Parallel Evolution of Insensitivity to Cardenolides in an Herbivore Community: A Molecular Basis

            Similar to foxglove and oleander, plants from the family Apocynaceae such as milkweed produce toxic steroids called cardenolides as a host-plant resistance mechanism to defend against predatory herbivory. Intriguingly, the larvae and adults of certain insect species are able to feed on and obtain nutrition from Apocynaceae without suffering from adverse effects that plague most herbivores. Moreover, some of these insects can sequester cardenolides as a defense against their own predation. For example, the larvae of the monarch butterfly feed on milkweed plants and sequester large amounts of cardenolide which renders them unpalatable to predators. Recently, scientists from Princeton University have disclosed studies on the genetic origins of this phenomenon (For more information about the study and some really great insect photos, see this).

            The Princeton team, led by Peter Andolfatto, surveyed the protein sequence of the alpha subunit of the sodium pump (ATPalpha) in 14 herbivorous insect species that feed on cardenolide-producing plants. Cardenolides bind to ATPalpha and inhibit essential physiological processes that depend on cation transport such as muscle contraction. It was shown by the Princeton team that amino acid substitutions associated with host-plant specialization are highly clustered, with many parallel substitutions. The mutations block ingested cardenolides from binding to ATPalpha.

            To examine the effects of the amino acid substitutions that were observed in the insect lineages on cardenolide binding affinity, molecular docking simulations were studied using an ATPalpha crystal structure bound to ouabain. Ouabain docking onto the wild-type (pig) protein was compared with a protein modified to incorporate the individual amino acid substitutions. A single parallel substitution, N122H, the replacement of an asparagine residue with histidine, occurs in five distinct insect lineages and is known from mutagenesis studies to confer resistance to ouabain in cells. This mutation appears to act by sterically blocking ouabain from entering the cardenolide-binding pocket of ATPalpha (depicted above). Stabilizing hydrogen bonds seem to occur between three polar amino acids in the ATPalpha binding pocket and hydroxyl groups located at positions C14 and C19 on the convex face of ouabain’s steroidal framework. The study provides a fascinating example of parallel evolution by diverse species in the context of a poison protection mechanism and enhances our understanding of the molecular basis for the observed physiological effects associated with cardenolide-type steroids.

Riley's Think-Time

Recent Synthetic Studies Targeting the Steroidal Alkaloid Batrachotoxin

            The steroidal alkaloid batrachotoxin (BTX, 1) functions as a selective agonist of voltage-gated sodium channels and is one of the most potent non-peptidic toxins known. BTX-induced lethality results from a permanent blockade of nerve signal transmission to muscles. The LD₅₀ of 1 (0.1 micrograms per mouse or 2 micrograms kg^-1) is approximately tenfold more potent than that of tetrodotoxin and attests to the severe toxicity associated with the pharmacological properties of this small family of steroidal alkaloids. Batrachotoxins have proven historically valuable as neurochemical agents for the study of voltage-dependent sodium-ion transport in nerve and muscle. Batrachotoxins share certain structural motifs with subcategories of cardiotonic steroid glycosides known as cardenolides and bufadienolides. These architectural commonalities include a steroidal core skeleton with A/B and C/D cis-ring junctions and oxygenated functionality at C3 and C14 in a b orientation. However, certain structural features of the batrachotoxins are unique as compared with other naturally occurring pregnanes, such as the intramolecular C3-hemiketal (BTX’s unique 3b-hydroxy-3a,9a-oxido arrangement is the first reported occurrence in Nature), the seven-membered 14b,18b-heterocyclic oxazapane ring spanning the C/D-ring junction, the 9a-hydroxyl, the D^16,17 unsaturation and the substituted pyrrole ester linked to the C20a-hydroxyl. BTX has received limited attention from the synthetic community (previous efforts reviewed here) and a practical and modular synthetic approach to analogues of 1 is not currently available. Du Bois and co-workers have recently disclosed a novel strategy to access BTX congeners bearing a fully elaborated eastern C/D/E substructural framework.

            Du Bois’ approach, reminiscent of Kishi’s previous total synthesis of BTX, relies on a key furan (as diene) Diels–Alder cycloaddition reaction to fashion a segment of the steroidal architecture. Whereas Kishi’s late-stage intramolecular cycloaddition forged the eastern C/D ring system, Du Bois’ related intermolecular [4+2] reaction, conducted on furanyl intermediates bearing an intact C/D/E substructure, introduces the western A-ring moiety. The furan, in the latter case, serves as the B-ring surrogate and offers flexibility to facilitate A-ring analogue production. Du Bois’ forward synthesis begins with nucleophilic addition of 2-lithio-3-bromofuran, a vicinal dianion equivalent, to the diketo-aldehyde 2 to provide a diastereomeric mixture of hemiacetals. Subsequent exposure of this product distribution to MOM-chloride under basic conditions then affords the cis-fused bicyclic tetrahydrofuranyl-cyclopentanones 3 in a 2:1 diastereomeric ratio favoring the endo isomer (depicted above). The major diastereomer (endo-3) is now nicely predisposed for formation of the C/D ring system of BTX with control of the relative stereochemistry between positions C11 and C14. In the event, intramolecular anion addition to the C14 carbonyl of 3 furnishes 4 in a highly stereocontrolled fashion. Intermediate 4 is then easily elaborated to 5, the substrate for a critical reduction of the C20 enone. Diastereoselective reduction of 5 with DIBAL-H generates a product of 1,2-reduction (6) that is consistent with a chelate addition model. The advanced intermediate 6 is then elaborated to the fully functionalized C/D/E core structure 7 in a straightforward fashion.

            Diels-Alder reactions between the advanced tetracyclic furan derivative 7 and ring-strained dienophiles including a cyclohexyne (derived from 8) and a cyclopropene (derived from 10) produce highly complex A-ring-containing cycloadducts (9 and 11), albeit with a modest degree of selectivity for approach of the dienophile from the desired b-face of the furan. A logical synthetic pathway leading to 1 from the advanced intermediates 9 or 11 is not readily apparent. The authors note that the evaluation of these BTX derivatives as modulators of voltage-gated sodium channels is currently underway. The unassuming architectural complexities embedded within the polycyclic steroidal framework of BTX (1) become more readily apparent when one considers that Du Bois and co-workers have previously synthesized daunting targets including zaragozic acid, saxitoxin and tetrodotoxin, with relative ease. However, as of 2013, a modular and scalable synthesis of 1 still eludes the modern synthetic chemist.