scheme 2).

In addition to the A-B unit of 1 (C27H32N2O8), the rest of the molecule (C unit) consists of C3H5NOS, which contains two degrees of unsaturation, including an ester carbonyl at 174.1 ppm, linking unit C to unit B. COSY and HMBC data for 1 showed that the spin system for -CHCH2OCO-, observed in all other Et's for C-1, C-22, and the ester carbonyl of unit C, respectively, is also present in 1. HMQC data showed a broad singlet CHN proton at 3.22 ppm and carbon at 54.3 ppm. The proton shifted to 4.53 ppm on acetylation of 1 and correlated to an exchangeable NH proton at 5.48 ppm in 1a, confirming a primary amine in 1. A sulfur attached to C-4 in 1 was identified by NMR resonances for H-4 (4.51 ppm) and C-4 (43.1 ppm), very similar to those in other Et's (cf. 5 and 7).1c,g;3a,b A methylene carbon at 35.4 ppm and two very broad protons at 2.20 ppm were also assignable to a sulfide-substituted -CH2-.1c-f Although correlation spectra (COSY, HMBC) failed to connect this broad sulfide methylene9 with the CHN proton to the ester carbonyl, due to the small sample size and broad, weak signals, these two groups must be connected to form a 10-membered sulfide-containing lactone as in other Et's. The relationship was also assured by the isolation of cysteine itself (cf. below). Thus, the structure of 1 was assigned as depicted in Chart 1.10

Ecteinascidin 583 (2) is the N12-demethyl analog of 1. In the 1H NMR spectrum of 2 only three methyl groups are observed in the region from 2.0 to 2.5 ppm, whereas four methyl signals appeared in the spectrum of 1. The molecular formula of 2, C29H35N3O9S, also corresponds to the lack of a methyl. FABMS/CID/MS spectra of 2 gave product ions a and b at m/z 190 and 204, 14 Da less than those from 1 (Table 3). COSY and HMQC data for 2 were compared with other ecteinascidins' NMR data, allowing assignment of all the protons and protonated carbons. Protonated carbons (Table 2) C-11 and C-13 were shifted upfield compared to those of 1 as a result of the demethylation at N-12, while 1H NMR signals were shifted downfield. These shifts in the NMR data were previously observed between the N12-methyl and N12-demethyl analogs of Et's.1c,g;3a,b,d All the above data are consistent with the assignment of ecteinascidin 583 (2) as the N12-demethyl analog of Et 597 (1), just as Et 729 (5) and Et 722 (7) are N12-demethyl analogs of Et 743 (6) and Et 736 (8).1g

Absolute Stereochemistry. A ROESY NMR spectrum of 1a showed an NOE between the amide proton and the methyl protons of the N-acetyl group, and an NOE between the N-acetyl methyl group and the 16-methyl group of unit A revealed the relative stereochemistry at C-1' as in Scheme 2. Treatment of 1 with HgCl2 followed by NaBH4 and methanolysis gave a mixture containing cysteine methyl ester. This product was derivatized with trifluoroacetic anhydride (TFAA) and trifluoroacetic acid (TFA) and then analyzed by chiral GC (Scheme 3). Injection of the analyte with D,L-TFA-Cys-OMe showed that the derivatized Cys (A) in the mixture coeluted with the L-isomer of the derivatized racemate (Figure 1). Thus, the absolute stereochemistry of 1 at C-1' was determined to be R. Since the relative stereochemistry of the C unit and that of the A-B unit was related by the above ROESY experiment, the stereochemistry of 1 is assigned as 1R,2R,3R,4R,11R,13S,21S,1'R. This agrees with that presumed from the X-ray study but differs from that proposed initially for 5.3a,b

Figure 1 Gas chromatogram for: reaction product A (below) and coinjection of A and bis-TFA-D,L-Cys-OMe (above) on Chirasyl Val-III (isothermal, 100 C).

Et 743 (5) must have the same absolute stereochemistry, since the CD spectrum for 1 was very similar to that of 5 (Figure 2). Moreover, it is most plausible biosynthetically that all other Et's have the same absolute stereochemistry for the A-B unit's chiral centers.

