1312

Agriculture (USDA, Beltsville, MD) from collectors worldwide. The extracts were sent to screening contractors for biological evaluation. Usually the 9KB (human nasopharyngeal carcinoma) in vitro cytotoxicity assay and the 3PS (P388) (methylcholanthrene-induced murine leukemia) in vivo assay were used. Turn-around times for these bioassays were length; 9KB required about one month, while 3PS required up to six months, and likely as not, the 3PS results were ambiguous and required retesting, which consumed precious time and materials. Nonetheless, the 3PS system, when it worked, provided leads that translated into actives in animal solid tumor systems. Most of the antitumor botanicals in clinical use today were found by following in vivo activities in such murine leukemias. The contract work provided us with several actives that were processed using cytotoxicities and 3PS for activity-directed fractionation;4 we eventually established our own cytotoxicity panel through the Cell Culture Laboratory of the Purdue Cancer Center. The turn-around times were, thus, reduced to one week, but the costs were high (about 120 per sample for six cell lines). These costs would soon deplete our research budget if this were the only available assay.

Cytotoxicity results often do not translate into in vivo activities. Millions of dollars have been spent detecting and isolating in vitro active compounds that are found later to be inactive when tested in vivo. Anticancer research may have been set back 30 years by the decision in the early 1980s to abandon in vivo screening at NCI and replace it with the in vitro panel of 60 human tumor cell lines. Even the notion that selectivity of susceptible tumor types, as suggested by such in vitro tests, has any meaning in predicting in vivo activity in the same tumor type dogmatically persists and remains unproven. Consequently, the use of so many cell lines is probably a waste of money, and, indeed, this has more recently been reduced to only three cell lines for initial screening. A satisfactory replacement for the in vivo murine leukemias has still not emerged. The high cost of using athymic mice, bearing xenografts of human solid tumors, precludes their use in academic research in monitoring the dozens of fractions encountered in natural product, bioactivity-directed, isolation work, so we felt a proven bioassay that could predict in vivo activity was sorely needed.

Murine toxicity is a frequent problem encountered when using the 3PS assay itself. Extracts that killed the mice were sometimes abandoned when a little more work with further dilutions would have led to doses within a therapeutic index. Such was the case with several of our most potent plant species, e.g., Goniothalamus giganteus Hook. f. and Thomas (Annonaceae), Asimina triloba (L.) Dunal (Annonaceae), and Kalanchoe tubiflora (Harv.) Hamet. (Crassulaceae). Consequently, these “toxic” species from the contract work were abandoned by NCI and turned over to my laboratory by the late Matt Suffness, the NCI contract officer, for fractionation to identify possible new pesticidal components. At this time (1980), I had just completed a sabbatical leave at the USDA laboratory (Peoria, IL); while there I had struggled with a laborious bioassay employing European corn borers to isolate some pesticidal cardenolides from Thevetia thevetioides (HBK.) K. Schum. (Apocynaceae).5 I had also waited patiently for 3PS assay results to guide the first isolation of 10-deacetylbaccatin III, 10-deacetyltaxol, and 10-deacetylcephalomannine from Taxus wallichiana Zucc. (Taxaceae).6

These experiences convinced me that the bioassays, as being employed by the NCI contractors, were the major road block in such fractionation work. Consequently, I directed my small research team to explore a number of alternatives. Brian Meyer developed the brine shrimp lethality bioassay,7 and Nelson Ferrigni perfected the potato disk bioassay (inhibiting crown gall tumors).8 Matt Suffness provided us with a blind set of test compounds, and Jon Anderson quickly demonstrated that these tests were statistically predictive of 3PS activity.9 These two benchtop methods soon began leading us to a diversity of 3PS active compounds. For example, the stilbene piceatannol,10 isolated from Euphorbia lagascae Spreng. (Euphorbiaceae), is now being studied as a protein tyrosine kinase inhibitor.11 Pamela Boner, with Nelson Ferrigni as her laboratory mentor, isolated two new bufodienolides from the murine toxic extracts of Kalanchoe tubiflora. The success of these bioassays eventually led to R01 grant support (1984–1997) from NCI, and I subsequently taught these procedures12 in 22 workshops worldwide sponsored by UNESCO, IOCD, CONICET, and other agencies. The brine shrimp method has now been cited thousands of times in the natural product literature. Using these simple bioassays, we were able to isolate over 350 significantly cytotoxic compounds, which are described in four successive review papers.13-16 In addition to the several hundred plant accessions that were accumulated during the NCI contracts, we were able to obtain through Burroughs Welcome and Co. (Tuckahoe, NY) a part of the plant collection of the late Morris Kupchan. From the USDA (Peoria, IL) and FMC Corporation (Princeton, NJ), we were given hundreds of additional, previously unscreened, species that had originally been collected by the USDA (Beltsville, MD) for the NCI. Eventually, over 4000 accessions were accumulated and housed in the Purdue warehouses. About 3500 of these were extracted, screened for bioactivity, and submitted under contracts to Eli Lilly and Company (Greenfield, IN), Merck Pharmaceuticals (Princeton, NJ), FMC Corporation (Princeton, NJ), and/or Xenova Limited (Berkshire, UK), where they were incorporated into libraries for high-throughput drug and pesticidal screens. At the time of my retirement (1999), Tom McCloud came from NCI (Frederick, MD) and salvaged some 500 species that had never been screened. The remainder of this huge plant collection was unfortunately later destroyed. At least 56 uninvestigated, active, species were among the accessions lost.

