New Structures and Bioactivity Properties of Jasplakinolide (Jaspamide) Analogues from Marine Sponges

Abstract

The goal of this study was to isolate and study additional jasplakinolide analogues from two taxonomically distinct marine sponges including two Auletta spp. and one Jaspis splendens. This led to the isolation of jasplakinolide (1) and eleven jasplakinolide analogues (3 – 13) including seven new analogues (6 – 10, 12, and 13). Structure elucidation of the new compounds was based on a combination of 1D and 2D NMR analysis, optical rotation, circular dichroism, and preparation of Mosher’s esters. Five of the new compounds are oxidized tryptophan derivatives of 1, including a unique quinazoline derivative (9). Compounds 1, 3, 5 – 8, and 11 were evaluated in the NCI 60 cell line screen and all compounds were tested in a microfilament disruption assay. Jasplakinolide B (11) exhibited potent cytotoxicity (GI₅₀ < 1 nM vs. human colorectal adenocarcinoma (HCT-116) cells) but did not exhibit microfilament-disrupting activity at 80 nM.

Introduction

In 1986 two publications appeared describing the structure of an unusual depsipeptide-polyketide compound named jasplakinolide (1),¹ by our lab, or jaspamide, by the Ireland-Faulkner-Clardy consortium.² The scope of this initial discovery has been greatly extended by the more than 400 papers published to date referencing 1 or its analogues, with the greatest number regarding cellular biology studies utilizing 1 (see Table S1, Supporting Information).³ An initial stimulus for further research was the impressive data from the US National Cancer Institute 60 cell screen data showing a significant cytotoxicity pattern for 1 (NSC: 613009 and P3933). The results indicated selectivity against renal, prostate, and CNS tumor cell lines. Subsequent biological evaluation revealed a mechanism of action involving microfilament disruption in vitro and an ability to permeate cells and alter the actin skeleton organization in vivo. This effect has also been exhibited by the close analogue chondramide C (2) isolated from the myxobacterium Chondromyces crocatus.⁴ Sustained interest in 1 can also be attributed to its ability to induce apoptosis, ^5–11 along with its additional actions as an anthelmintic,¹ insecticide,² and fish toxin.¹²

Development of 1 as a biological probe has transpired owing to its commercial availability based on natural sources.¹³ An evaluation of 20 marine sponges in the NCI repository identified samples in the Jaspis or Doryplores genera from seven Indo-Pacific regions as sources of 1.¹⁴ Similarly, the Crews and Zampella groups have isolated 1 from sponge collections spanning two different orders as summarized in Table 1.^{12, 15–18} The Vanuatu collections of Jaspis splendens studied by Zampella from 1999 – 2009 afforded 1 plus 14 analogues (renamed here using our favorite synonym):¹⁹ jasplakinolides B – H and J – P. ^{16, 20–22} Analogous results obtained by Schmitz in 1998, remaining yet unpublished, involved the examination of sponges, Jaspis or Dorypleres, one each from Truk lagoon and Yap Micronesia, which contained 1, debromo 1, 13-demethyl 1, plus jasplakinolides B, C, E, and three other Trp modified analogues.²³ Most recently, Proksch reported jasplakinolides Q and R, debromo and dibromo 1, from Kalimantan, Indonesia.²⁴

Table 1.

Marine Sponges Containing the Jasplakinolide (1) Family

Order	Family	Genus	Species	Collection Site	Reference
Astrophorida	Ancorinidae	Jaspis	splendens	Fiji	1, 2, this work
		Jaspis	sp.	Palau	2
		Jaspis	splendens	Vanuatu	16, 20–22
		Jaspis	splendens	Kalimantan, Indonesia	24
Halichondrida	Axinellidae	Auletta	cf. constricta	Milne Bay, Papua New Guinea	17
		Auletta	sp.	Papua New Guinea	this work

There have been two attempts to assess the structure activity pattern of the jasplakinolide framework based on natural material. Primarily through semi-synthesis we were able to study 1 plus six analogues as inhibitors of human prostate adenocarcinoma (PC-3) cell proliferation and observed that changes to the polyketide moiety involving transformation of the double bond into an epoxide or diol formation were detrimental to biological activity.²⁵ The Zampella lab evaluated their collection of fifteen compounds for cytotoxicity effects against human breast adenocarcinoma (MCF-7) and colon adenocarcinoma (HT-29) cells and found the levels of activity tracked with the anti-microfilament activity. Changes not well tolerated included: (a) replacement of the polyketide framework methyl groups, (b) migration of the polyketide double bond to an exocyclic position, and (c) modification of the Trp residue. Alternatively, replacement of the Ala side chain methyl did not substantially diminish the activity.

The eight total syntheses of 1, with the first appearing in 1988, have reaffirmed its absolute configuration, originally established by X-ray,² as 3R, 5S, 7S, 9S, 13R, and 15S.^26–32 These works also provide insight into routes for the preparation of 1 in small amounts (usually less than 10 mg). Some of these schemes have been used to prepare small sets of analogues for structure activity relationship studies, but none have shown biological activity on a par with that of 1. Highlights of the deleterious changes include inserting amino acids into the polyketide synthase (PKS) segment and replacing the methyl groups at C-13/C-15.^{33, 34} Similarly, total synthesis of 2 and three non-natural diastereomers revealed the sensitivity of cytotoxic action against solid tumors³⁵ and relative ability to induce actin polymerization to changes in configurations of substituents on the polyketide portion of the molecular framework. These results mesh well with the structure activity relationship (SAR) patterns described above for 1.

The most recent report by Zampella²² described research similar to that ongoing in our lab for the last decade on Indo-Pacific sponges rich in jasplakinolide and analogues. At one point we designated these sponges as Jaspis johnstoni,¹ but revised their identification to Jaspis splendens. It is important to underscore that the synonym, Dorypleres splendens, can be found in the literature for this organism.^{36, 37} The research reported below is based on three collections: two from Papua New Guinea (Auletta spp. coll. nos. 02137 and 02118) and one from Fiji (Jaspis splendens coll. no. 00101). As anticipated LCMS profiling of each extract showed that 1 was present and further research was begun to identify the several minor components present that appeared to be jasplakinolide analogues. A total of 11 jasplakinolide analogues were isolated and evaluated; these results are reported below.

Results and Discussion

The molecular formula and structural characteristics of 1 (C₃₆H₄₅BrN₄O₆) provided the benchmark to evaluate the three extracts and their chromatographic fractions. While the isolation steps, outlined in the experimental section, were straightforward, focusing on the intense LCMS molecular ion cluster for 1 ([M+H]⁺ m/z 709.2/711.2) provided an unambiguous handle to pinpoint fractions rich in this compound. Further evaluation of fractions containing known and/or new analogues of 1 was accomplished using a table of molecular formulas for 16 known jasplakinolides (see Table S10, Supporting Information). Strikingly, all congeners of 1 could be subdivided into three classes based heavy atom formulas as follows: C_35–37N₄O_6/7/9 (single MS peak), C_35–37BrN₄O_6–7 (doubled MS peak cluster), or C₃₆Br₂N₆O₆ (tripled MS peak cluster). This realization provided a rapid approach to dereplicate compounds suspected as being known, and also pinpointed compounds that appeared to be unknown. In addition, all previously described jasplakinolide analogues were divided into two groups as shown in Figure 1. Establishing the hybridization of C-31 by ¹H NMR provided the basis to discern between these groups: Group 1 = δ_H 1.5 – 1.6 (3H) and Group 2 = δ_H 5 – 6 (2H).

