Understanding how plants produce cocaine

The traditional use of plants that promote health, that cure many diseases, or that affect the human spirit has been adopted over thousands of years by human civilizations around the globe. Their biological activities, attributed to various specialized metabolites that defend plants against the biotic environment, have found numerous applications in medicine. Today, many modern drugs such as chemotherapy agents (vinblastine and camptothecin), pain killers (codeine and morphine), and other alkaloids continue to be extracted and purified from plants where they occur at low levels, and they are difficult to isolate. As a result, understanding the molecular biology and biochemical pathways involved in their assembly has become a popular area of research, as a prelude to application of this knowledge in metabolic pathway engineering and synthetic biology. In recent years, the multistep pathways for the assembly of morphine and vinblastine have been reported, and synthetic biology efforts have generated yeast strains that accumulate each of these drugs (1, 2). In the present study, Chavez et al. (3) complete the description of the pathway for biosynthesis of the tropane alkaloid (TA) cocaine and report that it has evolved independently of the pathway for biosynthesis of other TAs such as littorine or scopolamine. Besides completion of the cocaine pathway, these discoveries offer remarkable opportunities for metabolic engineering to supply needed medical drugs not available from natural sources and for production of medicinally superior hybrid TAs not found in nature with uses as new nonaddictive therapeutics for various disease treatments, as safer surgical anesthetics or to provide possible solutions for addiction to cocaine.

Cocaine is a TA produced by several coca species (Erythroxylum spp.) found in various parts of North and South America, with South American Erythroxylum coca and Erythroxylum novogranatense being major cultivated species used for licit and illicit production (4). The chewing of coca leaves as a general stimulant by indigenous people of the Americas is well documented over thousands of years. Coca became economically important and was cultivated widely by South American civilizations, particularly during the growth and expansion of the Inca Empire (4). In more modern times, beverages and tonics containing “Coca de los Incas” appeared in the late 19th century and were later removed as their addictive properties were recognized and the beverages became broadly adopted by consumers. However, the illicit use of cocaine as a stimulant has greatly increased the level of addiction and associated deaths, particularly in North America.

“In the present study Chavez et al. (3) complete the description of the pathway for biosynthesis of the tropane alkaloid (TA) cocaine and report that it has evolved independently of the pathway for biosynthesis of other TAs such as littorine or scopolamine.”

Much is known about the biosynthesis of the TAs, littorine, atropine, and scopolamine found in members of the Solanaceae family, while the pathways leading to the formation of cocaine remain to be characterized. The 3α stereochemistry of Solanaceae TAs compared to the 3β configuration in the cocaine molecule has contributed to their different pharmaceutical effects and uses in medicine. The formation of Solanaceae TAs involves decarboxylation of ornithine to form putrescine, N-methylation to form N-methylputrescine and its conversion to N-methylpyrrolinium catalyzed by ornithine decarboxylase (5), putrescine-N-methyltransferase (PMT) (6), and N-methylputrescine oxidase (MPO) (7), respectively. The polyketide synthase condensation of N-methylpyrrolinium with malonyl-CoA triggers formation of 4-(1-methyl-2-pyrrolidinyl)-3-oxobutanoate (MPOB) (8). The conversion of MPOB to tropinone is catalyzed by tropinone synthase (CYP82M3) (8), and tropinone is reduced to tropine catalyzed by tropinone reductase (9) and esterification of tropine with 1-O-β-phenyllactoylglucose to form littorine catalyzed by a serine carboxypeptidase-like acyltransferase, littorine synthase (10).