Figure 2 CD spectra for Et 597 (1) (solid) and Et 743 (5) (broken) in CH3OH.

Ecteinascidins 594 (3) and 596 (4). Compound 3 was obtained as a methanol adduct, giving a protonated "molecular" ion (M + H - H2O + MeOH) at m/z 627 in the magic bullet matrix.11 The FABMS spectra of the methanol adduct of 3 in glycerol matrix alone gave only the [M + H - H2O + MeOH]+ ion at m/z 627 (627.2020, HRFABMS); however, peaks at m/z 595.1750 (M + H - H2O, relative intensity 27%), 613.1827 (M + H, 52%), and 687.2205 (M + H - H2O + glycerol, 60%) were observed when a small amount of aqueous oxalic acid was added. 13C and 1H NMR data showed that the A-B unit of 3 is very similar to that of 5, and a FABMS/CID/MS spectrum of the methanol adduct (m/z 627) gave product ions a-d at m/z 204, 218, 463, and 493, respectively, fragments common to the A-B unit in 5. If the subunit A-B of 3 is the same as that of 5, the adducted methanol must be located on unit C. The C unit of 3 consists of C3H2O2S (three double bond equivalents) which is assigned to be a 2-oxo-2-deaminocysteine unit, as depicted in Chart 1. The 13C NMR spectrum for 3 in CD3OD showed a signal at 100.5 ppm assignable to a hemiketal at C-1'. An IR spectrum (film) of 3, with a shoulder at 1734 cm-1 as opposed to a sharp single absorption for esters as seen in 1, supported the presence of an -keto ester group in 3.12

Et 596 (4) was observed only in a mixture as a methanol adduct; it gave a protonated "molecular" ion at m/z 629.2171 (C31H37N2O10S), which corresponds to that of 3 with two more hydrogens. This could either be due to the replacement of the methylenedioxy group by OCH3 and OH groups in the B unit as in 1 or to the presence of the C-21 reduced structure seen in Et 745 (9), which usually gives a true M + H ion.1c,3a An attempt to obtain NMR data for this compound failed due to the minute amount of 4 in the mixture, but FABMS/CID/MS of the m/z 629 (M + H - H2O + MeOH) parent gave daughter ions a and b at m/z 204 and 218 for unit A, and ions f and g at m/z 465 and 495 for unit A-B, respectively (Table 3), suggesting that 4 has the intact A unit of Et 743 and the B unit plus H2, as seen in 1. Addition of excess cyanide to a methanolic solution of the mixture followed by a FABMS measurement gave ions for a monocyano (at m/z 624) and a dicyano adduct of 4 at m/z 651 (Figure 3), indicating the presence of two electrophilic centers, the C-21 carbinolamine and the C-1' oxo groups. Although these results are tentative, they are consistent with the structure of Et 596 as 4, a C-1'-deamino, C-1'-oxo derivative of 1.

Figure 3 FAB mass spectrum for Et 596 (4) in the presence of CH3OH (m/z 629, M + H + CH3OH - H2O)+ and in the presence of CN- (m/z 624, M + H + HCN - H2O; m/z 651, M + H + 2HCN - H2O).

Biogenesis of the Ecteinascidins (Scheme 4). As proposed earlier,1c,3a A-B units of Et's are most likely formed by condensation of two Dopa-derived building blocks, and the tetrahydroisoquinoline ring in unit B is closed by condensation (Pictet-Spengler) with a serine- (or glycine-) derived aldehyde as in the case of the related saframycins.13 S-Adenosylmethionine is the likely source of methyl groups at C-6, O-7, C-16, O-17, and N-12. The scheme has been partially confirmed by incorporation of labeled tyrosine,14 cysteine,14 methionine,14b glycine,14b and tryptophan.14b The sequence of the oxidation, methylation, and acetylation at the aromatic rings is not yet clear.

