Search PMN  


PDF version
for printing

© 2008 Plant Management Network.
Accepted for publication 28 May 2008. Published 22 August 2008.

Spinetoram: How Artificial Intelligence Combined Natural Fermentation with Synthetic Chemistry to Produce a New Spinosyn Insecticide

James Dripps, Senior Research Scientist, Brian Olson, Research Scientist, Thomas Sparks, Advisor, and Gary Crouse, Research Scientist, Dow AgroSciences, 9330 Zionsville Road, Indianapolis, IN 46268

Corresponding author: Brian Olson.

Dripps, J., Olson, B., Sparks, T., and Crouse, G. 2008. Spinetoram: How artificial intelligence combined natural fermentation with synthetic chemistry to produce a new spinosyn insecticide. Online. Plant Health Progress doi:10.1094/PHP-2008-0822-01-PS.

Natural products have been and continue to be an excellent source of inspiration for new insect control products. In the early 1980s, a vacationing chemist collected a soil sample from an abandoned rum still in the Caribbean as part of a program to search for soil microorganisms with unique biological activity. From this soil sample, a new species of actinomycete (Saccharopolyspora spinosa) was isolated (2). Extracts from the fermentation broth of S. spinosa showed both contact and ingestion activity against southern armyworm (Spodoptera eridana). Considering the rarity of natural products with contact activity against Lepidoptera, this discovery spurred further studies leading to the identification of a series of new macrocyclic structures (7), later named "spinosyns." Spinosyns A and D demonstrate the highest level of insecticidal activity and are produced in the greatest quantity among the many spinosyns extracted from the fermentation broth. The naturally occurring mixture of spinosyns A and D was assigned the common name of spinosad (Fig. 1).


Fig. 1. Spinosad: A naturally occurring mixture of spinosyns A and D.


The mode of action of spinosyns is not via the target sites of the avermectins, neo-nicotinoids, pyrethroids, or any other known insecticides (1). Spinosyns act through a novel site in the nicotinic receptor that is distinct from neo-nicotinoids or any other nicotinic actives (1). Selection for spinosad resistance in Drosophila melanogaster and subsequently sequencing the genes from this spinosad-resistant strain identified the spinosyn target site as an α7-like nicotinic acetylcholine receptor known as Dmα6-nAChR (3). It is the activation of this α6-nAChR by the spinosyns that begins the cascade of events leading to insect death. This mode of action is effective in controlling a variety of insect pests, including Lepidoptera, Diptera, Thysanoptera, some Coleoptera, termites, and ants.

Spinosad has low human toxicity and short environmental persistence, and its distinct pest selectivity allows it to be used in many integrated pest management programs. Spinosad was registered in 1997 under the USEPA Reduced Risk Pesticide Initiative and received the Presidential Green Chemistry Challenge Award in 1999. Several spinosad products are currently in use, including SpinTor®, Success®, Entrust®, and Conserve® insecticides, and GF-120® NF Naturalyte® fruit fly bait. The spinosad formulations Entrust® and GF-120® NF Naturalyte® fruit fly bait are listed for use on organically grown crops in the United States.

Following the discovery of spinosad, Dow AgroSciences set a goal to determine if new and different spinosyn insecticides were possible. Because the spinosyns are large, complex molecules, identifying modifications that would improve activity was difficult. Techniques such as biotransformation and genetic engineering, as well as screening for organisms that produce new spinosyns, did yield new compounds. However, none were superior to spinosad. Synthetic modification through systematic substitutions would potentially require producing a huge number of new compounds. For example, examining eight peripheral sites on the spinosyn molecule with four potential substitutions per site would result in more than 65,000 possible chemical combinations. During the early years of the effort, over 400 semi-synthetic spinosyn analogs were made and, with one exception, none showed improved activity compared to spinosad. Early attempts at quantitative structure-activity relationship (QSAR) modeling were not successful in identifying synthetic changes to improve activity.

During a trip to the West Coast, a Dow AgroSciences scientist met with a friend who happened to be working on a robotic vacuum cleaner that used an artificial neural network (ANN) as a means to learn the layout of an owner’s house (4). The scientist recognized artificial neural networks as an alternative approach to classic QSAR analysis. Artificial neural networks are a form of software-based artificial intelligence, essentially a learning machine that mimics neural connections of the brain. Artificial neural networks are very good at pattern recognition and work well with incomplete data (5). Using ANN-based QSAR modeling to analyze the spinosyns, the 2′,3′,4′-tri-O-ethyl analog of spinosyn A was identified as having strong potential for improved biological activity. Further ANN analysis determined that the 3′-O-ethyl group was the most potent in altering nicotinic function in the insect nervous system, showing improved activity against corn earworm (Helicoverpa zea) and beet armyworm (Spodoptera exigua) (Fig. 2) (5,6).


Fig. 2. Effect of rhamnose substitution pattern on the insecticidal activity of spinosyns.


Earlier structure-activity studies had suggested that hydrogenation of the 5,6 double bond in spinosyn A would improve photostability of the molecule and thereby increase residual control. The combination of a reduced 5,6 double bond and the 3′-O-ethyl group on spinosyn J showed a greater level of activity against corn earworm (Helicoverpa zea) and sweetpotato whitefly (Bemisia tabaci) than either the 3′-O-ethyl modification alone or the 5,6 double bond reduction alone (Fig. 3). Additional testing confirmed that this combination of synthetic modifications increased activity and residual control across a wide range of insect pest species. The active ingredient that resulted from combining these two synthetic modifications is spinetoram (pronounced spĭn·ET·ō·răm) (Fig. 4). The conceptual path that led to spinetoram is outlined in Figure 5.


