Beyond olfaction: New insights into human odorant binding proteins.

1
INTRODUCTION
In this review, we discuss the roles and function of the two known human odorant binding proteins (OBPs) beyond their putative scope of olfaction, allowing us to understand human OBPs in a broader biological context. Since their first discovery in the olfactory mucus of cows (Bignetti et al., 1985; Pelosi et al., 1982), OBPs are believed to play a role in the initial stages of olfactory perception (Grolli et al., 2006; Lacazette, 2000; Lazar et al., 2002). Due to their occurrence in the nose, the subsequent studies mainly focused on understanding their role in the olfactory sense (Melis et al., 2021), where they are now recognized as players in peri‐receptor events (Heydel et al., 2013), that is, any biochemical interactions that can occur between odorants and proteins in the aqueous interface of the nose (Pelosi, 1996). Hereby, OBPs are thought to bind and transport hydrophobic odorant molecules by lowering the energy barrier through the aqueous environment to olfactory receptors (Paesani et al., 2025). They possess a strong hydrophobic calyx that can facilitate the binding of hydrophobic molecules like odorants. The typical beta‐barrel structure and the attached C‐terminal alpha‐helix exhibit a conserved structural feature across all mammalian OBPs, which suggests similar functions throughout different species (Pelosi & Knoll, 2022). While mammals possess hundreds of different olfactory receptors (ORs), which are able to distinguish between thousands of volatile odorant molecules, only a small number of OBPs, usually 2–3, were identified in the mammalian nasal mucus (two in the case of humans: OBP2A and OBP2B). Even with the exception of porcupines with eight OBPs (Dal Monte et al., 1991; Ganni et al., 1997; Pelosi & Knoll, 2022), the number is significantly smaller than the ORs and suggests that OBPs are mostly non‐specific binders that can transport a wide range of structurally different odorants and volatiles (Pelosi & Knoll, 2022). Humans have approximately 400 different ORs, which allows for the combinatorial detections of different odorants by the receptors (Orecchioni et al., 2022). The small number of OBPs together with their low specificity towards their ligands is in accordance with a putative transporter function (D'Onofrio et al., 2020; Pevsner et al., 1990). Still, despite their low specificity, it has been shown that it is possible to enhance their binding affinity towards targeted ligands through mutagenesis experiments (Zhu et al., 2020), as well as post‐translational modifications in vivo (Bouclon et al., 2017). Besides odorant transport, other functionalities, such as scavenging, that is, clearing ORs after molecular detection and preventing overstimulation and oversaturation, may also be relevant in the olfactory sense (Nakanishi et al., 2024). Figure 1 shows a simplified overview of the olfactory process at the olfactory cleft mucus (OCM), which represents the interface between air and the environment surrounding the olfactory neurons. OBPs are usually active in the monomeric form of the proteins. This is made possible through a highly conserved disulfide bond that fixates the C‐terminal alpha helix to the wall of the beta‐barrel. An exception is bovine OBP, which lacks the necessary cysteine residues and is only stable in the form of a dimer (Ramoni et al., 2008; Vincent et al., 2004). It should also be noted that despite their names, mammalian OBPs are not in any way related to insect OBPs. The latter belong to their own distinct insect OBP superfamily of proteins of alpha helical fold and have been reported to have very distinct functions and mechanisms (Abendroth et al., 2023; Rihani et al., 2021). Selected examples of insect OBPs depicting their characteristic alpha helical structure are provided in the supporting information (see Figure S1). The stable hydrophobic beta‐barrel structure of mammalian OBPs is a characteristic of the lipocalin protein superfamily to which they belong (CATH lipocalin/calycin fold class—unrelated to insect OBPs by fold class) (Orengo et al., 1997). This review, therefore, focuses only on mammalian OBPs and their link to other members of the lipocalin superfamily. These proteins possess a typical beta‐barrel formed calyx (with a binding pocket of approximately ~500 Å3) (Paesani et al., 2025) that harbors the characteristic hydrophobic pocket capable of transporting hydrophobic odorants in the aqueous media. The hydrophobicity distribution over the protein structure of hOBP2A is shown in Figure S2. Lipocalins are well studied and known to transport a variety of different hydrophobic molecules such as retinol, retinoic acid, aromatic compounds, and fatty acids (Breustedt et al., 2006). Altogether, 37 members of the lipocalin family have been identified in the human genome as reported by Du et al. (Du et al., 2015) who undertook a large‐scale systematic survey of human proteins to define the lipocalin family. In humans, lipocalins are found in blood plasma, tears, genital secretion, and are mainly active in transporting hydrophobic molecules and binding to cell surface receptors (Charkoftaki et al., 2019). Selected human lipocalins, all showing the typical beta‐barrel calix and the leaning alpha helix, and some of the corresponding ligands are depicted in Figures S3 and S4, respectively. Tables S1 and S2 provide an overview of the different members of the lipocalin family and their reported functions as listed in the Human Protein Atlas (HPA) and of the conserved regions within the human lipocalin family identified by Du et al., respectively.
Among the many human lipocalins, at least three are implicated in cancer (Table S1). Lipocalin 2 (LCN2), also known as NGAL, has been the most extensively studied. In ovarian cancer, LCN2 is markedly overexpressed and associated with tumor differentiation, where it contributes to iron sequestration and modulates the immune response by influencing inflammatory signaling and immune cell recruitment (Cho & Kim, 2009). In breast cancer, it induces epithelial‐to‐mesenchymal transition (EMT) by downregulating the estrogen receptor α (ERα), which increases the expression of Slug, a key EMT transcription factor, thereby promoting the mesenchymal phenotype and enhancing cell invasiveness and metastasis (Yang et al., 2009). In colorectal cancer, LCN2 is implicated in promoting tumor progression through the induction of the SRC/AKT/ERK signaling cascades (Zhang et al., 2021). These findings suggest that lipocalins may be involved in the tumor microenvironment and cancer cell physiology. However, the precise pathways through which lipocalins influence cancer remain to be fully elucidated. It should be noted that next to the impact of LCN2 in cancer, ORs have also been shown to regulate cancer cell activities, in addition to their core function of odor detection in olfaction (Chung et al., 2022).
So far, two OBPs have been reported in humans: hOBP2A and hOBP2B, that were initially identified based on their sequence identity (15.3%–40.6%) (Tegoni et al., 2000) to their rat OBP counterparts (OBPI and OBPII) (Pes et al., 1998; White et al., 2009). Overall, this level of conservation is consistent with the evolutionary divergence typically observed among lipocalin family members across mammalian species who, despite low pairwise similarity, share a common structural framework and conserved function. Previous MSA results across mammalian OBPs reveal low primary‐sequence identity but preservation of the lipocalin structural fold, including the conserved residues forming the β‐barrel and ligand‐binding pocket (Tegoni et al., 2000). Thus, despite limited sequence similarity, key structural motifs and overall topology are conserved throughout lipocalins (see Figure S3). hOBP2A has been reported to be expressed in the nasal structures, salivary and lacrimal glands, and the lung, while hOBP2B is expressed in genital sphere organs such as the prostate and mammary glands (Lacazette, 2000). A deeper understanding of hOBPs emerged with the elucidation of the crystal structure of OBP2A and with the detection of OBP2A within samples of the olfactory cleft mucus (OCM) (Débat et al., 2007). While it has been shown that olfactory detection is possible in the absence of OBPs (El Kazzy et al., 2025; Staiano et al., 2007), their reported presence in the mucus covering the olfactory cleft, where the sensory olfactory epithelium is located, suggests their role as odorant carriers (Briand et al., 2002), which would be in agreement with a function of “lowering” the barrier for odorant transfer from the gas to the solvated phase. To exert such a transporter function, OBPs need to be concentrated and operational at the air‐mucus interface. Several studies indicate that OBPs may facilitate the solubilization of hydrophobic odor molecules in the aqueous mucus layer, thus facilitating odorant transport through the hydrophilic mucus (Hajjar et al., 2006; Paesani et al., 2025). Nevertheless, labeling these newly discovered lipocalin members as odorant transporters may have restricted their study to putative functions in the olfactory sense, thus narrowing the possibility to identify and understand their full functionalities in other areas of the body. While the primary role of OBPs is traditionally associated with olfactory detection, emerging evidence suggests that these proteins also play significant roles in non‐olfactory tissues (Ferrer et al., 2016; Sun et al., 2018). Experimental evidence of OBPs localization in nasal glands and secretion emphasizes their secretory nature and potential involvement in protective and modulatory roles in epithelial barriers (Pevsner et al., 1986). OBPs show a functional diversity across organs, noting their expression in a variety of tissues, including reproductive and digestive organs, suggesting broader physiological roles (Sun et al., 2018). This could expand the relevance of OBPs from mere odorant carriers to multifunctional proteins essential in diverse physiological contexts beyond olfaction‐related tissues. Their expression in the reproductive organs, together with their ability to transport hydrophobic molecules such as hormones, is particularly intriguing as several cancer types are hormone dependent. Elucidating these two points in light of the existing literature and available data in several databases is the main objective of this review. We hypothesize and verify whether OBPs could have a function beyond mere odorant transport in olfaction and suggest a role as hormone transporters, with a contribution to tumorigenesis. While the focus of this review remains on OBP expression in cancer and tissues outside the olfactory sense, we mention potential implications in other diseases, such as Parkinson's disease (PD), when reported in the literature (Melis et al., 2019). In the case of PD, recent studies identified the tendency of bOBP to form larger protein aggregates under different experimental conditions, such as temperature, pH, concentrations of the protein, as well as through mechanical agitation and the addition of solvents such as hexafluoro‐2‐propanol (Stepanenko et al., 2023). The work highlights the potential usefulness of using OBPs to study amyloid fibril formation, and as model systems to better understand and mimic the oligomerization mechanisms underlying neurodegenerative diseases (Stepanenko et al., 2024; Sulatskaya et al., 2024). Still, in this structured review, we focus on investigating the potential involvement of OBPs in cancer, thus opening additional new avenues to understand and exploit the function of these proteins. This hypothesis is further supported by the diverse roles of lipocalins such as LCN2 in modulating immune responses and cancers (Rodvold et al., 2012), and isolated reports of potential hOBP2A involvement in prostate cancer (Jeong et al., 2023) and colorectal cancer (Cervena et al., 2021). In this review, we address the question of whether OBPs may have a broader impact and more sophisticated functions than anticipated so far. We focus on the putative functions of hOBP2A and hOBP2B outside of the olfactory sense, exploring their implications in cancer biology and the broader physiological roles they may play in human health. Figure 2 provides an overview of our approach, which leverages available information from the literature and various databases, such as the human protein atlas (HPA) (Uhlen et al., 2019; Uhlén et al., 2015, 2019) and the cancer genome atlas (TCGA) (Tomczak et al., 2015), to carry out a data‐driven study to elucidate the underlying function of OBPs under normal and cancerous conditions. In addition, insights from structural bioinformatics allow us to analyze the potential impact of mutations on the functions of these two proteins. Finally, we conclude our overview with a discussion of future directions and the most relevant applications of OBPs and other members of the lipocalin family, as provided by current literature.