The present study offers the most plausible route for the formatiuon of unit C to be that in Scheme 4, with the side chain alcohol at C-22 being acylated by cysteine, followed by oxidation and addition (or substitution) at C-4 by the SH group of the side chain with closure of the 10-membered lactone ring to give 1. Transamination at C-1' forms an oxo derivative, 4, then oxidation of the 7-OCH3 of 4 yields the methylenedioxy group to form 3.15 The electrophilic ketone in 3 can be condensed in a Pictet-Spengler reaction with a dopamine derivative to form the third tetrahydroisoquinoline group in Et 743 (5) or with tryptamine to form the tetrahydro--carboline group in Et 736 (7).1g A similar biogenetic origin can be proposed for Et's 729 (6) and 722 (8), involving Et 583 (2) as an intermediate. This scheme differs from that suggested earlier1c,3a in which the C unit was formed independently by condensation of O-methyldopamine or tryptamine with mercaptopyruvate, then coupled to unit A-B. Biosynthetic studies to confirm or modify the present scheme are in progress.14

Bioactivities of New Et's. Biological activities of new Et's 1-3, Et 743 (5), and Et 729 (6), including cytotoxicity against several cell lines, antimetabolic activities, enzyme inhibitions and antimicrobial activity, are summarized in Table 4. New Et's which lack the aromatic C unit were generally 10-50 times less active than 5 and 6, except for 1, against the MEL 28 and CV-1 cell lines. All tested Et's showed potent inhibition of DNA and RNA synthesis and of RNA polymerase activity, but much less inhibition of DNA polymerase activity. Previously we proposed a possible binding model of Et 729 (6) and a DNA oligomer since it had been shown to bind DNA.1g,16 In this study, however, it is shown that Et's possess a stronger inhibitory effect against RNA synthesis and RNA polymerase activity relative to DNA. Saframycin A (13) was also shown to inhibit RNA and DNA synthesis but not protein synthesis following activation by dithiothreitol.17 It is interesting to note that the new Et's showed some inhibition of protein synthesis, while 5 and 6 were less active. More detailed biochemical studies on these Et analogues will be necessary to specify the mechanism of action of the ecteinascidins.

Experimental Section

General. NMR spectra were obtained on a 500 MHz FT NMR spectrometer using either CDCl3, CD3OD, or a mixture of both as solvents [internal standards: 7.26 (1H) and 77.0 (13C) ppm for CHCl3; 3.30 (1H) and 49.0 (13C) ppm for CD3OD or a mixture of CD3OD-CDCl3]. FABMS, FABMS/CID/MS, and HRFABMS data were recorded on a VG 70-SE-4F spectrometer. Optical rotations were measured with a JASCO DIP 370 digital polarimeter with a Na lamp (589 nm) using a 5 cm × 0.35 cm (1.0 mL) cell. CD spectra were obtained on a JASCO J720 spectropolarimeter. Reversed phase HPLC was performed using a C-18 semipreparative column (Altex, 10 m gel, 10 × 250 mm), a UV detector (254 nm), and a flow rate of 1 mL/min. An Ito coil (P.C. Inc.) was used for high-speed countercurrent chromatography (HSCCC) at 800 rpm.

Separation. A specimen of E. turbinata, collected in Puerto Rico in September 1992, was stored frozen until use. A frozen sample (2.8 kg) was extracted twice with 2-propanol (4 L each time, at less than 5 C) for 10 h. The alcoholic extract was evaporated to an aqueous emulsion (2.5 L) which was extracted with EtOAc (1 L × 1, 0.5 L × 1). The organic layer was concentrated and then partitioned between the lower and upper layers of a solvent system of EtOAc-heptane-MeOH-H2O, 7:4:4:3 (200 mL). The upper and lower layer gave 1.03 and 0.226 g of solids, respectively, upon concentration. The solid obtained from the lower layer was separated by C18 (25 g) flash chromatography. The first eluent (MeOH-0.4 M aqueous NaCl, 9:2, 50 mL) afforded bioactive fraction A (89.3 mg), and the second fraction (MeOH-CHCl3 wash) gave mostly lipids (116.5 mg). Fraction A was flash chromatographed over silica gel (pretreated with NH3, 0.5% w/w). The first (1:9 MeOH-CHCl3 eluate, fraction B) and second (1:4 MeOH-CHCl3 eluate) fractions exhibited activity against Bacillus subtilis (12 mm zone at 0.3 g/disk).