Fig. 3. The individual and combined effects of 3′-O-ethyl modification and 5,6 double bond reduction on insecticidal activity of the spinosyns.



Fig. 4. Spinetoram: A result of synthetic modifications of spinosyns J and L.



Fig. 5. Conceptual evolution of 3′-O-ethyl-5,6-dihydro spinosyn J (principal component of spinetoram) from spinosyn A (principal component of spinosad), and the role of artificial neural network (ANN)-based modeling.


Production of spinetoram begins with the naturally occurring mixture of spinosyn J (major component) and spinosyn L (minor component), which have a reactive hydroxyl group at the 3′ position. Spinosyns J and L are both modified through the addition of the 3′-O-ethyl group and the reduction of the 5,6 double bond on spinosyn J (Fig. 4). Reduction of the 5,6 double bond in spinosyn L is hindered by the adjacent methyl group.

Spinetoram provides an excellent combination of activity and residual control while maintaining the low mammalian toxicity and short environmental persistence of the spinosyn family. A field-lab bioassay in cabbage showed improved residual control of beet armyworm and corn earworm compared to spinosad six days after application (Fig. 6). In a comparison of spinetoram (Delegate™), spinosad (SpinTor®), and azinphos-methyl (Guthion) against codling moth (Cydia pomonella) in a field-lab bioassay, spinetoram provided control of codling moth similar to azinphos-methyl, but at a lower application rate (Fig. 7). Control was superior to spinosad at the same application rate. Table 1 lists some significant toxicological and environmental fate values for spinetoram.


Fig. 6. Comparison of residual control of spinosad and spinetoram (field-lab assay, 6 days after treatment, 100 g ai/ha in cabbage).



Fig. 7. Control of codling moth (Cydia pomonella) fruit entries; Fowler, IN, 2005 (field-treated fruit collected at specified time points and infested with codling moth eggs).


Table 1. Toxicological and environmental fate values for spinetoram.

Rat acute oral LD50 >5000 mg/kg
Rat acute dermal LD50 >5000 mg/kg
Mallard duck acute LD50 >2250 mg/kg
Rainbow trout acute LC50 >3.46 mg/liter (ppm)
Earthworm acute LC50 >1000 mg/kg soil
Terrestrial dissipation half-life (field soil) 2 to 4 days*
Aquatic dissipation half-life (natural surface water) 0.5 to 0.6 days*

 * Minor factor and major factor, respectively.

Spinetoram was registered in the United States in September 2007 under the USEPA Reduced Risk Pesticide Initiative. Submission for Annex I inclusion in the European Union was completed in 2007, and EU member state submissions are expected to begin in 2008. Registration and development of spinetoram in many other countries around the world are anticipated. There are currently two spinetoram products registered in the United States: Delegate™ WG insecticide for tree fruits, tree nuts, citrus, bush berries, cane berries, and grapes; and Radiant™ SC insecticide for vegetables, strawberries, and herbs.

Spinetoram is a new, semi-synthetic spinosyn insecticide with improved efficacy, expanded spectrum of activity, improved residual control, and favorable toxicological and environmental attributes. It was discovered by recognizing and building on other discoveries in microbiology and fermentation, synthetic chemistry, and artificial intelligence. The integration of these discoveries resulted in a new and better tool for managing insect pests. The unique attributes of spinetoram and the novel approach used to discover it were recognized by USEPA when the agency presented spinetoram with a 2008 Presidential Green Chemistry Challenge Award in the category of Designing Greener Chemicals.

Trademark Symbols

® ™ Trademark of Dow AgroSciences LLC

Literature Cited

1. Crouse, G. D., Dripps, J. E., Orr, N., Sparks, T. C., and Waldron, C. 2007. DE-175 (Spinetoram), a new semisynthetic spinonsyn in development. Pages 1013-1031 in: Modern Crop Protection Chemistry. W. Kramer and U. Schirmer, eds. Wiley-VCH, Weinheim, Germany.

2. Mertz, F. P., and Yao, R. C. 1990. Saccharopolyspora spinosa sp. nov. isolated from soil collected in a sugar mill rum still. Int. J. Syst. Bacteriol. 40:34-39.

3. Orr, N., Watson, G. B., Hasler, J., Michael, J., Geng, C., Cook, K. R., Salgado, V. L., and Chouinard, S. 2006. Sequences of Drosophila melanogaster nicotinic receptor alpha-6 and alpha-7 subunits for bioassay. WO 2006091672 PCT Intl Appl. 102 pp.

4. Pope, G. T. 1993. Homer hoover. Discover. 14(3):28.

5. Sparks, T. C., Anzeveno, P. B., Martynow, J. G., Gifford, J. M., Hertlein, M. B., Worden, T. V., and Kirst, H. A. 2000. The application of artificial neural networks to the identification of new spinosoids with improved biological activity toward larvae of Heliothis virescens. Pestic. Biochem. Physiol. 67:187-197.

6. Sparks, T. C., Crouse, G. D., Dripps, J. E., Anzeveno, P., Martynow, J., DeAmicis, C. V., and Gifford, J. 2008. Neural network-based QSAR and insecticide discovery: Spinetoram. J. Comput.-Aided Mol. Des. 22:393-401.

7. Sparks, T. C., Thompson, G. D., Kirst, H. A., Hertlein, M. B., Mynderse, J. S., Turner, J. R., and Worden, T. V. 1999. Fermentation-derived insect control agents: The spinosyns. Pages 171-188 in: Biopesticides: Use and Delivery. F. R. Hall and J.J. Menn, eds. Humana Press, Totowa, NJ.