2
OCCURRENCE OF OBPs IN TISSUES OUTSIDE OF THE OLFACTORY SENSE
2.1
OBP2A and OBP2B are expressed in tissues outside the olfactory sense and hormone related cancer types
To see whether OBPs are expressed outside of the olfactory sense, the HPA database was used (Sjöstedt et al., 2020; Thul et al., 2017; Uhlén et al., 2015), which provides an overview of reported RNA and protein expression levels throughout the body. For the two known human OBPs, OBP2A and OBP2B, the RNA‐expression data from the HPA showed that both proteins are noticeably expressed in healthy tissues unrelated to olfaction. Figure 3 shows an overview of the RNA expression levels in different tissues, including male and female tissues. Both OBP genes show high RNA expression in the male and female reproductive tissues. While OBP2A is expressed at a high level in female tissues (e.g., the fallopian tubes), OBP2B is significantly expressed in male tissues (testis). Since both OBP2A and OBP2B are predominantly associated with the olfactory system, this high expression in the female and male tissues raises the intriguing question of whether they contribute to the function or pathology of these non‐olfactory tissues. Additionally, as OBPs (and lipocalins in general) are known to bind hydrophobic molecules, it is noticeable that they are both concentrated in tissues whose functions are regulated by hormones. Given that the hormones of the reproductive systems, that is, estrogen, estradiol, progesterone, testosterone, are hydrophobic sterane derivatives similar to odorants (Oren et al., 2004), OBPs are therefore ideal candidates to function as hormone transporters. Based on this observation, we hypothesize that OBPs may play a role in hormone transport and/or hormone regulation in these tissues. The binding size of OBPs and lipocalins (~500 Å3) is sufficiently large to carry diverse ligands of the size of hormones (see also Figure S4 for an overview of reported ligands). In addition, Figure 3 shows that skin tissue also exhibits sensible levels of expression. Like all the expressions in tissues outside the olfactory sense, this suggests additional unexplored roles of OBPs within the human body. While this remains speculative, it could indicate possible binding and transport of volatile or hydrophobic compounds at the skin's surface. OBPs may therefore contribute to chemical communication, protection against xenobiotics, or local metabolic processes. A recent study by Nakanishi et al. supports this direction (Nakanishi et al., 2024).
For a more comprehensive investigation, we dive into protein expression levels of OBP2A and OBP2B in different reproductive tissues as reported in the Protein Abundance database (PaxDB) (Huang et al., 2023; Wang et al., 2015) and other proteomics repositories (Lonsdale et al., 2013; Thangudu et al., 2024). Interestingly, the reported protein expression levels concentrated in the male and female reproductive tissues are higher than the expression levels in the olfactory system. This may indicate that only low expression levels of OBPs are required for functionality in the olfactory sense and potentially support the hypothesis that odorant transport through the mucus may also be achieved spontaneously without OBPs; however, most likely at a slower rate.
In addition to expression levels in healthy tissues, the HPA further reports the RNA and protein expression of OBPs in different cancer tissues, such as breast and ovary (see Figure S5 for details). Moreover, from the Proteomic Data Commons (PDC) portal (Thangudu et al., 2024), which shows protein expression of OBP2A and OBP2B in cancer cases, OBP2A is differentially expressed in ovarian cancer and OBP2B in breast cancer. Since it is known that hormones of the reproductive system play a pivotal role in the development and progression of certain cancers, such as breast (Yang et al., 2009), ovarian (Cho & Kim, 2009), and prostate cancers (Jeong et al., 2023), we hypothesize that OBPs are likely to interact with the hormones present in these tissues. If OBPs indeed have a hormone‐transporting function in reproductive tissues, it is worthwhile investigating whether they may also be implicated in the molecular mechanisms underlying hormone‐driven cancers. An overview of the RNA and protein expression of OBP2A and OBP2B reported for both normal and cancer tissues in different databases is provided in Table 1. Due to their characteristics as small secreted proteins that can transport hydrophobic molecules, they can play roles in lots of physiological activities such as immunity, metabolism, and cellular signaling (Steinbrecht, 1998). Olfactory receptors are included in this overview to demonstrate that at least two receptors, OR51E1 and OR51E2, are also expressed in non‐olfactory tissues such as normal reproductive organs (see Figure S6) and cancer tissues (Chung et al., 2022; Li et al., 2021; Orecchioni et al., 2022). For a complete overview of the reported RNA and protein expression patterns of human lipocalins and ORs across different tissues, and for an overview of the known mechanisms of lipocalins in cancer, see supporting materials (Table S1 and Figure S7). In the case of human lipocalins, almost every family member displays RNA expression in cancer (except for LCN9). However, protein expression in cancer tissues is only reported for half of them. Interestingly, the reported protein expressions of lipocalins are mainly related to reproductive cancers and several lipocalins have been directly linked to tumorigenesis and metastasis (Bratt, 2000; Du et al., 2015), for example, LCN1 and LCN2 in breast (Yang et al., 2009) and uterine cancer (Mannelqvist et al., 2012) (Table S1). The ectopic expression of ORs is shown across multiple cancer types, including breast, prostate, and lung cancers (Chung et al., 2022). The functional roles of ORs appearing in reproductive tumor microenvironments could be a potential connection with the influence of OBPs in those cancers. Especially with ORs frequently being expressed ectopically in reproductive tissues and in multiple tumor types, where they act as bona fide signaling GPCRs that modulate proliferation, migration, metabolism and immune‐related pathways (Chung et al., 2022; Gelis et al., 2016). If OBPs are present in the same tumor microenvironment, as suggested in some studies (Jeong et al., 2023), they could change the local presence of hydrophobic ligands (such as steroids, fatty acids, volatile metabolites, or terpenoids). There, they could deliver OR agonists to receptors on tumor or immune cells, reduce OR activation (via ligand binding) or scavenge signaling molecules (e.g., steroid metabolites) that act on ORs. Such OBP–OR interactions could therefore provide a plausible mechanistic bridge whereby OBP expression in tumors influences OR‐driven signaling and so affects tumor cell behavior and the immune/tumor microenvironment (e.g., growth, therapy resistance, myeloid infiltration). Furthermore, the pro‐metastatic role of OR5B21 in breast cancer reveals that ORs promote cancer invasion and metastasis through epithelial‐mesenchymal transition (EMT) (Li et al., 2021). This may hint that ORs play a role in aggressive tumors. Expanding on this, the overexpression of OR2T6 has been reported to the invasive capabilities of cancer cells (Li et al., 2019). The above evidence provides an idea that ORs can act beyond sensory perception and contribute to oncogenesis. It should be noted that ORs (see also Introduction section), which are the key to olfactory detection mechanisms, have been shown to play a role in cancer progression (Chung et al., 2022). Nevertheless, it is possible that OBPs are simply co‐expressed as a side‐effect of ORs or lipocalins. Another interesting aspect about the function of OBPs may be if they are able to bind specifically to ORs or membranes although this has not been explicitly shown for mammal OBPs (Pelosi, 1998). At a higher level, lipocalins are also known to interact with membrane receptors, e.g., megalin (LRP2) is a broad receptor of lipocalins and LCN1 and LCN2 have known binding sites for molecular uptake (Cabedo Martinez et al., 2016; Wojnar et al., 2003). Although direct binding to an OR may not be necessary for OBPs within their function in the olfactory sense, it cannot be ruled out that interactions within other membrane proteins or membranes in general play a role in other tissues or cancerous environments.