A stored (-20 C) specimen (50 kg) collected during 1992 and 1993 was processed similarly, giving fraction B'.

Fraction B' (80 mg) was separated by high-speed countercurrent chromatography (HSCCC, solvent toluene-Et2O-MeOH-H2O, 6:6:6:3; lower layer mobile phase; flow rate 1.8 mL/min) into 9 fractions (fractions C-1 to C-9). Fraction C-1 (9.9 mg) was separated by RPHPLC (MeOH-NaCl 0.02 M, 3:1) to give mainly four fractions. The first and second fractions were combined and then separated on a silica gel column (1.5 × 25 cm column, CHCl3-MeOH, 6:1) to give pure Et 597 (1, 1.45 mg) and Et 583 (2, 1.43 mg).

Ecteinascidin 597 (1): Light brown solid; []25D -49 (c 0.17, MeOH); IR (film) 3400, 2934, 1741, 1456, 1419, 1236, 1197, 1066, 995, 754 cm-1; UV (max) 207 ( 46 000), 230 (sh, 15 000), 278 (3500), 285 (3800); CD (ext) 207 (E 15.4), 219 (-7.4), 237 (6.5), 280 (-2.8), 287 (-3.3). For NMR data, see Table 2. Anal. calcd for C30H36N3O8S (M + H - H2O) Mr 598.2223, found Mr 598.2219 (HRFABMS).

Et 583 (2): Light yellow solid; []22D -47 (c 0.14, CHCl3-MeOH, 6:1); UV (max) 207 ( 48 000), 230 (sh, 9200), 280 (2100), 290 (2300). For NMR data, see Table 2. Anal. calcd for C30H36N3O8S (M + H - H2O) Mr 598.2223, found Mr 598.2219 (HRFABMS).

Fraction C-8 was purified by RPHPLC (same conditions as above). A broad peak (tR = 33-42 min) gave Et 594 (3, 1.2 mg).

Et 594 (3): Light yellow solid; []22D -58 (c 1.1, MeOH); UV (max) 207 ( 60 500), 230 (sh, 11 000), 287 (2900); FABMS m/z 627 (M + H + MeOH - H2O, magic bullet matrix), 595 (M + H - H2O), 613 (M + H), 687 (M + H + glycerol - H2O, glycerol matrix in presence of oxalic acid and water). For NMR data, see Table 2. Anal. calcd for C30H31N2O9S (M + H - H2O) Mr 595.1750, found Mr 595.1716 (HRFABMS).

Fraction C-6 was separated by HPLC (MeOH-NaCl 0.04 M, 3:1) to yield a fraction (0.5 mg) which showed a FABMS peak at m/z 629 for compound 4 as the major Et component.

N-Acetyl Et 597 (1a): Et 597 (1 mg) was treated with Ac2O (50 L) and Et3N (5 L) at room temperature for 30 min. The product was passed through a Sep-pak silica gel column with CHCl3-MeOH (9:1) and then purified by RPHPLC (9:2 MeOH-NaCl, 0.04 M) to give a monoacetyl derivative (0.5 mg): FABMS m/z 641 (M + H - H2O); 1H NMR (CDCl3) 6.70 (1H, s), 5.48 (1H, brm), 5.12 (1H, d, J = 12.0 Hz), 5.10 (1H, brs), 4.87 (1H, brs), 4.53 (1H, m), 4.32 (1H, dd, J = 11.5, 2 Hz), 4.22 (1H, brd, J = 2.5 Hz), 4.00 (1H, brd, J = 8.5 Hz), 3.82 (3H, s), 3.80 (3H, s), 3.47 (1H, d, J = 18.5 Hz), 3.10 (1H, dd, J = 18.5, 8.5 Hz), 2.58 (3H, s), 2.36 (3H, s), 2.27 (3H, s), 2.08 (3H, s), 1.87 (3H, s), and a small amount of diacetyl derivative (only enough to obtain FABMS data). Anal. calcd for C32H39N3O9S (M + H - H2O) Mr 641.2407, found Mr 641.2398 (HRFABMS). Anal. calcd for C34H41N3O10S (M + H - H2O) Mr 683.2513, found Mr 683.2492 (HRFABMS).

