A virus-induced gene silencing approach to understanding alkaloid metabolism in Catharanthus roseus

Abstract

The anticancer agents vinblastine and vincristine are bisindole alkaloids derived from coupling vindoline and catharanthine, monoterpenoid indole alkaloids produced exclusively by the Madagascar periwinkle (Catharanthus roseus). Industrial production of vinblastine and vincristine currently relies on isolation from C. roseus leaves, a process that affords these compounds in 0.0003-0.01% yields. Metabolic engineering efforts to either improve alkaloid content or provide alternative sources of the bisindole alkaloids ultimately rely on the isolation and characterization of the genes involved. Several vindoline biosynthetic genes have been isolated, and the cellular and subcellular organization of the corresponding enzymes has been well studied. However, due to the leaf-specific localization of vindoline biosynthesis, and the lack of production of this precursor in cell suspension and hairy root cultures of C. roseus, further elucidation of this pathway demands the development of reverse genetics approaches to assay gene function in planta. The bipartite pTRV vector system is a Tobacco Rattle Virus-based virus-induced gene silencing (VIGS) platform that has provided efficient and effective means to assay gene function in diverse plant systems. A VIGS method was developed herein to investigate gene function in C. roseus plants using the pTRV vector system. The utility of this approach in understanding gene function in C. roseus leaves is demonstrated by silencing known vindoline biosynthetic genes previously characterized in vitro.

Figure 1

Proposed biosynthesis of vindoline from tabersonine in C. roseus, and incorporation of vindoline into the bisindole alkaloids, vinblastine and vincristine. Enzyme names are indicated in bold above solid black arrows. Cognate cDNAs have been isolated for all steps in the conversion of tabersonine to vindoline, except for the so-called hydration step (dotted arrow). Grey arrows indicate multiple enzymatic steps. Abbreviations: T16H, tabersonine 16-hydroxylase; 16OMT, 16-hydroxytabersonine 16-O-methyltransferase; NMT, 16-methoxy-2,3-dihydro-3-hydroxytabersonine N-methyltransferase; D4H, desacetoxyvindoline 4-hydroxylase; DAT, deacetylvindoline 4-O-acetyltransferase; PRX1, anhydrovinblastine synthase.

Figure 2

Silencing of protoporphyrin IX magnesium chelatase gene (ChlH) as a visual marker for virus-induced gene silencing. (a) Representative C. roseus plant 2 weeks after infiltration with pTRV1/pTRV2-ChlH(F3R1) (1:1; ChlH-vigs) exhibiting photobleached phenotype. (b) Representative C. roseus plant 2 weeks after infiltration with pTRV1/empty pTRV2 (1:1; EV). (c) Fold change in ChlH and vindoline (VIND) biosynthetic gene transcripts levels in leaves of EV and ChlH-vigs plants relative to mock-infiltrated (M) plants, as determined by quantitative real-time PCR. Data represent mean ± SE of at least three technical replicates performed with cDNA prepared from 8 (EV) and 3 (ChlH-vigs) individual plants. Asterisks indicate significant differences between EV and mock, or EV and ChlH-vigs plants using Student’s t-test (*, P < 0.05; **, P < 0.0001). (d) Alkaloid composition and total alkaloids in leaves of M, EV, and ChlH-vigs plants. Data represent mean ± SE alkaloid composition (as % of total) or total alkaloid content (normalized to M) from 5 (M), 8 (EV) and 3 (ChlH-vigs) individual plants. Asterisks indicate significant differences between EV and M, or ChlH-vigs and M plants using Student’s t-test (*, P < 0.05). Total leaf alkaloid content of from ChlH-vigs plants compared to EV plants was not significantly different (P > 0.05). Abbreviated alkaloid and biosynthetic gene (enzyme) names are defined in Figure 1.