Figure 1.

The two groups of jasplakinolide frameworks.

*Structures of jasplakinolide B and C have been switched in various publications.¹⁶, ²⁰–²²

Both known and new compounds, possessing either Group 1 or Group 2 frameworks, were isolated from the extracts of the three sponges used in this study. A total of 12 compounds were obtained consisting of eight (four new) from Group 1 and four (three new) from Group 2. Overall, each metabolite was completely characterized and the absolute configuration shown here was largely assigned based on analogy to that unequivocally established (as described above) for 1. The constituents of the Auletta sp. coll. no. 02137 proved to be the most diverse. The known compounds included: Group 1 - 1 (C₃₆H₄₅BrN₄O₆),^{1, 2} jasplakinolide E (3, C₃₆H₄₅BrN₄O₇),²⁰ jasplakinolide F (4, C₃₅H₄₃BrN₄O₆),²⁰ jasplakinolide P (5, C₃₇H₄₈N₄O₉);²² and Group 2 - jasplakinolide B (11, C₃₆H₄₃BrN₄O₇).¹⁶ The new analogues from this sponge consisted of: 21-epi-jasplakinolide P (6, C₃₇H₄₈N₄O₉), jasplakinolide S (7, C₃₆H₄₆N₄O₈), 21-epi-jasplakinolide S (8, C₃₆H₄₆N₄O₈), jasplakinolide C_a (12, C₃₇H₄₆N₄O₁₀) and jasplakinolide C_b (13, C₃₇H₄₆N₄O₁₀). Less complex mixtures were present in the other two sponges. The Auletta sp. coll. no. 02118 was a source of the new compound jasplakinolide U (10, C₃₇H₄₆N₄O₁₀); and Jaspis splendens, coll. no. 00101, provided the new compound jasplakinolide T (9, C₃₆H₄₆N₅O₆). The stereo-structures of the known compounds were set by thorough examination of their NMR and MS physical properties compared to those in literature and structure elucidation of the new compounds proceeded as described below.

Two compounds possessing the molecular formula and overall structure of jasplakinolide P (C₃₇H₄₈N₄O₉)²² were isolated and designated as diastereomers 5 and 6. The unique heavy atom formulas of these compounds matched that of known jasplakinolide P (see Table S10, Supporting Information), and these data plus the aliphatic NMR resonance positions of C-31 for 5 (δ_H/C 1.59/18.9) and 6 (δ_H/C 1.58/18.8) facilitated completing the first steps in the dereplication process. The extensive NMR data in Table 2 and Table 3 alongside the selected 2D NMR correlations shown in Figure 2 confirmed that all structural elements previously shown in jasplakinolide P,²² including the unusual benzoxazinone moiety, were present in 5 and 6. We eventually decided that the former compound was identical to that previously described and that the latter was a new diastereomer. Part of the decision network involved assessing differences in ¹³C chemical shifts (Δδ_C) between all three data sets as shown in Figure 3. Overall, we observed very similar δ_C values for 5 and 6, with the greatest differences occurring at C-2 (Δδ_C 3.3 ppm) and C-20 (Δδ_C 1.6). This implied epimerization at adjacent chiral centers such as C-3, C-5, or C-21. Then, comparison of the Δδ_C data for 5, 6, and known jasplakinolide P provided the basis for the conclusion noted above that jasplakinolide P²² and 5 were identical, because (see Figure 3) this pair provided the best match at key sites including C-20, C-21, and C-2. While the original publication was silent on the assignment of the configuration at C-21 for jasplakinolide P,²² all other chiral centers were assigned in this compound, based on semi-rigorous analysis, as being identical to that of 1. Our final configurational assignment at C-21 was based on analogy to the structures of two new jasplakinolide diastereomers described next.

Table 2.

¹³C NMR Data of Selected Compounds Including 5 – 10, 12, and 13 in CD₃OD at 150 MHz

Compound	5	6	7	8	9	10a	12	13
pos #	δC, typeb	δC, typeb	δC, typeb	δC, typeb	δC, typeb	δC, type b	δC, typeb	δC, typeb
1	172.4, qC	172.5, qC	170.8, qC	171.1, qC	169.7, qC	172.0, qC	170.7, qC	170.5, qC
2	41.9, CH2	38.6, CH2	40.3, CH2	39.8, CH2	40.2, CH2	41.7, CH2	40.7, CH2	39.7, CH2
3	50.8, CH	50.9, CH	49.2, CH	49.2, CH	48.5, CH	50.9, CH	49.5, CH	49.4, CH
4	170.6, qC	170.3, qC	169.1, qC	168.6, qC	167.9, qC	171.1, qC	169.8, qC	169.8, qC
5	53.6, CH	54.4, CH	52.6, CH	52.8, CH	54.1, CH	54.1, CH	56.0, CH	56.3, CH
6	174.7, qC	175.1, qC	173.3, qC	173.7, qC	172.2, qC	175.1, qC	174.9, qC	174.9, qC
7	47.6, CH	47.7, CH	45.9, CH	46.4, CH	44.3, CH	46.7, CH	45.4, CH	45.6, CH
8	178.3, qC	178.2, qC	176.6, qC	176.6, qC	174.2, qC	177.7, qC	177.3, qC	177.1, qC
9	40.8, CH	40.4, CH	39.2, CH	38.4, CH	37.8, CH	40.4, CH	39.8, CH	38.2, CH
10	42.2, CH2	42.2, CH2	40.6, CH2	40.3, CH2	41.6, CH2	37.6, CH	38.8, CH	36.3, CH
11	134.3, qC	134.1, qC	132.6, qC	132.5, qC	131.2, qC	146.6, qC	148.5, qC	147.8, qC
12	130.3, CH	130.0, CH	128.7, CH	127.9, CH	129.2, CH	207.2, CH	77.4, CH	78.2, CH
13	31.1, CH	31.1, CH	29.5, CH	29.4, CH	28.8, CH	37.5, CH	30.8, CH	31.1, CH
14	44.6, CH2	44.7, CH2	43.2, CH2	43.0, CH2	42.4, CH2	42.5, CH2	35.7, CH2	36.3, CH2
15	72.0, CH	72.0, CH	70.4, CH	70.4, CH	70.3, CH	71.5, CH	69.8, CH	68.7, CH
16	132.8, qC	132.4, qC	130.9, qC	130.6, qC	131.4, qC	133.7, qC	131.5, qC	132.0, qC
17	128.7, CH	128.5, CH	126.8, CH	126.9, CH	126.9, CH	128.8, CH	127.2, CH	127.2, CH
18	116.4, CH	116.4, CH	114.8, CH	114.7, CH	114.8, CH	116.5, CH	114.7, CH	114.7, CH
19	158.1, qC	158.1, qC	156.5, qC	156.5, qC	156.4, qC	158.1, qC	156.3, qC	156.3, qC
20	39.9, CH2	41.5, CH2	34.1, CH2	34.2, CH2	31.7, CH2	40.4, CH2	23.0, CH2	22.9, CH2
21	108.6, qC	109.2, qC	75.0, qC	74.8, qC	167.6, qC	108.6, qC	109.1, qC	108.9, qC
22	119.0, qC	118.5, qC	130.9, qC	130.1, qC	123.4, qC	118.9, qC	127.2, qC	127.3, qC
23	127.2, CH	127.0, CH	124.0, CH	123.8, CH	125.1, CH	127.3, CH	117.6, CH	117.5, CH
24	124.7, CH	124.9, CH	122.2, CH	122.4, CH	127.7, CH	124.7, CH	119.1, CH	119.0, CH
25	132.1, CH	132.1, CH	129.3, CH	129.6, CH	133.9, CH	132.1, CH	121.3, CH	121.3, CH
26	115.7, CH	115.7, CH	110.1, CH	110.0, CH	128.2, CH	115.8, CH	110.1, CH	110.1, CH
27	137.5, qC	137.7, qC	141.2, qC	142.0, qC	148.8, qC	137.6, qC	136.5, qC	136.5, qC
28	152.2, qC	152.0, qC	180.2, qC	179.3, qC	153.4, CH	152.2, qC	108.9, qC	109.1, qC
29	18.3, CH3	18.2, CH3	16.5, CH3	16.4, CH3	17.5,c CH3	18.6, CH3	12.9, CH3	14.6, CH3
30	20.6, CH3	20.6, CH3	19.0, CH3	19.3, CH3	19.6, CH3	18.9, CH3	14.8, CH3	15.1, CH3
31	18.9, CH3	18.8, CH3	17.3, CH3	17.3, CH3	17.6, cCH3	129.1, CH2	111.7 CH2	111.2, CH2
32	22.5, CH3	22.5, CH3	20.9, CH3	21.0, CH3	21.6, CH3	17.8, CH3	16.8, CH3	16.8, CH3
33	19.8, CH3	19.8, CH3	18.2, CH3	18.1, CH3	19.2, CH3	20.6, CH3	19.5, CH3	19.7, CH3
34	31.6, CH3	31.8, CH3	30.0, CH3	30.3, CH3	30.6, CH3	32.4, CH3	31.7, CH3	31.2, CH3
O-CH3	51.2, CH3	51.1, CH3				51.1, CH3

