Identification of a Noroxomaritidine Reductase with Amaryllidaceae Alkaloid Biosynthesis Related Activities

Abstract

Amaryllidaceae alkaloids are a large group of plant natural products with over 300 documented structures and diverse biological activities. Several groups of Amaryllidaceae alkaloids including the hemanthamine- and crinine-type alkaloids show promise as anticancer agents. Two reduction reactions are required for the production of these compounds: the reduction of norcraugsodine to norbelladine and the reduction of noroxomaritidine to normaritidine, with the enantiomer of noroxomaritidine dictating whether the derivatives will be the crinine-type or hemanthamine-type. It is also possible for the carbon-carbon double bond of noroxomaritidine to be reduced, forming the precursor for maritinamine or elwesine depending on the enantiomer reduced to an oxomaritinamine product. In this study, a short chain alcohol dehydrogenase/reductase that co-expresses with the previously discovered norbelladine 4'-O-methyltransferase from Narcissus sp. and Galanthus spp. was cloned and expressed in Escherichia coli Biochemical analyses and x-ray crystallography indicates that this protein functions as a noroxomaritidine reductase that forms oxomaritinamine from noroxomaritidine through a carbon-carbon double bond reduction. The enzyme also reduces norcraugsodine to norbelladine with a 400-fold lower specific activity. These studies identify a missing step in the biosynthesis of this pharmacologically important class of plant natural products.

Keywords: Amaryllidaceae alkaloids; crystal structure; enzyme structure; norbelladine; noroxomaritidine reductase; reductase; secondary metabolism; transcriptomics.

FIGURE 1.

Amaryllidaceae alkaloid biosynthesis pathway. Enzymes involved in the pathway are indicated in blue. PAL, phenylalanine ammonia lyase, N4OMT, and NR are indicated.

FIGURE 2.

Phylogenetic relationship of NR to other plant reductases and dehydrogenases. Amino acid sequences of all of the enzymes were taken from NCBI (PDB 2AE2, Q071NO.1, AJO70763.1, AKF02528.1, AAC48976.1, ABW24501.1, and AAS76634.1). The phylogenetic tree was constructed using MEGA (43), and the amino acid sequences were aligned using MUSCLE. The LG model was used to construct a Maximum Likelihood tree using a bootstrap replicate of 500. The percent identity of the plant reductases and dehydrogenases used in the tree were calculated using the Needleman-Wunsch Global Alignment tool from NCBI with the gap costs set at an existence of 11 and an extension of 2.

FIGURE 3.

Noroxomaritidine reduction by NR. Enzymatic activity of NR for reduction of noroxomaritidine (m/z 274.3) was monitored using LC-MS/MS with the same LC time program found in Ref. . A, enzyme assays of NR with 100 mm sodium phosphate buffer, pH 7.0, 100 μm noroxomaritidine, 1 mm NADPH, and 10 μg of pure protein in 100 μl at 30 °C for 2 h. Traces from top to bottom are: noroxomaritidine standard; complete assay with NR, noroxomaritidine, and NADPH; assay without NADPH; assay without noroxomaritidine; and assay with TALON resin purified E. coli empty vector protein extract substituted for NR protein. B, EPI MS/MS of NR product. C, EPI MS/MS of noroxomaritidine reduced with NaBH₄.

FIGURE 4.

NR consumption of 20 μm (10bS,4aR or 10bR,4aS)-noroxomaritidine using the same assay conditions as for specific activity assays incubated 2 h. A, LC-MS using a Chrom Tech, Inc. Chiral-CBH 100 × 4.0-mm, 5-μm column and the same LC setup and time program as for chiral separations in Ref. . Samples were monitored at m/z 272.3 with CE (25) and DP (70) on the QTRAP 6500 for the following samples top to bottom: (10bS,4aR and 10bR,4aS)-noroxomaritidine mixed standard; complete assay with NR, (10bS,4aR and 10bR,4aS)-noroxomaritidine and NADPH; assay without NADPH; assay without (10bS,4aR and 10bR,4aS)-noroxomaritidine substrate; assay without NR enzyme; and assay with TALON resin purified E. coli empty vector protein extract substituted for NR protein. B, the two enantiomers of noroxomaritidine and corresponding expected products.