The present study (3) tests the hypothesis that plant TA biosynthesis evolved independently in the Solanaceae and Erythroxylaceae by identifying the remaining steps for cocaine biosynthesis. Evidence for independent evolution is partly based on previous studies showing that polyketide synthases that catalyze the condensation of two malonyl-CoA esters to yield 3-oxoglutarate (Fig. 1) display different phylogenetic origins (11) and that the last two steps in cocaine biosynthesis involve an aldo-keto reductase (methylecgonine reductase) (12) and a BAHD acyltransferase (cocaine synthase) (13) (Fig. 1), rather than the short-chain dehydrogenase/reductase (9) and the serine carboxypeptidase-like acyltransferase like gene (10) involved in the last two steps in littorine biosynthesis. This study combines metabolomic studies with a yeast discovery platform (high-throughput screen of candidate genes) to show that pyrrolidine ring formation involves the combined action of a bifunctional spermidine synthase/N-methyltransferase (EcSPDS/SPMT), flavin oxidase (EcAOF1), and copper-dependent oxidase (EcAOC1/2) (Fig. 1) instead of PMT and MPO required in the littorine pathway. The same approach identified a SABATH [SAM-salicylic acid carboxyl-MT (SAMT); benzoic acid-MT (BAMT); THeobromine synthase-MT]-family methyltransferase [4-(1-methyl-2-pyrrolidinyl)-3-oxobutanoate methyltransferase (EcMPOMB)] responsible for the conversion of MPOB to 4-(1-methyl-2-pyrrolidinyl)-3-methyloxobutanoate (MPMOB) and a cytochrome P450 enzyme that converts MPMOB to methylecgonone, that is the substrate for the second to last reaction in cocaine biosynthesis (Fig. 1). Together, a combined metabolomic and yeast discovery platform was essential for the successful identification of five genes (ExSPDS/EcSPMT, EcAOF1, EcAOC1/2, EcMPOBMT, and EcMecgoR) that were functionally characterized and that complete the remaining enzymatic steps of cocaine biosynthesis in the Erythroxylaceae. For example, the yeast platform for the discovery of methylecgonone synthase was composed of a two-plasmid coexpression system for the assembly of the MPMOB precursor and a third system for testing 24 separate candidate genes that led to the discovery of EcCYP81AN15 that converted MPMOB to methylecgonine. The recruitment of nine independent enzyme families compared to those involved in TA biosynthesis within the Solanaceae provide convincing evidence that TA biosynthesis arose independently within the Solanaceae and Erythroxylaceae.

Fig. 1.

The pathway for biosynthesis of cocaine. Putrescine, derived by decarboxylation of ornithine or from agmatine after decarboxylation of arginine, is combined with the aminopropylation cofactor, decarboxylated S-adenosyl-L-methionine (dcSAM) by a spermidine synthase (EcSPDS) that also catalyzes N-methylation of spermidine (EcSPMT) in the presence of the methylation cofactor, S-adenosyl-L-methionine (SAM). The reaction product, N-methylspermidine is cleaved by N-methylspermidine oxidase (EcAOF1) to form N-methylputrescine that is converted to 4-methylaminobutanal by a flavin-dependent amine oxidase (EcAOC1 and EcAOC2) followed by nonenzymic cyclization to N-methylpyrrolinium. Type III polyketide synthases (not shown) catalyze the condensation of two malonyl-CoA esters to yield 3-oxoglutarate, which condenses spontaneously with N-methylpyrrolinium to form MPOB. The SAM-dependent O-methylation of this molecule by MPOB-O-methyltransferase yields 4-(1-methyl-2-pyrrolidinyl)-3-methoxyoxobutanoate (MPMOB). Methylecgonone synthase (CYP81AN15) catalyzes ring closure of MBMOB to generate methylecgonone that is then reduced by methylecgonone reductase (EcMecgoR) to form methylecgonine. The formation of cocaine involves cocaine synthase (EcCS) acylation of methylecgonine with benzoyl CoA.

The completion of the cocaine pathway triggered experiments to see whether hybrid TAs could be produced by coexpressing selected genes from the Solanaceae and Erythroxylaceae in a yeast background. The yeast strain coexpressing type III polyketide synthase together with 4-(1-methyl-2-pyrrolidinyl)-3-oxobutanoate methyltransferase (EcMPO3MT) did accumulate methylecgonone when coexpressing tropinone synthase (AbCYP82M3) from Atropa belladonna. The additional coexpression of E. coca methylecgonone reductase led to the accumulation of methylecgonine. Additional experiments showed promise for production of hybrid TAs in yeast and set the synthetic biology stage for their production as recently shown for yeast strains that produce opioids (1), anticancer drugs (2), and the TAs, hyoscyamine and scopolamine (14).