^aMeasured at 125.7 MHz.

^bCarbon type established from ¹³C DEPT and/or gHMQC experiments.

^cChemical shifts are interchangeable.

Table 3.

¹H NMR Data of Selected Compounds Including 5 – 10, 12, and 13 in CD₃OD at 600 MHz

Compound	5	6	7	8	9	10	12	13
pos #	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])	δH, (int., m., J [Hz])c	δH, (int., m., J [Hz])c
2	2.65 (1H, dd, 15, 7.7)	2.61 (1H, dd, 14.5, 7)	2.63 (1H, dd, 14.8, 7.5)	2.64 (1H, dd, 14.7, 3.2)	2.59 (1H, dd, 15.1, 6.5)	2.70 (1H, dd, 16.2, 4.2)	2.72 (1H, dd, 15.7, 5.0)	2.67 (1H, dd, 16.0, 4.8)
	2.73 (1H, dd, 15.3, 4.8)	2.69 (1H, dd, 15, 4.5)	2.72 (1H, dd, 14.1, 4.0)	2.67 (1H, dd, 14.1, 4.7)	2.72 (1H, dd, 15.0, 4.5)	2.82 (1H, dd, 16, 7.0)	2.84 (1H, dd, 15.7, 7.0)	2.86 (1H, dd, 16.0, 7.2)
3	5.17 (1H, dd, 7.9, 4.7)	5.08 (1H, dd, 6.5, 4.8)	5.13 (1H, m)	5.04 (1H, m)	5.10 (1H, m)	5.19 (1H, dd, 9.5, 4)	5.20 (1H, bdd, 12.3, 7.0)	5.15 (1H, dd, 12.4, 6.6)
5	5.56 (1H, dd, 8.0, 5.2)	5.30 (1H, dd, 9.5, 3.5)	5.46 (1H, dd, 10.5, 4.9)	4.82b	5.85 (1H, bt, 7)	5.45 (1H, dd, 8.5, 4.5)	5.80 (1H, dd, 10.0, 6.5)	5.73 (1H, dd, 8.8, 7.5)
7	4.70 (1H, q, 6.7)	4.87b	4.65 (1H, m)	4.80 (1H, q, 7.0)	4.62 (1H, dd, 6.8, 7.2)	4.53 (1H, q, 6.5)	4.61 (1H, p, 7.0)	4.58 (1H, p, 6.6)
9	2.72 (1H, m)	2.71 (1H, m)	2.71 (1H, m)	2.71 (1H, m)	2.56 (1H, m)	2.61 (1H, m)	2.33 (1H, m)	2.51 (1H, ddd, 14.4, 11.8, 6.9)
10	1.94 (1H, d, 16.0)	1.91 (1H, d, 16.3)	1.92 (1H, brd, 15.3)	1.89 (1H, brd, 16.3)	1.81 (1H, dd, 15.5, 4)	2.33 (1H, dd, 13.2, 11.4)	2.12 (1H, dd, 13.9, 8.8)	2.14 (1H, dd, 14.4, 3.8)
	2.27 (1H, dd, 15.5, 10.8)	2.30 (1H, dd, 15.9, 11.1)	2.29 (1H, dd, 15.7, 11)	2.31 (1H, dd, 15.9, 11.5)	2.13 (1H, dd, 15.5, 9.5)	2.55 (1H, dd, 12.6, 3.6)	2.25 (1H, dd, 14.6, 4.7)	2.34 (1H, dd, 14.9, 7.1)
12	4.86b	4.85b	4.86b	4.82 (1H, d, 7.0)	4.82 (1H, d, 9.2)		3.91 (1H, 4.0)	3.72 (1H, d, 6.2)
13	2.32 (1H, m)	2.31 (1H, m)	2.33 (1H, m)	2.30 (1H, m)	2.27 (1H, m)	3.32 (1H, m)	1.63b	1.61b
14	1.22 (1H, m)	1.14 (1H, m)	1.20 (1H, m)	1.10 (1H, m)	1.27 (1H, ddd, 13.8, 6, 4.5)	1.40 (1H, ddd, 13.8, 6, 4.8)	1.17 (1H, ddd, 14.4, 6.2, 5.5)	1.05 (1H, ddd, 14.4, 8.4, 3.5)
	1.43 (1H, m)	1.31 (1H, m)	1.39 (1H, m)	1.27 (1H, m)	1.48 (1H, m)	2.12 (1H, ddd, 15.6, 14.4, 7.8)	1.56 (ovl)	1.80 (1H, ddd, 14.6, 10.2, 2.2)
15	4.70 (1H, m)	4.66 (1H, m)	4.65 (1H, m)	4.62 (1H, m)	4.66 (1H, dd, 13.0, 6.5)	4.79 (1H, m)	4.82 (1H, bp, 6.3)	4.87 (1H, m)
17	7.11 (2H, d, 8.7)	7.01 (2H, d, 8.5)	7.08 (2H, d, 8.5)	6.96 (2H, d, 8.8)	7.08 (2H, d, 8.5)	7.12 (2H, d, 8.4)	6.92 (2H, d, 8.5)	6.77 (2H, d, 8.5)
18	6.72 (2H, d, 8.8)	6.70 (2H, d, 8.5)	6.71 (2H, d, 8.8)	6.68 (2H, d, 8.6)	6.68 (2H, d, 8.5)	6.74 (2H, d, 8.4)	6.61 (2H, d, 8.5)	6.55 (2H, d, 8.5)
20	2.44 (1H, dd, 13.3, 7.8)	2.56 (1H, dd, 15.5, 9.5)	2.35 (1H, dd, 14.8, 4.1)	2.51 (1H, dd, 14.7, 9.7)	3.40b	2.53 (1H, dd, 15.6, 8.4)	3.16 (1H, dd, 15.2, 10.0)	3.11 (1H, dd, 14.9, 9.0)
	2.72 (1H, dd, 15, 4.5)	2.78 (1H, dd, 15.5, 3.5)	2.44 (1H, dd, 14.9, 9.6)	2.58 (1H, dd, 14.9, 6.2)	3.92 (1H, dd, 15.5, 7.5)	2.66 (1H, dd, 15.0, 4.2)	3.41 (1H, dd, 15.2, 6.6)	3.40 (1H, dd, 14.9, 7.3)
23	7.32 (1H, dd, 8.1, 1)	7.27 (1H, dd, 7.5, 1)	7.43 (1H, d, 7.5)	7.30 (1H, d, 8.0)	8.41 (1H, d, 8.4)	7.32 (1H, dd, 7.8, 1.2)	7.53 (1H, d, 7.8)	7.58 (1H, bd, 7.7)
24	7.11 (1H, dt, 7.8, 1.2)	7.10 (1H, dt, 8,1)	6.99 (1H, dt, 7.6, 0.8)	7.02 (1H, dt, 7.6, 0.8)	7.76 (1H, ddd, 8.4, 8.5, 1.2)	7.13 (1H, dt, 7.2, 1.2)	7.08 (1H, dt, 8.0, 1.0)	7.09 (1H, t, 8.4, 0.9)
25	7.34 (1H, dt, 7.6, 1.2)	7.35 (1H, dt, 7.7, 1.2)	7.22 (1H, dt, 7.8, 1.2)	7.24 (1H, dt, 7.7, 1.2)	7.99 (1H, bt, 8.5)	7.36 (1H, dt, 7.8, 1.2)	7.13 (1H, dt, 8.0, 1.0)	7.14 (1H, dt, 8.4, 1.2)
26	6.89 (1H, dd, 7.9, 0.7)	6.89 (1H, dd, 8.0, 1.0)	6.84 (1H, d, 7.7)	6.84 (1H, d, 7.7)	7.98 (1H, dd, 8.5, 1.2)	6.89 (1H, dd, 8.4, 1.2)	7.25b	7.24b
28					9.08 (1H, s)
29	1.16 (3H, d, 6.7)	1.31 (3H, d, 7.0)	1.05 (3H, d, 6.5)	1.35 (3H, d, 6.8)	0.86 (3H, d, 6.8)	1.12 (3H, d, 6.6)	0.75 (3H, d, 7.1)	0.84 (3H, d, 6.9)
30	1.13 (3H, d, 7.0)	1.14 (3H, d, 7.0)	1.12 (3H, d, 6.9)	1.14 (3H, d, 7.2)	0.97 (3H, d, 6.6)	1.15 (3H, d, 7.2)	1.13 (3H, d, 6.7)	1.15 (3H, d, 6.8)
31	1.59 (3H, s)	1.58 (3H, s)	1.58 (3H, s)	1.57 (3H, s)	1.51 (3H, s)	5.82 (1H, s)	5.07 (1H, bs)	5.06 (1H, bs)
						6.15 (1H, s)	4.95 (1H, bs)	4.94 (1H, bs)
32	0.86 (3H, d, 6.6)	0.83 (3H, d, 7.0)	0.86 (3H, d, 6.6)	0.81 (3H, d, 6.7)	0.83 (3H, d, 6.6)	1.03 (3H, d, 6.6)	0.79 (3H, d, 6.8)	0.87 (3H, d, 6.9)
33	1.06 (3H, d, 6.3)	1.02 (3H, d, 6.0)	1.07 (3H, d, 6.5)	1.01 (3H, d, 6.4)	1.04 (3H, d, 6.6)	1.14 (3H, d, 7.2)	1.04 (3H, d, 6.1)	0.95 (3H, d, 6.2)
34	2.91 (3H, s)	3.06 (3H, s)	2.89 (3H, s)	2.97 (3H, s)	3.00 (3H, s)	2.87 (3H, s)	3.08 (3H, s)	3.13 (3H, s)
O-CH3	3.01 (3H, s)	3.11 (3H, s)				3.04 (3H, s)