2.2
OBP2A and OBP2B share a high sequence similarity
As protein structure is directly linked to protein function, we have looked into the experimental crystal structure of OBP2A (PDB‐ID: 4RUN; Schiefner et al., 2015) and the AlphaFold3 (Jumper et al., 2021) predicted structure of OBP2B along with their sequence and corresponding point mutations to further elucidate their potential functions. The two proteins exhibit a sequence identity of 90% with a total of 17 different residues (over the full length of the protein – 170 residues for OBP2A). The corresponding pairwise alignment shown in Figure 4b illustrates the high similarity of the two sequences. Figure 4a highlights the positions of the amino acids that differ from OBP2B. These 17 non‐conserved residues are spread over the entire protein chain, including the alpha‐helix and the beta‐barrel, which do not suggest different functionalities between the two proteins (see Table S3 and Figures S8 and S9 for details). Rather, with the high sequence similarity, the structure of OBP2B can be expected to be close to identical to OBP2A, and both proteins will probably be able to transport a large amount of similar hydrophobic molecules. In addition to our own preliminary docking experiments using OBP2A combined with testosterone or estradiol (see Figures S15 and S16 and Tables S7 and S8), lipocalins are known to have binding pockets sufficiently large to carry large hydrophobic molecules (such as fatty acids, terpenes—see also Figure S3 for an overview). The conserved hydrophobic residues throughout the structures are crucial to maintain the typical beta‐barrel structure of the transporter protein. Both proteins are therefore also likely to perform similar functions. Still, the five residues that shift from hydrophilic to hydrophobic physicochemical properties that are partially located close to the binding pocket may impact the affinity or binding properties of proteins towards different ligands or to yield different roles in a given tissue type (see Figure S9). Changes may include tissue‐specific expression or regulatory properties. Despite the high similarity at the protein level, they may be differentially regulated by different promoters or tissue‐specific signals. Using porcine OBP as a case study, previous works showed that selected mutations within the binding pocket can even lead to chiral discrimination between the enantiomers of carvone (Paesani et al., 2025). To get a full understanding of hOBPs and their relationship to other members of the lipocalin family, as well as ORs, we verified the chromosomal location reported in genecards (Safran et al., 2021). Conserving sites of the coded gene of human odorant‐binding proteins, the hOBP2A and hOBP2B genes are located close by on chromosome 9, with hOBP2A and hOBP2B on Chr 9q34.2 and Chr 9q34.3, respectively (Safran et al., 2021). Figure 4c shows the distribution of the different lipocalins on chromosome 9. Interestingly, most kernel lipocalins are in the same position as OBP2A in Chr 9q34.3. LCN2 exhibits a unique locus in Chr 9q34.11. In contrast, ORs are located on chromosome 11.