the stilbene piceatannol,10 isolated from Euphorbia lagascae Spreng. (Euphorbiaceae), is now being studied as a protein tyrosine kinase inhibitor.11 Pamela Boner, with Nelson Ferrigni as her laboratory mentor, isolated two new bufodienolides from the murine toxic extracts of Kalanchoe tubiflora. The success of these bioassays eventually led to R01 grant support (1984–1997) from NCI, and I subsequently taught these procedures12 in 22 workshops worldwide sponsored by UNESCO, IOCD, CONICET, and other agencies. The brine shrimp method has now been cited thousands of times in the natural product literature. Using these simple bioassays, we were able to isolate over 350 significantly cytotoxic compounds, which are described in four successive review papers.13-16 In addition to the several hundred plant accessions that were accumulated during the NCI contracts, we were able to obtain through Burroughs Welcome and Co. (Tuckahoe, NY) a part of the plant collection of the late Morris Kupchan. From the USDA (Peoria, IL) and FMC Corporation (Princeton, NJ), we were given hundreds of additional, previously unscreened, species that had originally been collected by the USDA (Beltsville, MD) for the NCI. Eventually, over 4000 accessions were accumulated and housed in the Purdue warehouses. About 3500 of these were extracted, screened for bioactivity, and submitted under contracts to Eli Lilly and Company (Greenfield, IN), Merck Pharmaceuticals (Princeton, NJ), FMC Corporation (Princeton, NJ), and/or Xenova Limited (Berkshire, UK), where they were incorporated into libraries for high-throughput drug and pesticidal screens. At the time of my retirement (1999), Tom McCloud came from NCI (Frederick, MD) and salvaged some 500 species that had never been screened. The remainder of this huge plant collection was unfortunately later destroyed. At least 56 uninvestigated, active, species were among the accessions lost.

The extracts of two annonaceous plant species, Asimina triloba (paw paw) and Goniothalamus giganteus, that had been toxic to the injected leukemic mice in the 3PS assay were sent in 1982 to Eli Lilly and Company (Greenfield, IN) for screening in a panel of seven indicator agricultural pests. The extracts of paw paw were more potent and were surprisingly very active against five of the seven pest species.17 Kent Rupprecht, with Tom McCloud as his laboratory mentor, used the brine shrimp assay to isolate asimicin18 (1), our first annonaceous acetogenin, as one of the major bioactive components of the bark. At the same time, Mike Mikolajczak at the USDA (Peoria, IL) isolated 1 from the seeds of paw paw rather than from the bark. Selective 1H decoupling using 200 MHz NMR helped to determine the regiochemistry by locating the molecule’s third hydroxyl at C-4 and ultimately solving or confirming the molecular structure. Compound 1 was 3PS and L1210 (murine leukemia) active, very active (with ED50 values <10-12 μg/mL) in cytotoxicity tests, immunosuppressive, antimalarial, and strongly pesticidal. Two U.S. patents were issued protecting the use of the acetogenins as pesticides and the composition of matter of 1.19,20

Ahmad Alkofahi fractionated the toxic extracts of Goniothalamus giganteus using the brine shrimp assay and quickly isolated goniothalamicin (2) and annonacin (3), two single tetrahydrofuran (THF) ring acetogenins, each, like 1, bearing a hydroxyl at C-4.21 John Cassady’s group, following cytotoxicity assays, had previously found 3 in Annona densicoma Mart.22 Compounds 1–3 were all cytotoxic, pesticidal, and antimalarial (Walter Reed Army Hospital, Washington, DC), and 1 and 3 were active in vivo in 3PS. We next solicited bark samples of annonaceous trees and shrubs from all over the tropical world; brine shrimp tests revealed that over 50 of the 80 some species evaluated were active and worthy of future fractionation. Some of these materials yielded active substances that were, surprisingly, not acetogenins.23 However, the acetogenins were found to be the most potent of all the annonaceous components. Currently the Annonaceae remains a “hot” family