^aMeasured at 500 MHz.

^bShift assigned based on gHMQC and gHMBC data.

^cSpectra recorded in CDCl₃.

Figure 2.

Selected 2D NMR (600/150 MHz, CD₃OD) correlations of the benzoxazinone substructure of jasplakinolide P (5), 21-epi-jasplakinolide P (6), and jasplakinolide U (10).

Figure 3.

NMR Δδ_C shift differences (ppm) between: (a) literature values for jasplakinolide P²² vs. experimental values of jasplakinolide P (5) or 21-epi-jasplakinolide P (6) and (b) shift differences between jasplakinolide S (7) and 21-epi-jasplakinolide S (8).

During LCMS screening, a second set of isomers was observed displaying an [M+H]⁺ at m/z 663 and molecular formula of C₃₆H₄₆N₄O₈. These new compounds (7 and 8) possessed the same carbon count as 1, were initially designated as jasplakinolides S, and their heavy atom formulas were closest to that of jasplakinolide P (C₃₇N₄O₉) and jasplakinolide O (C₃₅N₄O₇).²² Similar to the analysis above, the aliphatic resonance positions of C-31 for 7 (δ_H/C 1.58/17.3) and 8 (δ_H/C 1.57/17.3) indicated the presence of the Group 1 macrolide. Somewhat problematic for assigning the site(s) of configurational differences between 7 and 8 was that evaluation of their Δδ_C chemical shifts revealed no difference greater than δ 0.9 ppm. The major difference between 1 and 7 / 8 was quickly recognized as due to the bromo-indole moiety being replaced by a 3-hydroxyindolin-2-one group. The 2D NMR correlations seen in Figure 4 were key to establishing the connectivity network between the non-benzenoid ring atoms consisting of C₂H₂NO₂. The most significant Δδ_C’s between 1 vs. 7 / 8 were the shifts of C-28 (1: δ_C 109.2; 7: δ_C 180.2; and 8: δ_C 179.3) and C-21 (1: δ_C 110.4; 7: δ_C 75.02; and 8: δ_C74.8), both consistent with the substructure shown in Figure 4. At this point, though none of the data collected provided the basis for a definitive conclusion, it was assumed that a configurational change at C-21 was the sole source of differences between 7 and 8.

Figure 4.

Selected 2D NMR (600/150 MHz, CD₃OD) correlations for the 3-alkyl-3-hydroxyindolin-2-one substructure of jasplakinolide S epimers 7 and 8.

A rapid way to interrogate the hypothesis advanced above involved collecting circular dichroism (CD) data for 7 and 8. We believed that this information could be used in a direct comparison to literature CD traces for a pair of 3-hydroxy-2-oxytryptophan derivatives.³⁸ Shown in Figure 5 are the published CD data for two key models, (−)-3(S)-hydroxy-2-oxo-S-tryptophan and (+)-3(R)-hydroxy-2-oxo-S-tryptophan, each prepared from S-tryptophan. These compounds display antipodal-like Cotton effect curves, which dramatically shift from positive to negative as a function of the chirality at C-3. The CD response of 7 exhibited a negative curve at 242 nm and positive curves at 264 and 292 nm, which was analogous to that of the second tryptophan model indicating 21R assignment. Opposite CD results were obtained for 8 indicating 21S designation. As a next consideration it seemed useful to exploit a parallel pattern in the chiroptical data we observed between the jasplakinolide S (7 and 8) and P (5 and 6) diastereomers. These observations concerning optical rotation values and the conclusions drawn are as follows: (a) One isomer of each pair exhibited a relatively small optical rotation (7: [α]_D = +36.8, MeOH; 6: [α]_D = +39.2, MeOH); (b) One isomer of each pair exhibited a relatively large optical rotation (8: ([α]_D = +62.6, MeOH; 5: [α]_D = +61.6, MeOH); (c) The configurations established above for 7 and 8 appeared to be reasonable models to provisionally define the benzoxazinone ring configurations of 5 and 6 respectively as 21S and 21R; (d) Lastly, the configurational information shown at all other sites for 5 – 7 were based on biosynthetic analogy to 1.

