SUMMARYCyanogenic glucosides are amino acid-derived defence compounds found in a large number of vascular plants. Their hydrolysis by specific b-glucosidases following tissue damage results in the release of hydrogen cyanide. The cyanogenesis deficient1 (cyd1) mutant of Lotus japonicus carries a partial deletion of the CYP79D3 gene, which encodes a cytochrome P450 enzyme that is responsible for the first step in cyanogenic glucoside biosynthesis. The genomic region surrounding CYP79D3 contains genes encoding the CYP736A2 protein and the UDP-glycosyltransferase UGT85K3. In combination with CYP79D3, these genes encode the enzymes that constitute the entire pathway for cyanogenic glucoside biosynthesis. The biosynthetic genes for cyanogenic glucoside biosynthesis are also co-localized in cassava (Manihot esculenta) and sorghum (Sorghum bicolor), but the three gene clusters show no other similarities. Although the individual enzymes encoded by the biosynthetic genes in these three plant species are related, they are not necessarily orthologous. The independent evolution of cyanogenic glucoside biosynthesis in several higher plant lineages by the repeated recruitment of members from similar gene families, such as the CYP79s, is a likely scenario.

Cyanogenesis, the release of hydrogen cyanide from damaged plant tissues, involves the enzymatic degradation of amino acid-derived cyanogenic glucosides (a-hydroxynitrile glucosides) by specific b-glucosidases. Release of cyanide functions as a defense mechanism against generalist herbivores. We developed a high-throughput screening method and used it to identify cyanogenesis deficient (cyd) mutants in the model legume Lotus japonicus. Mutants in both biosynthesis and catabolism of cyanogenic glucosides were isolated and classified following metabolic profiling of cyanogenic glucoside content. L. japonicus produces two cyanogenic glucosides: linamarin (derived from Val) and lotaustralin (derived from Ile). Their biosynthesis may involve the same set of enzymes for both amino acid precursors. However, in one class of mutants, accumulation of lotaustralin and linamarin was uncoupled. Catabolic mutants could be placed in two complementation groups, one of which, cyd2, encoded the b-glucosidase BGD2. Despite the identification of nine independent cyd2 alleles, no mutants involving the gene encoding a closely related b-glucosidase, BGD4, were identified. This indicated that BGD4 plays no role in cyanogenesis in L. japonicus in vivo. Biochemical analysis confirmed that BGD4 cannot hydrolyze linamarin or lotaustralin and in L. japonicus is specific for breakdown of related hydroxynitrile glucosides, such as rhodiocyanoside A. By contrast, BGD2 can hydrolyze both cyanogenic glucosides and rhodiocyanosides. Our genetic analysis demonstrated specificity in the catabolic pathways for hydroxynitrile glucosides and implied specificity in their biosynthetic pathways as well. In addition, it has provided important tools for elucidating and potentially modifying cyanogenesis pathways in plants.

Flowers and leaves of Lotus japonicus contain α-, β-, and γ-hydroxynitrile glucoside (HNG) defense compounds, which are bioactivated by β-glucosidase enzymes (BGDs). The α-HNGs are referred to as cyanogenic glucosides because their hydrolysis upon tissue disruption leads to release of toxic hydrogen cyanide gas, which can deter herbivore feeding. BGD2 and BGD4 are HNG metabolizing BGD enzymes expressed in leaves. Only BGD2 is able to hydrolyse the α-HNGs. Loss of function mutants of BGD2 are acyanogenic in leaves but fully retain cyanogenesis in flowers pointing to the existence of an alternative cyanogenic BGD in flowers. This enzyme, named BGD3, is identified and characterized in this study. Whereas all floral tissues contain α-HNGs, only those tissues in which BGD3 is expressed, the keel and the enclosed reproductive organs, are cyanogenic. Biochemical analysis, active site architecture molecular modelling, and the observation that L. japonicus accessions lacking cyanogenic flowers contain a non-functional BGD3 gene, all support the key role of BGD3 in floral cyanogenesis. The nectar of L. japonicus flowers was also found to contain HNGs and additionally their diglycosides. The observed specialisation in HNG based defence in L. japonicus flowers is discussed in the context of balancing the attraction of pollinators with the protection of reproductive structures against herbivores.

The evolutionary appearance of non-cyanogenic hydroxynitrile glucosides in the Lotus genus is accompanied by the substrate specialization of paralogous b-glucosidases resulting from a crucial amino acid substitution SUMMARYLotus japonicus, like several other legumes, biosynthesizes the cyanogenic a-hydroxynitrile glucosides lotaustralin and linamarin. Upon tissue disruption these compounds are hydrolysed by a specific b-glucosidase, resulting in the release of hydrogen cyanide. Lotus japonicus also produces the non-cyanogenic c-and b-hydroxynitrile glucosides rhodiocyanoside A and D using a biosynthetic pathway that branches off from lotaustralin biosynthesis. We previously established that BGD2 is the only b-glucosidase responsible for cyanogenesis in leaves. Here we show that the paralogous BGD4 has the dominant physiological role in rhodiocyanoside degradation. Structural modelling, site-directed mutagenesis and activity assays establish that a glycine residue (G211) in the aglycone binding site of BGD2 is essential for its ability to hydrolyse the endogenous cyanogenic glucosides. The corresponding valine (V211) in BGD4 narrows the active site pocket, resulting in the exclusion of non-flat substrates such as lotaustralin and linamarin, but not of the more planar rhodiocyanosides. Rhodiocyanosides and the BGD4 gene only occur in L. japonicus and a few closely related species associated with the Lotus corniculatus clade within the Lotus genus. This suggests the evolutionary scenario that substrate specialization for rhodiocyanosides evolved from a promiscuous activity of a progenitor cyanogenic b-glucosidase, resembling BGD2, and required no more than a single amino acid substitution.

Lima bean, Phaseolus lunatus, is a crop legume that produces the cyanogenic glucosides linamarin and lotaustralin. In the legumes Lotus japonicus and Trifolium repens, the biosynthesis of these two α‐hydroxynitrile glucosides involves cytochrome P450 enzymes of the CYP79 and CYP736 families and a UDP‐glucosyltransferase. Here, we identify CYP79D71 as the first enzyme of the pathway in P. lunatus, producing oximes from valine and isoleucine. A second CYP79 family member, CYP79D72, was shown to catalyze the formation of leucine‐derived oximes, which act as volatile defense compounds in Phaseolus spp. The organization of the biosynthetic genes for cyanogenic glucosides in a gene cluster aided their identification in L. japonicus. In the available genome sequence of P. vulgaris, the gene orthologous to CYP79D71 is adjacent to a member of the CYP83 family. Although P. vulgaris is not cyanogenic, it does produce oximes as volatile defense compounds. We cloned the genes encoding two CYP83s (CYP83E46 and CYP83E47) and a UDP‐glucosyltransferase (UGT85K31) from P. lunatus, and these genes combined form a complete biosynthetic pathway for linamarin and lotaustralin in Lima bean. Within the genus Phaseolus, the occurrence of linamarin and lotaustralin as functional chemical defense compounds appears restricted to species belonging to the closely related Polystachios and Lunatus groups. A preexisting ability to produce volatile oximes and nitriles likely facilitated evolution of cyanogenesis within the Phaseolus genus.