2.3
Network and co‐expression patterns of OBP2A and OBP2B
Similar to mammalian OBPs, insect OBPs have been reported to be expressed in specific tissues including antennae as well as reproductive and gustatory tissues (Hu et al., 2016). However, they have been shown to be co‐expressed with olfactory receptors and to cluster on co‐expression experiments and corresponding protein homology with genes involved in signal transduction (Mika & Benton, 2021; Task et al., 2022), detoxification, and neural signaling, which all strongly suggest their implication in olfactory pathways (see reported network in the STRING database (Huang et al., 2023)—data not shown). In mammalian, and particularly human olfaction, this is not as straightforward. Although independent sources have identified OBP2A and OBP2B to be expressed in the nasal mucus (Verbeurgt et al., 2014), there are also studies reporting that unlike mouse OBPs, human OBP2A and OBP2B do not show signs of enhanced expression in sensory tissue (Olender et al., 2016). Figure 5 visualizes the known protein connection network between OBP2A and OBP2B as reported in the STRING database (Huang et al., 2023). There, OBPs are indeed only reported to be co‐expressed with other lipocalin family members, most likely due to their location on chromosome 9 (see also previous section), but there is no reported co‐expression with other proteins from the olfactory system. The lack of identified gene expression networks and pathways for human OBPs makes it difficult to straightforwardly associate their role in olfactory pathways. While their expression is most likely directly linked to lipocalin expression, this does not necessarily exclude the putative implication of OBP2A and OBP2B in cancer progression or other functions across the body. Note that the reported links are not solely based on text mining but that several co‐expression patterns are reported between OBP2A and OBP2B, as well as to other members of the family, namely LCN8 and LCN12, which are both predominantly reported in the male reproductive tissues in the HPA. It should also be noted that the STRING database does not report any link to LCN1 and LCN2 in the network. Altogether, it can be said that the role of OBPs in human olfaction is still disputed, these proteins are rather intriguing, and their initial name may indeed have been hurried and thus limiting to fully study and exploit the functionality within the context of olfaction and beyond. Within the notable expression profiles of OBPs in reproductive tissues and some cancer types, fully understanding them will require going beyond their expression profiles, and looking closer into their mutation profiles and the structural and functional implication that these may involve. This is done in the following sections.

2.4
The genomic landscape of different cancer types indicates alterations in OBP genes
Using the available data from the HPA, we performed differential gene expression (DGE) analysis to determine whether their expression is significantly altered in different cancer types. Specifically, DGE analysis was conducted to identify the expression patterns of OBP2A, OBP2B, and other lipocalins, as well as known cancer genes (cf. Table S4 for the list of included known cancer biomarkers) in six different cancer types: ovarian, breast, uterine, prostate, colorectal, and lung cancers (TCGA database; Tomczak et al., 2015) and normal tissue (GTEx database; Lonsdale et al., 2013). The datasets used were screened out based on the filter parameters shown in Table S5. The results of the DGE for ovarian cancer are depicted in the volcano plots shown in Figure 6 (see Figure S10 for the remaining breast, uterine, prostate, lung, melanoma plots). The analysis reveals that both OBPs are upregulated in ovarian and breast cancer, while OBP2B is upregulated only in ovarian cancer.
Given the strong link between genetic mutations and cancer, we also examined the mutational profiles of OBP2A, OBP2B, and other human lipocalins using data from TCGA. From the oncoprint shown in Figure 7, it can be seen that the FABP family is most frequently mutated, with a mutation frequency of 53% among the human lipocalins across the cancers in the TCGA. APOD ranked in second place with a mutation rate of 48%. Most of the lipocalins were affected in at least 40% of patients. The lipocalins were mostly affected by amplifications and deletions that can span several genes. There is a high co‐occurrence of these mutations across the lipocalins that is potentially due to the close proximity of these genes on chromosome 9.