Figure 3

Virus-induced silencing of 16-methoxy-2,3-dihydro-3-hydroxytabersonine N-methyltransferase (NMT) transcripts in C. roseus. (a) Selected ion chromatograms (m/z 385+457) showing relative vindoline (VIND, m/z 457) and 16-methoxy-2,3-dihydro-3-hydroxytabersonine (MHDHT, m/z 385) in mock infected (M), empty vector infected (EV), and least (NMT-vigs-1) and most (NMT-vigs-1a) extreme phenotypes observed for pTRV2-NMT infected plants. (b) Relative expression of NMT in EV and NMT-vigs plants compared to M plants, measured by quantitative real-time PCR. Data represent mean ± SE of at least three technical replicates performed with cDNA prepared from 8 (EV) and 7 (NMT-vigs) individual plants. Asterisk indicates significant difference between NMT-vigs and EV plants using Student’s t-test (P < 0.05). (c) Mean alkaloid composition and total alkaloid content (relative to M) in EV and NMT-vigs plants. Data represent mean ± SE alkaloid composition (as % of total) or total alkaloid content (normalized to M) from 8 (EV) and 7 (NMT-vigs) individual plants. (d) Selected ion chromatogram (m/z 327+385+399) of in vitro rescue assay demonstrating the S-adenosyl-L-methionine (SAM)-dependent conversion of MHDHT (m/z 385) accumulating in NMT-vigs plants to desacetoxyvindoline (DESV, m/z 399) by recombinant NMT. Abbreviated alkaloid names are defined in Fig. 1; I.S., ajmaline internal standard (m/z 327).

Figure 4

Virus-induced silencing of desacetoxyvindoline 4-hydroxylase (D4H) transcripts in C. roseus. (a) Selected ion chromatograms (m/z 399 and m/z 457) showing relative vindoline (VIND, m/z 457, grey trace) and desacetoxyvindoline (DESV m/z 399, black trace) in mock infected (M), empty vector infected (EV), and least (D4H-vigs-4b) and most (D4H-vigs-6b) extreme phenotypes observed for pTRV2-D4H infected plants. (b) Relative expression of D4H in EV and D4H-vigs plants compared to M plants, measured by quantitative real-time PCR. Data represent mean ± SE of at least three technical replicates performed with cDNA prepared from 8 (EV) and 7 (D4H-vigs) individual plants. Asterisks indicate significant difference between D4H-vigs and EV plants using Student’s t-test (P < 0.0001). (c) Mean alkaloid composition and total alkaloid content (relative to M) in EV and D4H-vigs plants. Data represent mean ± SE alkaloid composition (as % of total) or total alkaloid content (normalized to M) from 8 (EV) and 7 (D4H-vigs) individual plants. (d) Selected ion chromatogram (m/z 327+399+415) of in vitro D4H rescue assay demonstrating the α-ketoglutarate (α-KG)-dependent conversion of DESV (m/z 399) accumulating in D4H-vigs plants to deacetylvindoline (DAV, m/z 415) using crude, soluble protein extract from young leaves of wild type C. roseus plants. Abbreviated alkaloid names are defined in Fig. 1; I.S., ajmaline internal standard (m/z 327).

Figure 5

Virus-induced silencing of deacetylvindoline 4-O-acetyltransferase (DAT) transcripts in C. roseus. (a) Selected ion chromatograms (m/z 415+457) showing relative vindoline (VIND, m/z 457) and deacetylvindoline (DAV, m/z 415) in mock infected (M), empty vector infected (EV), and least (DAT-vigs-1b) and most (DAT-vigs-6b) extreme phenotypes observed for DAT-silenced plants. (b) Relative expression of DAT in EV and DAT-vigs plants compared to M plants, measured by quantitative real-time PCR. Data represent mean ± SE of at least three technical replicates performed with cDNA prepared from 8 (EV) and 7 (DAT-vigs) individual plants. Asterisks indicate significant difference between DAT-vigs and EV plants using Student’s t-test (P < 0.001). (c) Mean alkaloid composition and total alkaloid content (relative to M) in empty vector and DAT-vigs plants. Data represent mean ± SE alkaloid composition (as % of total) or total alkaloid content (normalized to M) from 8 (EV) and 7 (DAT-vigs) individual plants. (d) Selected ion chromatogram (m/z 327+415+457) of in vitro DAT rescue assays demonstrating the acetyl-CoA-dependent conversion of DAV (m/z 415) accumulating in DAT-vigs plants to VIND (m/z 415) using crude, soluble protein extract from methyl jasmonate-elicited C. roseus seedlings. Abbreviated alkaloid names are defined in Fig. 1; I.S., ajmaline internal standard (m/z 327).