Backgroundβ-Ionone is a fragrant terpenoid that generates a pleasant floral scent and is used in diverse applications as a cosmetic and flavoring ingredient. A growing consumer desire for natural products has increased the market demand for natural β-ionone. To date, chemical extraction from plants remains the main approach for commercial natural β-ionone production. Unfortunately, changing climate and geopolitical issues can cause instability in the β-ionone supply chain. Microbial fermentation using generally recognized as safe (GRAS) yeast offers an alternative method for producing natural β-ionone. Yarrowia lipolytica is an attractive host due to its oleaginous nature, established genetic tools, and large intercellular pool size of acetyl-CoA (the terpenoid backbone precursor).ResultsA push–pull strategy via genome engineering was applied to a Y. lipolytica PO1f derived strain. Heterologous and native genes in the mevalonate pathway were overexpressed to push production to the terpenoid backbone geranylgeranyl pyrophosphate, while the carB and biofunction carRP genes from Mucor circinelloides were introduced to pull flux towards β-carotene (i.e., ionone precursor). Medium tests combined with machine learning based data analysis and 13C metabolite labeling investigated influential nutrients for the β-carotene strain that achieved > 2.5 g/L β-carotene in a rich medium. Further introduction of the carotenoid cleavage dioxygenase 1 (CCD1) from Osmanthus fragrans resulted in the β-ionone production. Utilization of in situ dodecane trapping avoided ionone loss from vaporization (with recovery efficiencies of ~ 76%) during fermentation operations, which resulted in titers of 68 mg/L β-ionone in shaking flasks and 380 mg/L in a 2 L fermenter. Both β-carotene medium tests and β-ionone fermentation outcomes indicated the last enzymatic step CCD1 (rather than acetyl-CoA supply) as the key bottleneck.ConclusionsWe engineered a GRAS Y. lipolytica platform for sustainable and economical production of the natural aroma β-ionone. Although β-carotene could be produced at high titers by Y. lipolytica, the synthesis of β-ionone was relatively poor, possibly due to low CCD1 activity and non-specific CCD1 cleavage of β-carotene. In addition, both β-carotene and β-ionone strains showed decreased performances after successive sub-cultures. For industrial application, β-ionone fermentation efforts should focus on both CCD enzyme engineering and strain stability improvement.Electronic supplementary materialThe online version of this article (10.1186/s12934-018-0984-x) contains supplementary material, which is available to authorized users.