Current and future experimental strategies for structural analysis of trichothecene mycotoxins--a prospectus

Abstract

Fungal toxins, such as those produced by members of the order Hypocreales, have widespread effects on cereal crops, resulting in yield losses and the potential for severe disease and mortality in humans and livestock. Among the most toxic are the trichothecenes. Trichothecenes have various detrimental effects on eukaryotic cells including an interference with protein production and the disruption of nucleic acid synthesis. However, these toxins can have a wide range of toxicity depending on the system. Major differences in the phytotoxicity and cytotoxicity of these mycotoxins are observed for individual members of the class, and variations in toxicity are observed among different species for each individual compound. Furthermore, while diverse toxicological effects are observed throughout the whole cellular system upon trichothecene exposure, the mechanism of toxicity is not well understood. In order to comprehend how these toxins interact with the cell, we must first have an advanced understanding of their structure and dynamics. The structural analysis of trichothecenes was a subject of major interest in the 1980s, and primarily focused on crystallographic and solution-state Nuclear Magnetic Resonance (NMR) spectroscopic studies. Recent advances in structural determination through solution- and solid-state NMR, as well as computation based molecular modeling is leading to a resurgent interest in the structure of these and other mycotoxins, with the focus shifting in the direction of structural dynamics. The purpose of this work is to first provide a brief overview of the structural data available on trichothecenes and a characterization of the methods commonly employed to obtain such information. A summary of the current understanding of the relationship between structure and known function of these compounds is also presented. Finally, a prospectus on the application of new emerging structural methods on these and other related systems is discussed.

Keywords: trichothecene; NMR; NMR crystallography; antibiotic; crystallography; molecular dynamics; molecular modeling; mycotoxin; natural product; ribosome; solid-state NMR; structure; structure-function.

Figure 1

Chemical structure of the trichothecene core. Substituents R1 through R5 are depicted with their stereochemical configuration off the core.

Figure 2

Three-dimensional stereochemistry of the trichothecene core when (A) the A-ring is in a half-chair, and the B-ring in a chair conformation; and (B) the A-ring is a half-chair, and the B-ring in a boat conformation.

Figure 3

The general core structures for Type A, B, C, and D trichothecenes.

Figure 4

Structures of some non-trichothecene trichodienoid compounds: (A) sambucinol; (B) sambucoin; (C) 13-hydroxy-3α,11-epoxyapotrichothecene.

Figure 5

The ¹H NMR spectra of (A) T-2 toxin and (B) Deoxynivalenol (DON) at 300 MHz in CDCl₃. All ¹H resonances have been assigned and are labeled.

Figure 6

Two-dimensional NMR spectra of T-2 toxin. (A) Correlation Spectroscopy (COSY) spectrum showing ¹H-¹H homonuclear through-bond (J-coupling) correlations; (B) Nuclear Overhauser Effect Spectroscopy (NOESY) spectrum depicting ¹H-¹H homonuclear through-space (dipolar coupling) correlations; (C) Heteronuclear Single Quantum Correlation (HSQC) spectroscopy spectrum depicting ¹H-¹³C correlations for carbon nuclei which are directly bound to one or more ¹H nuclei; (D) Heteronuclear Multiple Bond Correlation (HMBC) spectroscopy spectrum depicting ¹H-¹³C correlations for carbon nuclei which are up to 4 bonds away from one or more ¹H nuclei.

Figure 7

Structural changes observed for NIV and DON under different solvent conditions: (A) chemical structure of NIV; (B) chemical structure of DON; (C) hemiketal formation from C-8 to C-15 observed for NIV and DON, where R is -OH and -H for NIV and DON, respectively; (D) cyclic ether formation from C-4 to C-15 observed for NIV only.

Figure 8

Deuterium exchange experiment performed on T-2 toxin. Regions exhibiting significant changes throughout the incremental addition of D₂O have been expanded to show peak structure. Of particular interest are the H_3OH and H₂O/HDO regions, which not only demonstrate significant changes in chemical shift, but also exhibit the retention of their sharp peak structure, indicating a slow chemical exchange process. H₃ is less affected in that the only observable changes are the loss of coupling to H_3OH as the latter peak is converted to H_3OD.

Figure 9

Appearance of NMR signals under (A) isotropic, and (B) anistropic experimental conditions. Samples rotationally averaged by the Brownian motion of a solvent in the solution-state; or solid-state samples which are rotationally averaged due to Magic-Angle Spinning (MAS), will have NMR lineshapes similar to that depicted in (A). Solid-state samples experiencing multiple orientations simultaneously will display powder patterns similar to that depicted in (B).

Figure 10

Solution- and solid-state NMR structures observed for T-2 toxin: (A) solution-state conformation of T-2 toxin observed in CDCl₃; (B and C) two different solid-state conformations for T-2 toxin, differing mainly in the orientation of the side chains.

Figure 11

Proposed water binding interaction for T-2 toxin in the (A) solution- and (B) solid-state. At least one molecule of water has been observed to bind in the pocket formed between the tetrahydropyranyl B-ring and cyclopentyl C-ring. The major chemical shift differences observed for T-2 toxin in the solid-state NMR spectrum suggest that a second water binding site may be present near C-12 of the epoxide ring.

Figure 12

Superposition of the solid-state (black) and solution-state (blue) spectra displaying the twinning observed solid-state signals, indicating that two distinct conformations for T-2 toxin are present in the unit cell. (A) methyl carbons; (B,C) methylene carbons; (D-G) methine, π-bonded, and quaternary carbons; (H) carbonyl carbons, respectively.