Figure 5.

Cotton effects (CE) exhibited by synthetic 3-hydroxy-S-tryptophan derivatives (top)³⁸ and CD spectra of jasplakinolide S epimers 7 (green) and 8 (blue).

An especially unique analogue, jasplakinolide T (9), was obtained through LCESIMS-guided isolation by focusing on [M+H]⁺ peaks of m/z 644.3. Its molecular formula was established as C₃₆H₄₅N₅O₆, which differed from 1 by the absence of a Br and the presence of an additional N atom. Existence of the Group 1 framework was clear because of the diagnostic shift for C-31 (δ_H/C 1.51/17.6) and Δδ_C’s < 1 ppm vs. those of 1 were observed for all of the remaining resonances of the core carbons. Defining the constitution of the R₁ appendage of the Group 1 structure, whose formula consisted of C₉H₇N₂ (7 sites of unsaturation), was addressed next. The ¹H/¹³C NMR data readily indicated this to be a 4-alkylquinazoline group fully consistent with the ¹J_CH = 201 Hz at C-28 (δ_C 153.4) as observed by 2D NMR (see Figure S1, Supporting Information). Additional sets of 2D gHMBC and gCOSY NMR data represented in Figure 6 further reaffirmed this proposal. One pattern involved gHMBC correlations from H-20 (δ_H 3.92) to C-21 (δ_C 167.6) and to C-22 (δ_C 123.4). Another consisted of ³J_H-C gHMBC correlations from H-28 (δ_H 9.08) to C-21 (δ_C 167.6) and to C-27 (δ_C 148.8). Also shown in Figure 6 are supporting data for the close model, 6,7-dimethoxy-4-propylquinazoline (14, also see Figure S2, Supporting Information). The stereostructure of 9 shown here is based on analogy to that of 1.

Figure 6.

Selected 2D NMR (600/150 MHz, DMSO-d₆) correlations and δ_H/C values of the 4-alkyquinazoline substructure of 9 and of the standard 6,7-dimethoxy-4-propylquinazoline (14).

The isolation of jasplakinolide U (10, C₃₈H₄₈N₄O₁₀) was prioritized once it was recognized that there was no Br atom present. An initial inference based on this single observation was that a modified indole array was present, similar to that seen in 5, 6, 7, and 8. The isolation of 10 was completed by tracking the LCMS peak having [M+H]⁺ m/z = 707. The incorporation of a Group 2 core was set from the diagnostic shift for C-31 (δ_H/C 5.82 & 6.15/129.1) and Δδ_C’s < 1 ppm vs. those of jasplakinolide B (11)¹⁶ were observed for all of the remaining resonances of the core carbons including that of the enone: C-12 (10: δ_C 207.2; 11: δ_C 207.2).¹⁶ The final structural element R₁ was determined to be a benzoxazinone appendage, analogous to that in 5 and 6: (δ_C-OCH3 5: 51.2; 6: 51.1; 10: 51.1 and δ_C-28 5: 152.2; 6: 152.0; 10: 152.2). As expected, a set of 2D NMR correlations parallel to those shown in Figure 2 were observed for 10. The pattern of optical rotations discussed above for 7 ([α]_D = +36.8°, MeOH) and 6 ([α]_D = +39.2°, MeOH) vs. that for 10 ([α]_D = +39.2°, MeOH) provided a putative R configuration at C-21, and, as above, the chirality shown at all other sites is based on biosynthetic analogy to 1.

The final pair of diastereomers isolated (12 and 13) possessed the same molecular formula as jasplakinolide C (C₃₆H₄₅BrN₄O₇), first described by Zampella in 1999.¹⁶ Challenges faced when comparing our material to this compound included the lack of a defined configuration at C-12 in the original description and the inadvertent switch of structures between jasplakinolides B and C in the last three reports of jasplakinolide congeners.^20–22 Clearly, both 12 and 13 possessed the Group 2 framework (12: δ_H-31/C-31 4.95, 5.04/113.1; 13: δ_H-31/C-31 4.97, 5.12/111.2), and they differed from 11 by an additional degree of saturation consistent with the presence of an allylic alcohol constellation at C-11—C-12—C-31 vs. the enone of 11. Further, the NMR shift positions for protons and carbons at these sites had the expected δ values in 12 and 13. Resolving the configuration at C-12 was not straightforward because most of the Δδ_C data between this pair was less than 1.6 ppm. Also, as shown in Figure 7, the Δδ_C’s for atoms in or near the allylic alcohol moiety (C-10 to C-14) were small, with the largest at C-14 (Δδ_C = 5.8), two bonds away from the chiral center. In spite of this circumstance the acquisition of NOESY data provided useful correlations such as: 12: H-12 (δ 3.81) to H-13 (δ 1.65), and 13: H-12 (δ 3.72) to CH₃-32 (δ 0.93). These data showed that for jasplakinolide C_a (12), H-12 and H-13 were on a common face indicating that C-12 OH was α, and for jasplakinolide C_b (13), H-12 and CH₃-32 were on a common face indicating that C-12 OH was β. Defining the absolute chirality at this position for both compounds could now be addressed by derivatization using modified Mosher’s method to form 12a, 12b, 13a, and 13b as seen in Figure 8.^{39, 40} The straightforward analysis of Δδ^SR concluded 12 possessed 12R configuration and 13 was the 12S diastereomer. At this point the CH₃-32 absolute configuration could be assigned as R and the configurations at the remaining sites were concluded to be unchanged relative to 1. Still unresolved was the relationship between 12 and 13 vs. the previously described jasplakinolide C. No headway could be made on this issue as a comparison of δ_C’s between 12 and 13 vs. literature (see Table S2, Supporting Information) indicated the substantial differences shown in Figure 7 and at other sites (Δδ_C) including: C-4 (5), C-29 (3), C-30 (5), C-31 (8), and C-32 (7).

Figure 7.

NMR Δδ_C shift differences (ppm) between literature values of jasplakinolide C¹⁶ vs. experimental values of jasplakinolide C_a (12) and jasplakinolide C_b (13).

Figure 8.

Modified Mosher analysis, Δδ^SR (Hz), of 12a, 12b, 13a, and 13b prepared from jasplakinolide C_a (12) and jasplakinolide C_b (13) respectively (for framework numbering see 1).