2.5
Statistical analysis of the mutational landscape in different cancer types shows that copy number alterations (CNAs) represent the only significant genomic alteration, while the reported single nucleotide variants (SNVs) are unlikely to affect protein function
The DGE analysis did not confirm an overexpression of OBP2A in prostate cancer, as others have observed in prostate cancer tissues that had shrunk during remission under androgen deprivation therapy (Jeong et al., 2023). However, the study suggests that OBP2A overexpression may be specifically induced by an androgen‐deprivation state (Jeong et al., 2023), implying that the overexpression of OBP2A might require a very specific cancer microenvironment. This is in agreement with other studies in mammals that indicate the expression of OBPs may be influenced directly by external factors and developmental conditions (Gonçalves et al., 2021; Kuntová et al., 2018; Todrank et al., 2011). As for lipocalins among the six cancer types (ovarian, breast, uterine, prostate, lung and colorectal) analyzed in this study, LCN2 serves as an active factor in oncogenesis, and APOD was found to be downregulated in four of the selected cancer types (Figure S10). To assess if the differential expression of OBPs is due to mutations affecting them, namely copy number alterations (CNAs) and single nucleotide variants (SNVs), we have compared the expression of OBPs between patients with and without these mutations. OBP2A and OBP2B showed a large increase in gene expression in patients with a copy number gain of the gene in ovarian cancer and a moderate decrease in expression when the gene was lost (Figure S11). Breast and uterine cancer showed less pronounced differences compared to ovarian cancer. SNVs showed no change in expression. To further assess if the mutations within OBP2A and OBP2B have an impact, we have utilized the computational method CIBRA to measure the system‐wide impact of these mutations (including gains and losses). From the CIBRA analysis, CNAs in OBP2A and OBP2B are associated with significant changes in the system‐wide expression profiles of ovarian, breast, and uterine cancer (Figures 8 and S12).
While the CIBRA results showed that copy number changes can affect the expression level of OBP2A and OBP2B, SNVs did not result in expression changes. The mutations could still affect the protein structure and thereby the function.
To determine whether patient‐derived SNVs may have an influence on protein function, the physical–chemical properties of the corresponding mutated amino acid residues were assessed. Of all 44 mutations that were detected in OBP2A, 10 residues changed from hydrophilic to hydrophobic, and three changed the other way around. For OBP2B, 31 mutations were observed (see Figure S13). Among them, four changed properties from hydrophilic to hydrophobic, two from hydrophobic to hydrophilic. Most of them belong to missense mutations. The location of those mutated residues highlighted in structures respectively in OBP2A (PDB‐ID: 4RUN) and OBP2B (AlphaFold3 predicted structure) is depicted in Figure 9. To calculate the expected number of mutations due to random chance, we first determined the average mutation counts from the cancer type exhibiting the highest frequency of mutations to verify that the frequency of mutations is higher than random chance and meaningful (see Table S6 for further information).
Altogether, this does not directly imply a drastic change of functionality, but it is possible that at least the six mutations that include hydrophobic residues also impact the uptake capability and alter the binding profile of the proteins. Previous experimental works showed that mutated porcine OBP exhibits higher binding specificity towards carvone and may even be used to distinguish between different enantiomers of the molecules (Mulla et al., 2015). It may also be interesting to explore whether the reported mutation can impact the ability of OBPs to bind membrane receptors or ORs, a topic on which little is known so far.

3
FUTURE PERSPECTIVES AND APPLICATIONS OF OBPs
While the exact biological function of OBPs in human olfaction remains largely unknown, OBPs and lipocalins in general remain interesting candidates for a variety of biomimetic applications. Their size and robustness provide ideal conditions to explore their usage in molecular detection devices, such as artificial noses and biomarker sensors, or as drug and flavor delivery agents. Current efforts in OBP research are mostly directed on applications related to their putative function in the olfactory sense (Zahra et al., 2024), thereby focusing on increasing their specificity for targeted air‐quality monitoring, food quality assessment, and medical diagnostics (Hellmann, 2020). Over the past decades, several studies emerged with the aim to fabricate individual sensors, tuned to identify and detect volatile organic compounds with high specificity (Hurot et al., 2019, 2020; Pelosi et al., 2018a). However, until now most electronic noses use only poor specificity gas sensors based on metal oxides and conducting polymers (Zhai et al., 2024). The major drawback of these gas‐sensing materials is doubtlessly their low selectivity and stability. Recent trends in gas‐phase biosensing showed an opening towards different protein‐based elements: odorants receptors, OBPs, and odorant receptor peptides (Barbosa et al., 2018). Hereby, OBPs stand out as suitable detecting elements for artificial gas and smell sensing (Pelosi et al., 2018b). In contrast to membrane‐bound receptors which, despite their high selectivity (Haag & Krautwurst, 2022; Li et al., 2016), are highly unstable and sensitive to external stress, and odorant receptor peptides for which only a very limited number of suitable peptides is known for chemical sensing, OBPs are extremely stable to high temperature, refractory to proteolysis and resistant to organic solvents (Pelosi et al., 2018b). Preliminary studies showed that porcine OBP can be tailored towards higher specificity to distinguish between carvone enantiomers (Paesani et al., 2025). Fine‐tuning the binding properties of OBPs and lipocalins towards the detection of specific molecules or molecular families provides a stepping stone for artificial sensor devices and can open up applications in the biomedical field (Haag & Krautwurst, 2022; Mulla et al., 2015; Pelosi et al., 2018a), leveraging OBPs to detect biomarkers in the breath of patients to diagnose diseases (e.g. hexanal in lung cancer (Mousazadeh et al., 2022)). Additionally, machine learning algorithms are being employed to analyze the complex data generated by these OBP‐based sensors, enabling improved odor recognition patterns and enhancing the overall performance of artificial olfactory systems (Hellmann, 2020; Lötsch et al., 2019).
Closely linked to olfactory research is flavor research, which combines taste (gustation) and smell (olfaction). In flavor research, further exploring the transporter functions of OBPs to improve the properties of flavor distribution in different food matrices could be an interesting opportunity (Boichot et al., 2022). OBPs may be leveraged to encapsulate small odor and flavor molecules, thus preventing them from reaching taste receptors too fast (Ren et al., 2024). They could be fine‐tuned towards “masking” odors or towards reducing the effect and perception of specific odor and flavor molecules (Gascon, 2007). Similarly, OBPs can be targeted to transport drugs to specific locations inside the body (Choi et al., 2022) or designed to target specific cells or tissues (Nästle et al., 2023). A benefit of leveraging these proteins as drug carriers is their limited immune response as direct members of the organism (Yang et al., 2023). Their potential to seize harmful compounds makes them favorable candidates for applications in detoxification (Grolli et al., 2006; Nakanishi et al., 2024). Thus, due to their ability to bind and transport hydrophobic molecules, OBPs and lipocalins may be engineered towards different applications that go beyond their putative roles in the olfactory sense. Finally, recent studies on the aggregation of bOBPs highlighted the usefulness of leveraging the properties of hydrophobic beta‐barrel to study fibrillogenesis in neurodegenerative diseases (Stepanenko et al., 2023; Sulatskaya et al., 2024). This could provide another exciting domain for OBPs in bio‐chemical applications and lead to the development of new therapeutic strategies to prevent pathological beta‐barrel aggregation.
While the present outlook emphasizes the roles of OBPs in olfaction and artificial biosensors, our results, together with recent evidence, suggest that OBPs and other members of the lipocalin family may have the potential to gain further relevance in oncology (Krizanac et al., 2024).Notably, the secreted lipocalin LCN2 is overexpressed in multiple solid tumors, where it correlates with poor prognosis and progression of hormone‐driven cancers such as prostate cancer, and is actively being explored as a biomarker and therapeutic target (Cho & Kim, 2009; Ding et al., 2016; Santiago‐Sánchez et al., 2020). Given that hOBP2A and hOBP2B are lipocalins expressed in hormone responsive tissues (e.g., prostate, mammary gland) and are known to carry hydrophobic ligands, these proteins may similarly serve as biomarkers (via detection of their secreted forms) or therapeutic targets (by modulating ligand transport or signaling) in hormone‐dependent malignancies (Lalis et al., 2025). Therefore, future study should evaluate (i) whether circulating levels of hOBP2A/B correlate with cancer stage or hormone treatment response, (ii) whether functional blockade of hOBP2A/B affects tumor cell viability in hormone‐dependent models, and (iii) whether ligand‐binding specificity can be leveraged in targeted delivery platforms for hormone‐driven malignancies.