FIGURE 5.

Norcraugsodine reduction to norbelladine by NR. Enzymatic activity of NR for reduction of norcraugsodine was monitored using the same assay conditions and LC-MS/MS setup as in specific activity assays. A, MRM (m/z 260.1/238.0) of NR assays. Traces from top to bottom are: norbelladine standard; complete assay with NR, tyramine, 3,4-dihydroxybenzaldehyde, and NADPH; assay without NADPH; assay without tyramine and 3,4-dihydroxybenzaldehyde; assay without enzyme; and assay with TALON resin purified E. coli empty vector protein. B, MRM (m/z 260.1/238.0) relative quantification of assays shown in panel A in triplicate. C, EPI MS/MS of norbelladine standard. D, EPI MS/MS of NR product.

FIGURE 6.

A comparison of the tyramine (CE (15), DP (60), m/z 138.1/121.0 and m/z 138.1/93.0), levels in daffodil organs (above ground leaf (agl), below ground leaf (bgl), leaf scale (ls), bulb scale (bs), bulb core (bc), root (r), flower stalk above ground (fsag), flower stalk below ground (fsbg), and flower (f)). These samples were collected during the months of January (well developed roots with well developed flower primordia), March (all plants form this month lack flower primordia due to sampling limitations), April (plants all flowering), and August (dormant bulbs). The instrument used for the experiment was a QTRAP 6500 with a Phenomenex Luna 5-μm C8(2) 250 × 4.60-mm LC column using the following LC method; A = 0.1% formic acid and B = acetonitrile, start with 10% B for 2 min followed by a linear increase to 40% B for 9.9 min, linear increase to 90% B for 0.1 min, hold for 3 min, linear decrease to 10% B for 0.1 min, and hold 7 min. The following were also monitored by MRM across these conditions but were not detected 3,4-dihydroxybenzaldehyde (CE (15), DP (60), m/z 139.0/111.0 and 139.0/93.0), norbelladine (CE (15), DP (50), m/z 260.1/138.0 and 260.1/121.0), noroxomaritidine (CE (25), DP (70), m/z 272.3/229.0 and 272.1/212.1), and oxomaritinamine (CE (20), DP (60), m/z 274.1/219.1 and 274.1/136.1). Alkaloid samples were prepared by utilizing liquid nitrogen, a mortar and pestle for grinding of frozen tissue, followed by 70% ethanol extraction, PTFE membrane filtration, extract drying, and extract resuspension in mobile phase.

FIGURE 7.

Three-dimensional structure of NR. A, ribbon diagram of the NR tetramer complexed with NADP⁺ and tyramine is shown. Each monomer is colored differently. NADP⁺ and tyramine are shown as space-filling models. B, NADP⁺ binding in NR. Residues interacting with NADP⁺ are shown as stick figures. Dotted lines represent hydrogen bond and charge-charge interactions.

FIGURE 8.

The ligand binding site of NR and loop movement. A, details of the tyramine binding site. B, surface view of the NR ligand binding site showing tyramine and NADP⁺. C, details of the piperonal binding site. D, comparison of the 209–220 loop movement in the NR·NADP⁺ (cream) and NR·NADP⁺·tyramine complexes. The positions of Thr²¹³ and Phe²¹⁶ are shown in each structure for reference. The positions of the hinge residues Arg²⁰⁹ and Lys²²⁰ are noted by triangles.

FIGURE 9.

Structure-based model for noroxomaritidine binding by NR. A, surface representation of noroxomaritidine docked into the open active site conformation of NR. B, residues in the proposed noroxomaritidine binding site.

FIGURE 10.

Proposed reaction mechanisms for NR catalyzed reduction of noroxomaritidine (A) and norcraugsodine (B).