At completion of isolation and identification of compounds 1, 3 – 13, a two pronged approach was used to evaluate their biological properties. This consisted of the following assays: (1) cytotoxicity assessment in the National Cancer Institute-Developmental Therapeutics Program (NCI-DTP) 60 cell line screen and (2) evaluation in the microfilament (MF) disruption assay in the UCSC-Chemical Screening Center (UCSC-CSC). We have demonstrated the value of such a parallel approach for 18-epi-latrunculol which exhibited cytotoxicity (IC₅₀ = 2.1 µM against HCT-116 cells) without MF disruption at 5 µM.⁴¹ NCI evaluation of seven compounds (1, 3, 5 – 8, and 11) against the NCI-DTP 60 cell line panel resulted in GI₅₀, TGI, and LC₅₀ data (see Table S3 – Table S9, Supporting Information)⁴² with selected GI₅₀ values from this assessment shown in Table 4. Based on results reported in Table 4 and Table 5, two compounds (3 and 11) were selected by the NCI Biological Evaluation Committee for further study in the hollow fiber assay, pending availability of more material. Further, as seen in Table 4 and Table 5, compound 3 exhibited slightly enhanced inhibition of leukemia cells (RPMI-8226) while compound 11 exhibited significant inhibition of both RPMI-8226 and colon adenocarcinoma (HCT-116) cells. Overall, these results shown in Table 4 and Table 5 counter the previously reported²⁰ cytotoxicity findings demonstrating greatest human breast adenocarcinoma (MCF-7) cell inhibition by 1 (IC₅₀ = 0.019 µM) and 3 (IC₅₀ = 0.02 µM) with modest inhibition by compound 11 (IC₅₀ = 3.4 µM).

Table 4.

Bioassay Data for 1, 3 – 13

		GI50 (µM)		MF Disruption Assayb
Cmpd	NSC #	HCT-116a	MCF-7a
1	613009	0.1	Not Tested		+
3	731257	0.14	0.18	−
4		Not Tested		+
5	731256	0.35	4.9	−
6	731253	0.38	2.4	−
7	731254	0.81	2.3	−
8	731255	> 10	> 10	−
9		Not Tested		−
10		Not Tested		−
11	731258	< 0.001	0.13	−
12		Not Tested		−
13		Not Tested		−

^aGI₅₀ data provided by the National Cancer Institute-Developmental Therapeutics Program (NCI-DTP). For the comprehensive data set against 60 cell lines use the NSCs above at http://dtp.nci.nih.gov/.

^bThe microfilament-disrupting effects were evaluated in HeLa cells: (+) active at 80 nM; (−) inactive at 80 nM.

Table 5.

Selected NCI 60 Cell Screen Testing Results for Jasplakinolide (1), Jasplakinolide E (3), and Jasplakinolide B (11)

	1	3	11
Cells	GI50 (µM)	GI50 (µM)	GI50 (µM)
RPMI-8226	0.031	0.022	0.0019
HOP-62	0.15	0.14	0.14
HCT-15	0.69	0.97	6.6
SF-539	0.30	0.17	0.064
M14	0.056	0.078	0.053
OVCAR-3	0.040	0.53	0.11
786-0	0.020	0.18	0.51

As stated above, all compounds (1, 3 – 13) were screened in the UCSC-CSC HeLa (human cervical cancer) cell MF disruption assay. Table 4 shows the results of the phenotypic assay, with the last column indicating changes in actin structural organization in the presence of 80 nM of each compound (see Figure S13, Supporting Information). Two compounds (1 and 4) exhibited microfilament disruption at 80 nM as seen in Figure 8. Jasplakinolides A – P have been previously tested for microfilament disruption and it was shown that antimicrofilament activity parallels cytotoxicity.²² Alternatively, the NCI-DTP 60 cell screen data in combination with the UCSC-CSC microfilament disruption assay data showed that 11 exhibits significant cytotoxicity against HCT-116 colon adenocarcinoma cells (< 1 nM) without microfilament disruption at 80 nM (Figure 8). Both 1 and 11 possess identical scaffolds in the nonribosmal peptide synthetase (NRPS) region but they differ by the enone group in the PKS region. Overall, these results are analogous to the above mentioned reports that oxidation or derivatization of the brominated tryptophan or β-tyrosine groups decrease potency,^{22, 25, 43–46} replacement of Ala for Ser does not significantly decreases potency,²⁰ and limited modifications in the PKS region are tolerated.^{22, 44, 45} Complementary conclusions were gained from SAR data published by two different groups on four synthetic chondramide C diastereomers obtained through total synthesis. A binding site analysis of synthetic 2 identified the presence of the double bond and the 15R configuration as two key elements required for optimal activity.⁴⁷ Alternatively, the 2 diastereomer with 13S, 15S configurations also exhibited similar conformation and bioactivity to natural 2.⁴⁸ In conclusion, further biological evaluation of jasplakinolide diastereomers having modified configurations at each chiral center is warranted.

The results reported above, including seven new compounds (6 – 10, 12, and 13), have both significant structural and biological impact. Firstly, the unique quinazoline substructure in 9 is the first reported sponge-derived natural product containing quinazoline functionality.⁴⁹ The benzoxazinone substructure of 5 and 6 and the 3-hydroxyindolin-2-one of epimers 7 and 8 represent rare tryptophan functionalization. Counter to previous findings,²² this is the first account where discrepancy between MF disruption and cytotoxicity has been exhibited by a jasplakinolide congener (11). Compound 11 exhibited considerable cytotoxicity (GI₅₀ < 1 nM vs. HCT-116 cells) without MF disruption at 80 nM, which has prompted both hollow fiber assessment by the NCI-DTP and further microfilament investigation at the UCSC-CSC. At this point it is difficult to explain the pattern of activity or inactivity pertaining to MF disruption results of Table 4. The presence of an exo double bond may be a factor toward contributing to the inactivity of compounds 11 – 13. Extensive tryptophan modifications were exhibited in compounds 5 – 10 and these changes greatly diminish the MF binding effects. Also of high impact is the loss of activity through the simple replacement of Ala (in 1) by Ser (in 3). Finally, changing R₃ from CH₃ (in 1) to H (in 4) has no impact on the MF disruption activity. It is tantalizing to ponder the results from a recent molecular genetics study of the pathways responsible for producing 2 by C. crocatus.^{4, 50} An interesting point to resolve in regard to this study pertains to the different mechanisms at work to produce PKS domains differing in geometry and constitution in the C-13 to C-15 region. Consequently, we have been reinvigorated to obtain and study genomic DNA from jasplakinolide-producing and non-producing taxa including J. splendens and A. cf. constricta available in our repository.

Experimental Section

General Experimental Procedures

The optical rotations were determined on a Jasco DIP 370 digital polarimeter and UV data were obtained on an Agilent 8453 UV/Vis spectrophotometer. All NMR spectra were recorded in CD₃OD, CDCl₃, or DMSO-d₆ with a 5 mm triple resonance (HCN) probe. Chemical shifts are reported in ppm relative to CD₃OD (δ_H 3.31 and δ_C 49.0), CDCl₃ (δ_H 7.27), or DMSO-d₆ (δ_H 2.50 and δ_C 39.5). A Mariner ESITOF mass spectrometer was used for low- and high-resolution mass measurements. Preparative reversed-phase (RP) separation was carried out utilizing a Waters 600E system controller and pumps with a Prep LC 25 mm radial compression column using 25 × 100 mm C18 Nova-Pak HR16 (6 µm) cartridges. Both ELSD and UV (254 nm) were used for peak detection. Semipreparative RP HPLC used a Phenomenex Synergi C18 4 µm column, 10 × 250 mm and UV peak detection (254 nm). Compound purity (>95%) was confirmed using both ¹H and ¹³C NMR and LCMS (UV and ELSD detection) experiments.

Animal Material

Samples of Auletta sp. sponges (coll. nos. 02137, 1.9 kg wet weight; and 02118, 2.0 kg wet weight) were collected in Papua New Guinea (S 9°43.969’, E 150°44.412’) June 2002. Samples of Jaspis splendens (coll. no. 00101, 1.9 kg wet weight) were collected in Fiji (S 18°22.28’, E 177°58.87’) February 2000. Taxonomic identification Auletta sp. and Jaspis splendens were performed by Dr. Rob van Soest of the Zoological Museum of Amsterdam. Underwater pictures are in Figure S12, Supporting Information, and voucher specimens are available from the corresponding author (P.C.).