4
CONCLUSION
In this systematic review, we showed that despite their fame in context with the olfactory sense, OBPs are overexpressed in various tissues throughout the human body, especially in the male and female reproductive organs. Based on this observation and the ability of OBPs to bind hydrophobic compounds, we hypothesized their potential function as hormone transporters. Based on their overexpression in the reproductive organs, we further explored their potential implications in cancer tissues of the reproductive organs, since several cancer types are hormone dependent. While our review clearly showed that human OBPs are overexpressed in both healthy tissues, the statistical analysis using CIBRA to evaluate the impact of the underlying genomic alterations in cancer tissues remained inconclusive. Still, it cannot be ruled out with certainty that OBP could function as a hormone transporter and participate in cancer progression and tumor genesis. This is also supported by the size of the binding pocket, the multiple reported hydrophobic ligands binding other members of the lipocalin family as well as our own preliminary in silico docking experiments of hOBP2A with testosterone and estradiol. Our suggestion that odorant‐binding proteins (OBPs) may also function as hormone transporters is therefore primarily based on their lipocalin fold, which is known for binding a broad range of hydrophobic ligands, and on the co‐expression patterns with hormonally responsive tissues. Regarding experimental evidence, in vitro studies have demonstrated that porcine OBP can bind steroids and fatty acids with measurable affinities, as shown by fluorescence spectroscopy assays using native and recombinant isoforms (Briand et al., 2019). However, direct in vivo evidence for OBP–hormone interactions in humans is still lacking. In future, to further strengthen the hypothesis of this work and verify its validity, more advanced experiments are needed, for instance using expression profiling and knockouts in cancer cells, possibly with a focus on ovarian and breast cancer. Overall, integrating the study of OBPs with other members of the lipocalin family and possibly other olfactory proteins such as ORs could help our understanding of the diverse functions of OBPs in the human body.

5
METHODS
5.1
Structured literature overview
To provide a structured overview of OBPs, their function, characteristics and potential applications, we systematically searched PubMed, Scopus, and Web of Science databases for publications related to OBP2A/2B gene expression and their relevance in reproductive cancers. The search covered all available literature without any restriction on the publishing date. The provided references cover publications from 1982 to 2025, using combinations of the terms: “OBP2A/2B”, “olfactory binding proteins”, “ovarian cancer”, “breast cancer”, “uterine cancer” and “gene expression”, connected by Boolean operators (AND, OR). Only English‐language, peer‐reviewed original research articles were included. While the focus of the work is on human olfaction and tissues, we mention non‐human examples and studies from insects and mammals when needed to get a full overview of the problem. The authors screened titles and abstracts independently, followed by full‐text screening and discussed the results and main findings together.

5.2
Metadata analysis
Figure 10 shows a schematic overview of the meta‐analysis approach using research questions from the structured literature overview.
Starting from RNA‐seq expression data to (Q1) determine whether OBPs are overexpressed in any cancer types, (Q2) to confirm links to underlying genomic alteration, (Q3) verify their significance, (Q4) check if the overexpression of OBP2A and OBP2B genes is caused by neighboring driver genes of reproductive cancers that are located on the same locus (chromosome 9) as OBPs, and (Q5) gain insight from a structural perspective by leveraging pairwise alignments and previously reported multiple sequence alignments (MSA) and impact of identified structural mutants. The technical details for each of the individual steps are provided below. We first performed differential gene expression analysis of OBPs in several cancer types, with the focus on reproductive tissues: ovarian, breast, prostate and uterine cancer. To further understand the underlying mutational profile in terms of cancer related CNAs and SNVs, we used CIBRA, a computational method to assess the effect of these genomic alterations on system‐wide gene expression (Lakbir et al., 2024) along with MSA and structural variants analysis for full mutational overview.

5.3
Databases
5.3.1
HPA, PaxDB and PDC
The data is retrieved from the Human Protein Atlas (HPA), Protein Abundance Exchange Database (PaxDB), and Proteomic Data Commons (PDC). Queried series are generated based on clinical data of OBP2A and OBP2B in various cancers, including ovarian, breast, uterine, prostate, melanoma, lung, and colorectal cancers.

5.4
The Cancer Genome Atlas (TCGA) data
Clinical information and mutation calls for TCGA were obtained from cBioportal. Gene expression data quantified as counts were retrieved from the ovarian, breast, prostate, uterine, melanoma, lung, and colorectal cancers (harmonized with GTEx). The RNA‐Seq data has been normalized by calculating the Transcripts Per Kilobase Million (TPM) or Fragments Per Kilobase Million (FPKM).

6
ANALYSIS OF SEQUENCES, EXPRESSION, MUTATION SIGNIFICANCE, AND DRIVER GENES
6.1
Multiple sequence alignment (MSA)
The MSA was performed with Clustal Omega (Sievers et al., 2011) with default parameters using the Job Dispatcher from the European Bioinformatics Institute (EMBL‐EBI). Protein sequences from human lipocalins and mammalian OBPs were retrieved from UniProt. The MSA was conducted with these protein sequences and visualized using JalView (Troshin et al., 2011, 2018) with default parameters. The correlation identity was generated using the NCBI BLAST (Basic Local Alignment Search Tool) (Altschul et al., 1990; Camacho et al., 2023) with corresponding protein sequences from UniProt.