Degradation of Ecteinascidin 597. GC Analysis. Compound 1 (0.5 mg) in methanol was treated with HgCl2 (1 mg) followed by NaBH4. A dark precipitate was removed by filtration and the mixture was treated with acidic methanol (MeOH-10% HCl, 110 C, 30 min). The solvent was removed and the product was derivatized with trifluoroacetic anhydride (TFAA-TFA, 100 L each, 110 C, 15 min). The TFA derivative was then analyzed by chiral GC (Chirasil Val-III, 110 C, isothermal). The standard sample of D,L-TFA-Cys-OMe was also prepared as above by methylation and TFA derivatization. Acknowledgment

This paper is dedicated to our longtime colleague Professor Nelson Leonard, who is celebrating his 80th birthday. We thank Drs. D. Lednicer and S. Donohue, National Cancer Institute, for in vivo data. In vitro biological testing was performed by Drs. T. García de Queseda and D. García Grávalos, PharmaMar, S.A. This work was supported in part by research grants from the National Institute of Allergy and Infectious Diseases (AI01278, AI04769). Supporting Information Available

1H NMR spectra of 1-3, 13C NMR spectrum of 1, all in CD3OD, and ROESY spectrum (CDCl3) of 1a (5 pages). See any current masthead page for ordering and Internet access instructions.

* Author to whom correspondence should be addressed.

University of Illinois.

Abstract published in Advance ACS Abstracts, September 1, 1996.

1. (a) Holt, T. G. Ph.D. Thesis, University of Illinois, Urbana, 1986; Diss. Abstr. Int. B 1987, 47, 3771-3772; Chem. Abstr. 1987, 106, 193149u. (b) Rinehart, K. L.; Holt, T. G. U.S. Patent 5,089,273, Feb. 18, 1992; Chem. Abstr. 1992, 117, 63016q. (c) Rinehart, K. L.; Holt, T. G.; Fregeau, N. L.; Keifer, P. A.; Wilson, G. R.; Perun, T. J., Jr.; Sakai, R.; Thompson, A. G.; Stroh, J. G.; Shield, L. S.; Seigler, D. S. J. Nat. Prod. 1990, 53, 771-792. (d) Sigel, M. M.; Wellham, L. L.; Lichter, W.; Dudeck, L. E.; Gargus, J. L.; Lucas, A. H. In Food-Drugs from the Sea Proceedings 1969; Younghen, Jr., H. W., Ed.; Marine Technology Soc.: Washington, D.C., 1969; pp 281-294. (e) Lichter, L.; Ghaffar, A.; Wellham, L. L.; Sigel, M. M. In Drugs & Food from the Sea; Kaul, P. N.; Sindermann, C. J., Eds., University of Oklahoma: Norman, OK, 1978; pp 137-144. (f) Sigel, M. M.; Lichter, W.; Wellham, L. L.; McCumber, L. J. 4th Intl. Symposium on Marine Natural Products, Tenerife, Spain, July 26-30, 1982, Abstract C-12. (g) Sakai, R.; Rinehart, K. L.; Guan, Y.; Wang, A. H.-J. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 11456-11460. (h) Rinehart, K.; Sakai, R.; U.S. Patent 5,149,804; Sept. 22, 1992; Chem. Abstr. 1992, 117, 205189z.