Extraction and Isolation

All three sponge materials were preserved in the field in 1:1 CH₃OH : H₂O, decanted, and shipped back to UC Santa Cruz where they were then submersed in 100% CH₃OH and stored at 4 °C. For extraction, the CH₃OH was decanted and the sponge was soaked in 1L four more times to obtain the crude methanol extract. The extract was partitioned between hexanes and CH₃OH-H₂O (9:1). After separation from the hexane layer, H₂O was added to the CH₃OH-H₂O layer adjusting the solution to 1:1. The CH₃OH-H₂O layer was then extracted with CH₂Cl₂ to give the crude (FD) extract (2.8 g for sponge coll. no. 02137, 1.6 g for sponge coll. no. 02118 and 2.1 g for 00101 for sponge coll. no. 00101).

The FD extract of 02137 was subjected to preparative reversed-phase (RP) HPLC (10 – 70% CH₃CN : H₂O, 60 min) yielding eight fractions (H1 – H8). Fraction H4 (79.4 mg) and H5 (98.9 mg) were combined and subjected to semiprepartive RP HPLC using isocratic conditions (19:31, CH₃CN : H₂O) to afford 7 (4.1 mg, H4 H8 H2), 8 (2.5 mg, 02137 FD H45 H4 H3), 12 (5.6 mg, H45 H10 H3) and 13 (2.6 mg, H45 H11 H3). Fraction H6 (83.9 mg) was subject to multiple rounds of RP HPLC utilizing isocratic conditions (21:29, CH₃CN : H₂O) to afford 3 (1.5 mg, H6 HW H2 H2 H3 H2 H2), 4 (1.1 mg, H6 H2 H3), 5 (2.9 mg, H6 H3 H2), 6 (2.5 mg, H6 H2 H2), and 11 (4.0 mg, H6 H5 H3). Fraction H7 (221.1 mg) was purified by repetitive RP HPLC over isocratic conditions (11:13, CH₃CN : H₂O) to yield 1 (70.1 mg, H7 H1 H2).

The FD extract for 02118 was subjected to preparative RP HPLC (10 to 65% CH₃CN : H₂O, 60 min) yielding eight fractions (H1 – H8). Fraction H2 (22.9 mg) was subjected to three rounds of RP HPLC employing isocratic conditions (9:16, CH₃CN : H₂O) to yield compound 10 (4.3 mg, H2 H6 H2 H2).

Similarly, 00101 FD was subjected to preparative RP HPLC (10 – 65% CH₃CN : H₂O, 50 min) yielding 11 fractions (H1 – H11). Fraction H6 (133.2 mg) was subjected to multiple rounds of semipreparative RP HPLC using isocratic conditions (43% CH₃CN : H₂O, 40 min) to yield 9 (2.3 mg, H6 H11 H3 H3 H3). Fraction H8 (159.2 mg) was separated via multiple rounds of isocratic (77% CH₃OH : H₂O, 45 min) RP HPLC to yield 4 (1.9 mg, H8 H5 H5).

HeLa Cell Microfilament Disruption Assay

HeLa Cells were plated in 384 well tissue culture treated plates (Corning) at a density of 1500 cells per well. After incubating at 37 °C with 5% CO₂ overnight compounds were pinned into plates using a Janus MDT (PerkinElmer) automated liquid handler. After 24 hours cells were fixed in 4% formaldehyde for 20 minutes then washed with PBS using an automated plate washer (BioTek). The cells were then treated with 0.5% TritonX-100 in PBS for 10 minutes, washed, and then blocked with a 2% BSA PBS solution for 20 minutes. Actin was stained with rhodamine-phalloidin for 20 minutes and then washed. Lastly, the DNA was stained with hoechst 33342 (AnaSpec Inc.) followed by a wash with the automated plate washer. Images were taken using an automated fluorescence microscope (Image×press, MDS) at 10× magnification.

Jasplakinolide P (5)

colorless glass (2.9 mg). [α]²⁷_D = +61.6° (c, 0.062, MeOH); UV (MeOH) λ_max (log ε): 227 nm (3.19), 244 nm (2.87), 277 nm (2.50); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+H]⁺ m/z 693.3491 (calculated for C₃₇H₄₉N₄O₉, 715.3314).

21-Epi-Jasplakinolide P (6)

colorless glass (2.5 mg). [α]²⁷_D = +39.2° (c, 0.07, MeOH); UV (MeOH) λ_max (log ε): 228 nm (3.07), 244 nm (2.77), 278 nm (2.23); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+Na]⁺ m/z 715.3321 (calculated for C₃₇H₄₈N₄O₉Na, 715.3314).

Jasplakinolide S (7)

colorless glass (4.1 mg). [α]²⁷_D = +36.8° (c, 0.056, MeOH); UV (MeOH) λ_max (log ε): 230 nm (2.70), 253 nm (2.33); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+H]⁺ m/z 663.3392 (calculated for C₃₆H₄₇N₄O₈, 663.3388); CD (MeOH) λ_max (Δε): 242 nm (−16), 264 nm (+10), 292 nm (+3).

21-Epi-Jasplakinolide S (8)

colorless glass (2.5 mg). [α]²⁷_D = +62.6° (c, 0.054, MeOH); UV (MeOH) λ_max (log ε) 230 nm (2.75), 253 nm (2.43); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+Na]⁺ m/z 685.3209 (calculated for C₃₆H₄₆N₄O₈Na, 685.3208); CD (MeOH) λ_max (Δε): 239 nm (+32), 264 nm (−12), 293 nm (−1).

Jasplakinolide T (9)

white powder (2.3 mg). [α]²⁷_D = +26° (c, 0.056 MeOH); UV (MeOH) λ_max (log ε) 202 nm (4.10), 223 nm (3.98) 274 nm (3.07); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+H]⁺ m/z 644.3495 (calculated for C₃₆H₄₆N₅O₆, 644.3442).

Jasplakinolide U (10)

colorless glass (4.3 mg). [α]²⁷_D = +39.4° (c, 0.054, MeOH); UV (MeOH) λ_max (log ε): 228 nm (3.07), 244 nm (2.77), 278 nm (2.23); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+H]⁺ m/z 707.3315 (calculated for C₃₇H₄₇N₄O₁₀, 707.3287).

Jasplakinolide C_a (12)

colorless glass (4.3 mg). [α]²⁷_D = +37.0° (c, 0.054, MeOH); λ_max (log ε) 220 nm (3.41), 281 nm (2.61); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+H]⁺ m/z 707.3315 (calculated for C₃₇H₄₇N₄O₁₀, 707.3287).

Jasplakinolide C_b (13)

colorless glass (2.6 mg). [α]²⁷_D = +62.3° (c, 0.058 MeOH); λ_max (log ε) 222nm (3.32), 280nm (2.70); ¹H and ¹³C NMR (see Table 2 and Table 3); HRESITOFMS [M+Na]⁺ m/z 747.2373 (calculated for C₃₆H₄₅N₄O₇NaBr, 747.2364).

