Enzymes are powerful biological catalysts used in the manufacturing of diverse chemicals and pharmaceuticals due to their high efficiency and selectivity. Traditionally, enzymes are conventionally recognized for their remarkable catalytic specificity. Up to now, numerous enzymes have been found to exhibit catalytic promiscuity (Hult and Berglund, 2007; Tawfik, and S, D., 2010). Catalytic promiscuity is a subset of enzyme promiscuity, which refers to the ability of enzymes to catalyze secondary or non-native reactions beyond their primary biological functions. And enzyme promiscuity is generally categorized into three types: catalytic promiscuity, substrate promiscuity, and conditional promiscuity (Hult and Berglund, 2007) (Fig. 1).

Substrate promiscuity means that enzymes have a less strict substrate selectivity(Thakur and Pandit, 2022). They can catalyze reactions with non-natural substrates via the same catalytic mechanism as for its native substrate. For example, methane monooxygenase can hydroxylate 150 substrates besides methane(Elliott et al., 1997). The flexibility is largely attributed to the structure of catalytic active site, which allows small molecule substrates to bind. Substrate-promiscuous enzymes possess more spacious and adaptable active pockets, enabling interactions with multiple substrates. In contrast, substrate-specific enzymes possess highly selective active sites that typically accommodate only one specific substrate-analogous to a key fitting precisely into a corresponding lock. Enzymes with conditional promiscuity can catalyze reactions in different conditions, such as anhydrous media, extreme temperatures, or pH values. For instance, lipase is able to catalyze ester substrates in both aqueous solutions and organic solvents(Godoy et al., 2022). Enzyme activity in organic solvents offers various advantages, including enhanced substrate solubility, reversal of hydrolysis reactions, and even altered enzyme specificity, which may unlock novel enzymatic functions (Kumar et al., 2016). While the applied potential of conditional and substrate promiscuity has been recognized for decades and resulted in many applications, catalytic promiscuity has only recently emerged as a significant topic in biocatalysis and synthetic biology.

Catalytic promiscuity, also known as cross-reactivity or poly-reactivity, refers to enzymes that have additional activities besides their “native” activity and can catalyze multiple chemically distinct reactions (Bornscheuer and Kazlauskas, 2004). In recent years, both theoretical and experimental studies have confirmed that enzyme catalytic promiscuity is a widespread phenomenon in nature and plays a critical role in enzyme evolution (Duarte et al., 2013; Galmes et al., 2021; Gupta, 2016). The catalytic promiscuity of existing enzymes in nature is a crucial starting point for the evolution of new activities, and it holds immense significance in enriching the diversity of natural compounds. We summarize notable examples of catalytically promiscuous enzymes and their reactions documented in the literature, with a particular focus on applications in synthetic chemistry (Table 1).

According to the Yčas-Jensen theory, ancestral enzymes with dual catalytic functions existed at key evolutionary nodes (Jensen, 1976; Yčas, 1974). It is believed that catalytic promiscuity served as a driving force that allowed enzymes to evolve from multifunctional ancestors into specialized catalysts, while some specific enzymes retain latent promiscuous activities that can be reactivated under selective pressure (Glasner et al., 2020; Yang et al., 2020). These residual activities offer a foundation for evolving new enzymatic functions, providing a valuable strategy to expand the range of enzyme applications in the industrial biocatalysts. (Chen and Arnold, 2020; Leveson-Gower et al., 2019).

Although the alternative activity of a catalytically promiscuous enzyme is usually significantly lower than its native reaction, various technical strategies, such as directed evolution, rational design and computational enzyme design, have been employed to enhance catalytic properties, rendering these reactions practically viable (Bell et al., 2024; Yang et al., 2024). In recent years, the catalytic promiscuity of enzymes has attracted extensive attention, opening up new avenues in biocatalyst research. Several successful studies(Chen and Arnold, 2020; Leveson-Gower et al., 2019) in protein engineering highlight the crucial role of catalytic promiscuity in expanding new enzyme functions. This approach may lead to the development of efficient and stable catalysts exhibiting unprecedented activity for reactions that currently lack viable enzyme alternatives. For instance, the cytochrome P450 (referred to as CYP) enzymes have been successfully engineered to catalyze novel reactions not found in nature(Chen and Arnold, 2020; Yang and Arnold, 2021).

By tailoring the promiscuity of natural enzymes, it is now possible to design biocatalysts with novel or enhanced functions, which expands the functional versatility of single enzymes and unlocks new biotechnological applications(Copley et al., 2023; Jiang et al., 2017). Protein engineering technologies can be employed to design and modify the catalytic promiscuity of natural enzymes, enabling them to perform new or improved functions. Despite these advances, guiding promiscuous enzymes toward desired reactions remains challenging due to incomplete understanding of enzyme catalytic mechanisms and key influencing factors. Directed evolution or semi-rational remain the predominant tools, but they are often labor-intensive and time-consuming. With the development of computational methods, particularly artificial intelligence-aided enzyme design, has introduced powerful tools for optimizing enzyme performance and even uncovering previously unknown catalytic functions. Recently, machine learning methods can predict the functions of unannotated proteins, resulting in a variety of candidate enzymes for applications(Yu et al., 2023). More recently, increased efforts have focused on generating novel enzymes with low sequence similarity to known proteins, capable of catalyzing unprecedented reactions.

In this review, we provide a comprehensive overview of the factors that can potentially impact the promiscuity of enzyme catalysis, including differences in substrate binding modes in pre-reaction states, instability of key high-energy intermediates in catalytic mechanisms, and environmental perturbations that drive enzyme evolution. Additionally, we highlight current strategies in protein engineering and artificial intelligence-assisted design for enhancing enzyme activity, selectivity, and functional diversity. Our goal is to offer insights that contribute to the development of next-generation biocatalysts capable of performing new-to-nature chemistry with high efficiency and industrial properties.