2. G. Faircloth, PharmaMar U.S., Cambridge, MA, personal communication to K. L. R.

3. (a) Rinehart, K. L.; Holt, T. G.; Fregeau, N. L.; Stroh, J. G.; Keifer, P. A.; Sun, F.; Li, L. H.; Martin, D. G. J. Org. Chem. 1990, 55, 4512-4515. (b) Wright, A. E.; Forleo, D. A.; Gunawardana, P. G.; Gunasekera, S. P.; Koehn, F. E.; McConnell, O. J. J. Org. Chem. 1990, 55, 4508-4512. (c) Guan, Y.; Sakai, R.; Rinehart, K. L.; Wang, A. H.-J. J. Biomol. Struc. Dynamics 1993, 10, 793-818. (d) Sakai, R., Ph.D. Thesis, University of Illinois, Urbana, 1991.

4. (a) Arai, T.; Takahashi, K.; Kubo, A. J. Antibiot. 1977, 30, 1015-1018. (b) Ikeda, Y.; Idemoto, H.; Hirayama, F.; Yamamoto, K.; Iwao, K.; Asao, T.; Munakata, T. J. Antibiot. 1983, 36, 1279-1283.

5. (a) Remers, W. A. The Chemistry of Antitumor Antibiotics; Wiley: New York, 1988; Vol 2, pp 93-119. (b) He, H.-Y.; Faulkner, D. J. J. Org. Chem. 1989, 54, 5822-5824.

6. Saframycin A exhibited antitumor activity against B16 melanoma, T/C = 146 (1 mg/kg, 1-5 days ip), and safracin B had T/C 200 (0.5 mg/kg, 1-9 days, ip),5a while Et 729 had T/C = 253 (12.5 g/kg, 1-9 days, ip) with five survivors at day 42.3d

7. The present compounds (1-4) have been briefly described in a meeting report: Rinehart, K. L.; Tachibana, K. J. Nat. Prod. 1995, 58, 344-358.

8. Due to the reactive carbinolamine at C-21, compounds 1 and 2 gave only dehydrated molecular ions in positive FABMS spectra. Attempts to observe [M - H]- by negative FABMS failed (see ref 1e).

9. Signal broadening in medium-sized rings has been commonly observed for marine polyethers; cf. Murata, M.; Legrand, A. M.; Ishibashi, Y.; Yasumoto, T. J. Am. Chem. Soc. 1989, 111, 8929-8931.

10. ROESY data confirmed all other stereochemistry of the A-B unit of 1 to be the same as in 5; see Scheme 2. X-ray data for derivatives of 5 had revealed all relative stereochemistry for the A-B unit.1f

11. Magic bullet contains 10% methanol; cf. Witten, J. L.; Schaffer, M. H.; O'Shea, M.; Cook, J. C.; Hemling, M. E.; Rinehart, K. L., Jr. Biochem. Biophys. Res. Commun. 1984, 106, 811-813.

12. IR absorption for the -keto group is expected to be small, since compound 1 exists mostly as the hemiketal or ketal in solution, as evidenced by FABMS. In the solid form, it may also exist, at least in part, as the hemiketal or ketal form.

13. Mikami, Y.; Takahashi, K.; Yazawa, K.; Arai, T.; Namikoshi, M.; Iwasaki, S.; Okuda, S. Biol. Chem. 1985, 260, 344-348.

14. (a) Kerr, R. G.; Miranda, N. F. J. Nat. Prod. 1995, 58, 1616-1621. (b) Morales, J. J.; Rinehart, K. L. 1995 International Chemistry Congress Pacific Basin Societies, Honolulu, Dec 17-22, 1995, Abstract ORGN 184.

15. Enzymatic reactions to form a methylenedioxy bridge have been demonstrated for plant-derived tetrahydroisoquinolines; cf. Bauer, W.; Zenk, M. H. Phytochemistry 1991, 30, 2953-2961.

16. Guan, Y.; Sakai, R.; Rinehart, K. L.; Wang, A. H.-J. J. Biomol. Struct. Dyn. 1993, 10, 793-818.

17. Ishiguro, K.; Sakiyama, S.; Takahashi, K.; Arai, T. Biochemistry 1978, 17, 2545-2550.

| Chemistry Home | Dr. Sundin Home | Chemistry 351 | sundin@uwplatt.edu |