**1. Introduction**

Bee pollen as a natural food source is regarded as an excellent nutritional supplement for human consumption. It contains a variety of nutrients including proteins, carbohydrates, lipids, polyphenols, and many other nutrients [1]. Bee pollen also displays a variety of beneficial health properties, such as antioxidant, antibacterial, hepatoprotective, and cardioprotective activities [2,3]. By 2024, the global bee pollen market is expected to reach a value of USD 720 million according to Marketwatch.com. Global bee pollen consumption shows an upward trend with increasing numbers of consumers regarding bee pollen as a nutritional supplement. However, bee pollen consumption can cause a number of clinical allergic symptoms in certain individuals with allergies [4,5]. In addition, bee pollen can induce cross-allergic reactions when consumed with other foods [6–8]. The potential allergenicity of bee pollen has become one of the key issues limiting the development and utilization of bee pollen. Developing an efficient approach for reducing the allergenicity of bee pollen is thus necessary for expanding its further utilization.

Enzyme treatment is a critical food processing technique that can increase the nutritional value and reduce the allergenicity of foods. Furthermore, this technique is low cost and highly efficient [9]. Currently, α-chymotrypsin, trypsin and flavourzyme are widely used to reduce the allergenicity of certain allergenic foods, such as peanuts [10,11], soybean [11,12], and wheat [13]. Enzyme treatment has also been applied to break down the pollen wall and release nutrients contained within [14]. In our previous study, cellulase, pectinase and papain were combined and used for allergen degradation into small peptides and amino acids, resulting in decreased bee pollen allergenicity based on in vitro

assays [15]. In this study, we aim to further clarify the mechanisms of allergenicity alleviation of enzyme-treated bee pollen on the changes in serum allergic mediators, metabolic pathways, and gut microbiota composition.

Immunoglobulin (Ig) E-mediated type I hypersensitivity accounts for the vast majority of food allergies [16]. As a high-IgE response strain, the BALB/c mouse is suitable for IgE-mediated food allergy research, such as egg- [17,18], milk- [19], fish- [20] and peanutinduced [21,22] hypersensitive reactions. Recently, BALB/c mice were adopted to study the pathogenesis of allergy syndrome caused by oral pollen [23], providing the basis for the selection of an in vivo model in our study. Additionally, food allergies can activate the release of cytokines such as interleukin(IL)-4, IL-5, IL-13, and other mediators that can induce the production of IgE antibody in B cells [16]. Food allergies can also cause changes in host metabolism and gut microbiota. Some studies have applied metabolomics to investigate IgE-mediated food allergies [24–26]. A close association between food allergies and the dysbiosis of gut microbiota was also proposed in numerous studies [27–29]. The immune system can be influenced by host metabolic disorders and gut microbiota dysbiosis, although the related mechanism remains unclear [30].

To further clarify the mechanisms behind allergenicity alleviation of enzyme-treated bee pollen, the BALB/c mouse model was used to investigate the regulatory effects of enzyme-treated bee pollen on the production of serum allergic mediators, changes in metabolic pathways, and gut microbiota composition. Bee pollen samples were treated with a combination of cellulase, pectinase and papain, and mice were fed enzyme-treated and non-enzyme-treated bee pollen, respectively. The scratching behavior and histopathological injury were evaluated to identify the allergic state in mice. Subsequently, the mice serum was collected for allergic mediator and metabolite assays. Mice fecal DNA was extracted for gut microbiota composition analysis. Our findings might provide a basis for further development and utilization of hypoallergenic bee pollen products.