6.2
Differential expression analysis
For differential expression (DE) analysis between normal and cancer tissue, harmonized and normalized gene counts were retrieved from the UCSC Toil Recompute Compendium (Lonsdale et al., 2013) for the TCGA and GTEx samples to account for the differences between the cohorts. DE analysis was performed using the R package limma (version 3.28.14, default parameters) and edgeR (3.14.0 version, default parameters) following the protocol from Chen et al. (Chen & MacDonald, 2022). Genes with counts <10 or no variance were filtered from the analysis. The results were subsequently visualized using volcano plots using the EnhancedVolcano R package (version 1.27.0) (Blighe, 2018). Genes with an adjusted p‐value <0.05 were deemed differentially expressed. To estimate expected lipocalin exon mutations across cancer types, we calculated a background mutation rate using the most mutated cancer type, applied it to lipocalin exon lengths, and scaled the result by the number of samples per cancer type.

6.3
Computational identification of biologically relevant alterations (CIBRA)
To assess whether mutations in the OBP genes have an impact in cancer, we made use of CIBRA, a computational method to assess the system‐wide impact of mutations on gene expression. Unprocessed gene counts for the TCGA retrieved from cBioPortal were used with CIBRA. Only comparisons with at least 10 samples with and without mutations were considered (default parameters of the tool). DESeq2 (version 1.40.2) was used as a differential expression analysis method within CIBRA. All other parameters were left as default. Statistical evaluation of the CIBRA impact scores was performed by comparing the scores of the differential subsamples with a reference distribution of 1000 permuted CIBRA runs for each cancer type. Comparisons were made with the mutation categories of OBP defined as gain, loss, deep deletion, amplification, or single nucleotide variant (SNV) compared to WT samples without mutations in the OBP gene.

AUTHOR CONTRIBUTIONS

Mifen Chen: Software; investigation; visualization; formal analysis; writing – original draft; validation; data curation; writing – review and editing; methodology. Soufyan Lakbir: Investigation; writing – review and editing; methodology; validation; software; formal analysis; data curation; supervision. Mihyeon Jeon: Visualization; formal analysis; software; writing – review and editing. Vojta Mazur: Visualization; formal analysis; software; writing – review and editing. Sanne Abeln: Writing – review and editing; conceptualization; supervision. Halima Mouhib: Funding acquisition; supervision; conceptualization; writing – review and editing; project administration; writing – original draft; visualization.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.

Supporting information

DATA S1. Supporting Information.

FIGURE S1. Insect OBPs structures: helices (red), coils (yellow). It should be noted that Agam OBP1 (PDB‐ID: 2ERB) is reported as a dimer, indicating molecules passing through channels to both chains, while all others are reported as monomers. It should be noted that insect OBPs are not related to mammalian OBPs and lipocalins in any way, although they perform similar functions, i.e., transporting odorants. For instance, the olfactory processes involving insect OBPs are pH dependent. Here, the interested reader is encouraged to check the recent review by Briand et al. for further details.

FIGURE S2. Hydrophobicity visualization of hOBP2A (PDB ID: 4RUN). Cartoon representation of human OBP2A, color‐coded to illustrate hydrophobicity across the structure. The applied scale for the hydrophobicity visualization using Pymol is based on the work of Eisenberg et al.

FIGURE S3. Protein structures of different human lipocalins. The colors red, yellow, green depict the alpha helices, loops, and beta‐sheets of the structures, respectively. Note that the structure of Odorant Binding Protein 2B was generated using AlphaFold. For the experimental structure of OBP2A see Figure 1 in the manuscript.

FIGURE S4. Chemical formulas of known ligands bound by lipocalins. Beneath each formula, the corresponding lipocalin protein and the name of the ligand are listed.

FIGURE S5‐1. RNA and protein expression distribution of OBP2A and OBP2B across various cancer tissues from TCGA dataset. RNA expression of OBP2A (A) and OBP2B (B). The X axis presents cancer types. The Y axis displays Transcripts Per Kilobase Million (TPM). The central line within each box represents the median. The outlier data points that differ significantly from the rest are shown in the dots. Pink‐colored boxes indicate high expression levels.

FIGURE S5‐2. RNA and protein expression distribution of OBP2A and OBP2B across various cancer tissues from TCGA dataset. Protein expression of OBP2A (not available for OBP2B on the HPA). The X axis shows several cancer types. The Y axis presents the percentage of patients. Each box is color‐coded corresponding to each cancer type.

FIGURE S6. Differential gene expression analysis of odorant receptors in prostate Cancer. Volcano plots display the differential gene expression between normal and tumor samples in prostate cancer. The x‐axis represents the log2 fold change in gene expression between cancer and healthy tissues, while the y‐axis represents ‐log(False discovery rate), indicating statistical significance. The same human lipocalins and cancer markers are labeled as in Figure 6 of the manuscript, with genes colored blue for down‐regulated and red for upregulated. OR51E2 and OR51E1 genes were upregulated while OR2T6 and OR5B21 were filtered out due to their low counts.

FIGURE S7. Overview of the known mechanisms and effects of lipocalins in cancer. The figure provides an overview of selected members (LCN2, APOM, RBP4, PAEP and CRABP2) of lipocalin family proteins that contribute to cancer progression. It highlights the key pathways and reactions of them in the tumorigenesis, including up and down regulation on each track. Remarkably, the mechanism by which PAEP influences cancer remains to be elucidated.

FIGURE S8. Beta barrel and alpha helix displayed in sequences on crystal structure of OBP2A (PDB‐ID: 4RUN) to show in regions. Color‐coded: red (coils and loops); Purple (betabarrel); yellow (alpha‐helix).

FIGURE S9. Visualization of different residues between OBP2A and OBP2B on the structure of OBP2A (PDB‐ID: 4RUN). The five highlighted residues between OBP2A and OBP2B shift their properties from hydrophobic to hydrophilic or vice versa. Hydrophilic residues (R108) are highlighted in orange, hydrophobic residues (Y109, P118, L122, M144) are highlighted in blue (see also Table S2 for full information).

FIGURE S10‐1. Differential Gene Expression Analysis of OBP2A, OBP2B, human lipocalins in breast, uterine, and prostate (cf. Figure S9‐2 for full description).