Reaction of 12 to Form 12a and 12b

To a solution of 1.8 mg of jasplakinolide C_a (12) 20 µL of (+)-α-methoxy-α-trifluoromethylphenylacetic chloride (MTPA-Cl) was added, followed by 100 µl of pyridine. The reaction mixture was stirred at room temperature for 30 min and dried under N₂. The residue was then purified by RP HPLC using CH₃CN : H₂O (22:3; isocratic) yielding 1.1 mg of the (R)-MTPA ester 12a. The same procedure was used with 2.0 mg of 12 to furnish 1.3 mg of product for the (S)-MTPA ester 12b. In both cases the secondary as well as phenolic hydroxyl groups reacted with the MTPA-Cl.

(R)-MTPA Ester 12a

Obtained as a colorless glass. ¹H NMR (CD₃OD) δ_H 0.76 (3H, d, J = 6.5 Hz, CH₃-32), 0.84 (3H, d, J = 7 Hz, CH₃-29), 1.05 (3H, d, J = 6.5 Hz, CH₃-33), 1.13 (3H, d, J = 7 Hz, CH₃-30), (1H, m, H14a), 1.49 (1H, m, H14b), 1.87 (1H, m, H13), 2.17 (1H, dd, J = 15.5, 6.5 Hz, 10a), 2.41 (1H, dd, J = 16.5, 9 Hz, H10b), 2.61 (1H, dd, J = 14.8, 6.8 Hz, H9), 2.77 (1H, dd, J = 17.3, 5.3 Hz, H2a), 2.99 (1H, dd, J = 16.8, 8.8 Hz, H2b), 3.07 (1H, dd, J = 15, 10 Hz, H20a), 3.12 (3H, s, CH₃-34), 3.38 (1H, dd, J = 14.5, 6 Hz, H20b), 3.47 (3H, s, OMe), 3.68 (3H, s, OMe), 4.41 (1H, q, J = 7 Hz, H7), 4.96 (1H, s, H31a), 5.08 (1H, s, H31b), 5.16 (1H, dd, J = 6.5, 5 Hz, H3), 5.23 (1H, d, J = 4.5 Hz, H12), 5.59 (1H, dd, J = 10, 6.5 Hz, H5), 6.96–7.63 (9H, Ar)

(S)-MTPA Ester 12b

Obtained as a colorless glass. ¹H NMR (CD₃OD) δ_H 0.82 (3H, d, J = 7 Hz, CH₃-29), 0.82 (3H, d, J = 7 Hz, CH₃-32), 1.07 (3H, d, J = 7 Hz, CH₃-30), 1.11 (3H, d, J = 6 Hz, CH₃-33), 1.23 (1H, m, H14a), 1.62 (1H, m, H14b), 1.89 (1H, m, H13), 2.06 (1H, dd, J = 14.5, 5 Hz, H10a), 2.36 (1H, dd, J = 15, 8 Hz, H10b), 2.58 (1H, dq, J = 13.5, 6.5 Hz, H9), 2.74 (1H, dd, J = 16.8, 5 Hz, H2a), 3.00 (1H, dd, J = 17, 9 Hz, H2b), 3.07 (1H, dd, J=15, 10 Hz, H20a), 3.12 (3H, s, CH₃-34), 3.38 (1H, dd, J = 15, 6 Hz, H20b), 3.49 (3H, s, OMe) 3.68 (3H, s, OMe), 4.39 (1H, q, J = 6.5 Hz, H7), 4.80 (1H, s, H31a), 4.84 (1H, H15), 5.00 (1H, s, H31b), 5.15 (1H, dd, J = 9, 5 Hz, H3), 5.18 (1H, d, J = 4.5 Hz, H12), 5.61 (1H, dd, J = 9.8, 5.8 Hz, H5), 6.97–7.63 (9H, Ar).

Reaction of 13 to Form 13a and 13b

To a solution of 1.1 mg of jasplakinolide C_b (13) 20 µL of (+)-α-methoxy-α-triflouromethylphenylacetic chloride (MTPA-Cl) was added, followed by 100 µl of pyridine. The reaction mixture was stirred at room temperature for 30 min and dried under N₂. The residue was then purified by RP HPLC using CH₃CN:H₂O (22:3; isocratic) yielding 0.7 mg of the (R)-MTPA ester 13a. The same procedure was used with 1.0 mg of 13 with (−)-MTPA-Cl to furnish 0.5 mg of the (S)-MTPA ester 13b. In both cases the secondary as well as phenolic hydroxyl groups reacted with the MTPA-Cl.

(R)-MTPA Ester 13a

Obtained as a colorless glass. ¹H NMR (CD₃OD) δ_H 0.80 (3H, d, J = 7.0 Hz, CH₃-29), 0.91 (3H, d, J = 6.5 Hz, CH₃-32), 1.02 (1H, m, H14a), 1.09 (3H, d, J = 7.0 Hz, CH₃-30), 1.11 (3H, d, J = 6.0 Hz, CH₃-33), 1.69 (1H, m, H14b), 1.99 (1H, m, H13), 2.31 (2H, d, J = 6.0Hz, CH₂-10), 2.57 (1H, dd, J = 13.5, 6.5 Hz, H9), 2.74 (1H, dd, J = 17.5, 4.5 Hz, H2a), 2.97 (1H, dd, J=17.5, 9.5 Hz, H2b), 3.08 (1H, dd, J = 14.8,10.3 Hz, H20a), 3.13 (3H, s, CH₃-34), 3.52 (3H, s, OMe), 3.68 (3H, s, OMe), 4.38 (1H, q, J = 7 Hz, H7), 4.73 (1H, s, H31a), 4.96 (1H, s, H31b), 5.00 (1H, H15), 5.10 (1H, d, J=4.5 Hz, H12), 6.97–7.63 (9H, Ar).

(S)-MTPA Ester 13b

Obtained as a colorless glass. ¹H NMR (CD₃OD) δ_H 0.80 (3H, d, J = 7 Hz, CH₃-29), 0.82 (3H, d, J = 6.5 Hz, CH₃-32), 0.91 (1H, m, H14a), 1.06 (3H, d, J = 6 Hz, CH₃-33), 1.12 (3H, d, J = 7 Hz, CH₃-30), 1.61 (1H, dd, J = 11.8 Hz, H14b), 1.96 (1H, m, H13), 2.36 (2H, d, J = 8 Hz, CH₂-10), 2.61 (1H, dd, J=12.8, 6.8 Hz, H9), 2.73 (1H, dd, J = 17.8, 4.8 Hz, H2a), 2.96 (1H, dd, J = 17.5, 9.5 Hz, H2b), 3.08 (1H, dd, J = 15, 10 Hz, H20a), 3.13 (3H, s, CH₃-34), 3.35 (1H, dd, J = 15.3, 5.8 Hz, H20b), 3.47 (3H, s, OMe), 3.68 (3H, s, OMe), 4.39 (1H, q, J = 7 Hz, H7), 4.93 (1H, s, H31a), 4.97 (1H, H15), 5.05 (1H, s, H31b), 5.13 (1H, d, J = 4.5 Hz, H12), 5.19 (1H, dd, J = 9.5, 4.5 Hz, H3), 5.58 (1H, dd, J = 10, 6 Hz, H5), 6.98–7.63 (9H, Ar).

Figure 9.

Influence of jasplakinolide (1), jasplakinolide F (4), jasplakinolide B (11) and DMSO (negative control) at 80 nM on the actin cytoskeleton of HeLa cells after 24 hour incubation. The F-actin was labeled red (rhodamine-phalloidin) and nuclei and chromosomes labeled green (hoechst 33342, AnaSpec Inc.).

Abbreviations

PKS

polyketide synthase

NRPS

nonribosomal peptide synthetase

CD

circular dichroism.