Quinolizidine alkaloids (QAs) are a class of lysine-derived specialized metabolites that are important for the agricultural, pharmaceutical, and chemical sectors. QToxic and bitter QAs accumulate in the seeds of cultivated lupins (Lupinus spp.) and limit the potential of these nutritious crops for the manufacture of food products and feedstock. Furthermore, shortages in supply of commercially important QAs, such as sparteine, have prevented their exploitation for industrial processes. Compared to other alkaloid classes, the biosynthesis of QAs is poorly understood. In fact, only the first two enzymes in the pathway have been characterized. Filling the gaps in the QA pathway, will facilitate the breeding of lupin varieties with low QA levels and the production of commercially important sparteine. In this thesis, I report the results of the research I carried out during my PhD at the University of Copenhagen. The main objective of my work was to investigate the biosynthetic pathway of QAs in the promising crop species narrow-leafed lupin (L. angustifolius, NLL). I present five manuscripts to illustrate the most salient discoveries I made during my PhD. First, I present a critical review of the different QA pathway hypotheses that have been proposed in the literature, which helped me lay the groundwork for subsequent gene discovery. I then present the development of a virus- induced gene silencing protocol for the rapid and efficient functional characterization of genes in lupins. In the last three manuscripts, I explore the biochemical basis for the biosynthesis of (+)-lupanine from L-lysine. First, I describe the discovery of an entirely unexpected early step of the QA pathway catalyzed by an acetyltransferase (AT), and demonstrate that a mutation in AT is crucial for enabling breeding of low-QA varieties. In the next manuscript, I show how I used a combination of transcriptome mining and co- expression analysis in NLL to discover the remaining QA biosynthetic genes up to (-)- sparteine and how I reconstructed the entire biosynthetic pathway in Nicotiana benthamiana. In the last manuscript, I provide an alternative and – in my view – more economically viable route to producing (-)-sparteine by engineering the QA pathway in NLL. My work represents an extensive contribution to our understanding of QA biosynthesis and paves the way for the large-scale plant-based production of (-)-sparteine. The knowledge and methods presented here can contribute to the creation of novel, low- QA lupins as well as kick-start the accelerated domestication of related legumes.