FIGURE S10‐2. Differential Gene Expression Analysis of OBP2A, OBP2B, human lipocalins in lung and colorectal cancer. Volcano plots display the differential gene expression between normal and tumor samples in five cancer types. The x‐axis represents the log2 fold change in gene expression between cancer and healthy tissues, while the y‐axis represents ‐log(False discovery rate), indicating statistical significance. The same human lipocalins and cancer markers are labeled as in Figure 6 of the manuscript, with genes colored blue for down‐regulated and red for up‐regulated.

FIGURE S11‐1. Normalized RNA expression of OBP2A and OBP2B genes in (A) ovarian cancer, (B) breast cancer, and (C) uterine cancer. See continued caption of Figure S10‐2 for full description.

FIGURE S11‐2. Normalized RNA expression of OBP2A and OBP2B genes in (D) prostate cancer, and (E) melanoma cancer. The boxplots compare normalized RNA expression levels across different cancer types in different categories for the genes OBP2A (left) and OBP2B (right). In X‐axis, the categories include: “WT” (wild‐type), “Loss” (copy number loss), and “Gain” (copy number gain). WT (Wild‐Type): Represents tumors without mutations in OBP2A or OBP2B. Loss: Represents tumors where OBP2A or OBP2B has a deletion (loss of function). Gain: Represents tumors where OBP2A or OBP2B is amplified (increased expression). SNV (in uterine cancer): Represents tumors with single‐nucleotide variants (point mutations) in OBP2A. Colors indication: Green (WT); Red (Loss); Blue (Gain). Y‐axis is Normalized RNA expression (on a logarithmic scale). The spread of data points (gray dots) within the boxplots indicates the variability in normalized expression level across samples in each category. The wider spread suggests more variability in RNA expression among these samples, which has narrower variability. There are several outlier observations visible as points beyond the boxplot whiskers, especially in the WT category, highlighting some samples with exceptionally high expression levels. Some points in each category fall outside the whiskers, indicating potential extreme values in the dataset. Displayed median values in bold inside the boxes.

FIGURE S12. CIBRA impact scores for OBP2A and OBP2B copy number alterations (CNA) across breast (upper), and uterine cancers (below). This figure illustrates the CIBRA impact scores for loss and gain events in OBP2A and OBP2B across breast cancer (upper part of the figure) and uterine cancer (lower part of the figure). Each bubble represents the significant area of impact scores associated with either copy number loss or gain, as analyzed by the CIBRA pipeline. The size of the bubbles indicates the percentage of cases (%) affected by each alteration, and the color gradient corresponds to the p‐value, where darker red shades denote stronger statistical significance (lower p‐values); Light red suggest less significant results (higher p‐values). The vertical axis represents the CIBRA impact score (significant area), which reflects the extent to which each CNA (loss or gain) influences the molecular profile in these cancers; higher values suggest a greater impact of the genetic alteration. For X‐axis, loss: tumors where the gene is deleted (reduced copy number); gain: tumors where the gene is amplified (increase copy number). Error bars represent the interquartile range (IQR) of the data. A loss implies shallow deletion which could be a heterozygous deletion. The gain indicates a low level of additional copies, which is usually in a wide range.

FIGURE S13. OBP2A and OBP2B mutations across various cancers. 44 mutations in OBP2A (left, PDB‐ID: 4RUN) and 31 mutations in OBP2B (right, AlphaFold3 predicted structure). It can be seen that the mutations are almost evenly distributed over the two protein structures.

TABLE S1‐PART 1. Overview of relevant information on human lipocalins and ORs. The functions are provided as reported in the Human Protein Atlas, GTEx, PaxDB, and PDC.

TABLE S1‐PART 2. Overview of relevant information on human lipocalins and ORs. The functions are provided as reported in the Human Protein Atlas, GTEx, PaxDB, and PDC.

TABLE S2. Conserved regions of human lipocalins identified by
Du
et
al. (see main text for further details).

TABLE S3. List of the different residues between OBP2A/2B as obtained through MSA (with a total 17 residues that differ between OBP2A and OBP2B). The residue numbers correspond to the crystal structure of OBP2A (PDB‐ID: 4RUN). The four residues in blue bold change their properties from hydrophobic to hydrophilic, while the orange one changes from hydrophilic to hydrophobic (see also visualization in Figure S8). These changes are mostly likely to impact the binding specificity between the two proteins.

TABLE S4. Known biomarkers for different cancer types. List of known protein markers used in differential gene expression analysis for different cancer types. These biomarkers served as reference points to assess the relationship between OBP overexpression and significant genes in each cancer within the volcano plots generated in this review.

TABLE S5. Filter Parameters for Samples from GTEx and TCGA Database. This table presents the filter parameters used for selecting healthy (GTEx) and tumor (TCGA) samples for each cancer type. Six cancer types were investigated for differential gene expression analysis. For GTEx, filter parameters included primary site and primary tissue, while for TCGA, the primary site was used, encompassing all histological types.

TABLE S6. Background Correction for Mutation Cases in Human Lipocalins, OBP2A, and OBP2B in Melanoma and Endometrial Cancer. This table presents the expected number of mutation samples due to random chance for all human lipocalin genes, OBP2A, and OBP2B in melanoma and endometrial cancer. TCGA data for each cancer type were used, and mutation samples were counted based on missense mutations. The average mutation counts per sample were calculated.

FIGURE S14. Lewis structures of two hormones used in the docking experiments to estimate the fit of the ligands inside the binding cavity of hOBP2A.

TABLE S7. Results of the docking binding poses obtained for estradiol inside the hOBP2A binding pocket of hOBP2A (PDB‐ID: 4RUN).

FIGURE S15. Docking results of estradiol within hOBP2A (PDB‐ID: 4RUN). Lefthand side: best scoring binding pose (2 different views, ligand highlighted in magenta), righthand side: superposition of all seven binding poses of the ligand inside the binding cavity. Color‐code: beta‐strands, alpha helix, and loops highlighted in grey, salmon, orange, respectively; each binding pose highlighted in a different color.

TABLE S8. Results of the docking binding poses obtained for testosterone inside the hOBP2A binding pocket of hOBP2A (PDB‐ID: 4RUN).

FIGURE S16. Docking results of estradiol within hOBP2A (PDB‐ID: 4RUN). Lefthand side: best scoring binding pose (2 different views, ligand highlighted in magenta), righthand side: superposition of all seven binding poses of the ligand inside the binding cavity. Color code: beta‐strands, alpha helix, and loops highlighted in grey, salmon, orange, respectively; each binding pose highlighted in a different color.