An advantage of using an electrochemical method to detect E2 is that is its ability to directly sense the electron transfer generated during the oxidation of the hydroxyl group on the E2 molecular structure to the carbonyl group. Hence, this system does not require the design of a bioreceptor. To improve signal quality, a large number of nanomaterials are employed to increase electron transfer efficiency. For example, Nunes Da Silva et al. (2021) designed an electrochemical sensor to detect E2 in milk. This biosensor was based on an MCPE-MMIP (Magneto Carbon Paste Electrode, MCPE; Magnetic Molecularly Imprinted Polymer, MMIP), facilitating pre-concentration, separation, and manipulation of the analyte. This biosensor exhibited sensitivity to E2 across a range from 0.06 to 175 μM with an LOD of 0.02 or 0.06 μM. In another work, Supchocksoonthorn et al. (2021) selected a nanocomposite synthesized by polyaniline (PANI) and carbon dots (CDs) to improve the sensing performance of glassy carbon electrode (GCE) on detecting E2 (Figure 1A). The surface chemical properties of CDs are stable, and there are various functional groups for E2 to attach. Conductive polymer PANI increased the conductivity and sensitivity of this electrode. Cyclic voltammetry (CV) and linear sweep voltammetry (LSV) were used to record the electrochemical detection of E2 utilizing CDs-PANI/GCE across the range of 0.001–100 M with an LOD of 43 nM. More examples of application of nanomaterials in E2 electrochemical sensors are listed in Table 1.

In addition to the flexible utilization of nanoparticles, electrodes can also be modified structurally to increase the performance of electrochemical biosensing. Nanomaterials have a higher surface-to-volume ratio. Moreover, their three-dimensional networks assembled by one-dimensional or two-dimensional nanostructures can present porous structures (Li et al., 2011). This property maintains a larger area of the active surface, enabling the electrode to attach more biomolecules to a confined active region. For example, Xie et al. (2022) established GCE for the assay of E2 based on the change of the surface nanostructure (Figure 1B). The team used wrinkled mesoporous carbon (wMC) nanomaterials to provide abundant active electron and proton transfer sites. They selected wrinkled silicon nanoparticle (wSiO₂NPs) as supporting and sacrificial hard templates to form wMC nanomaterials with core-shell nanocage structures, thereby increasing the surface area, electron diffusion, and the number of attachment points for E2. More examples of nanostructure application in E2 electrochemical biosensors are listed in Table 1.

The electrochemical sensor based on non-bioreceptor mostly constructs a primary electrochemical detection platform in E2 detection without incorporating specific bioreceptors. Additionally, the electrode can be modified with specific nanomaterials based on diverse detection requirements. However, because of the absence of biometric components, the electrochemical sensor based on non-biometric molecules performs poor selectivity. If substances with redox molecular structure similar to E2 or substances with redox peak that overlap with E2 are present, the biosensor may be falsely identified as E2, thereby affecting the detection results.

E2 biosensors that utilize enzymes as bioreceptors are often combined with electrochemistry. The performance of such biosensors depends on the properties and architecture of the employed enzymes and the surface characteristics and structure of the electrodes (Spychalska et al., 2020). In addition, the way to immobilize the enzyme affects the sensing performance. The immobilized enzyme not only needs to maintain a tight binding to the electrode surface but also ensure its biological activity. There are two commonly used methods for immobilizing enzymes: directly onto the surface of the transducer or by first immobilizing them with a carrier and then attaching them to the transducer (Nguyen et al., 2019). The main methods mentioned above include adsorption, entrapment, covalent crosslinking, chemical crosslinking, and embedding.

Adsorption is one of the simplest fixation methods, presenting fewer destructive effects on enzyme structure than other methods due to the non-necessity of an additional reagent. However, this method is most susceptible to interference during detection, because it is based on weak bonds such as Van der Waals forces, electrostatic, and hydrophobic interactions (Sassolas et al., 2012). Additionally, some physiological conditions like high temperature or pH, substrate addition may weaken the bond (Sassolas et al., 2012). Thus, it is rarely employed in E2 detection.

Entrapment requires selecting a suitable polymer to provide the enzyme-free space for dispersion, ensuring high stability, and minimizing leaching (Nguyen et al., 2019). For example, Jijana, (2023) selected Horseradish Peroxidase (HRP) as a bioreceptor to achieve the detection of E2. In this work the nanocomposite 3-mercaptopropionic acid capped zinc selenide quantum dots trapped within the polyaniline (PANI: 3MPA-ZnSeQD) are used to immobilize the HRP on the gold electrode surface by entrapment, as shown in Figure 2A. This biosensor was able to accurately and efficiently measure concentrations of E2 from 0.2 to 4 μM and display an LOD of value 0.2 μM.

Covalent crosslinking presents a powerful method to offer stable complexes between enzymes and supports. Spychalska et al. (2020) used the same enzyme, HRP, as a bioreceptor. The platinum (Pt) electrode was modified by an electroconductive polymer (CP) poly (4,7-bis(5-(3,4-ethylene-dioxythiophene) thiophenes-2-yl) benzothiadiazole), which created an appropriate microenvironment to immobilize the protein and act as a transducer during the transfer of electric charge, as shown in Figure 2B. Then, HRP was immobilized on the surface of the modified electrode by physical adsorption and covalent cross-linking with glutaraldehyde. This biosensor detected the E2 concentration in the medication with a detection range of 0.1–200 M and an LOD of 105 nM.

Chemical crosslinking prevents the active site of an enzyme from being buried in its complex protein structure (Povedano et al., 2017). Wang et al. (2019) modified the GCE with laccase loading with citric acid-graphene (CA-GR) and electropolymerized L-lysine film (Lac/PLLY/CA-GR/GCE), as shown in Figure 2C. Under the best experimental conditions, this biosensor presented a reliable linear detection and LOD. In this system, graphene was treated with citric acid, which solve the problem of the high hydrophobicity and lack of functional group of GR (Liu Y. et al., 2012). Glutaraldehyde was used as a cross-linker to connect laccase and nanocomposite material, generating cross-linking reaction between laccase and nanocomposite material.

Collectively, electrochemical enzyme sensors offer a range of electrode materials, structures, and enzymes to choose from. Various nanomaterials can be incorporated into electrodes to enhance sensing performance. Methods of immobilizing enzymes on electrode surfaces are varied and easy to operate, which are convenient for designing and applying electrochemical enzyme sensors. Furthermore, electrochemical biosensors based on enzymes have the potential to realize the measurement of hormones in vivo, which is one of the directions worth developing in the future. However, enzymes are often buried in complex protein structures or modified electrode materials, thus affecting the detection performance. Therefore, it is necessary to select appropriate enzymes and immobilization methods to avoid this situation. Moreover, the electrode surface is structurally modified or modified of nanomaterials with biologically active molecules (Rodriguez-Abetxuko et al., 2020), enhancing the electrochemical reactivity of biomolecules and promoting electron transfer, signal amplification, and enzyme immobilization.

For biosensors, the creation of biosensors in vitro and in vivo has been dramatically promoted due to the high affinity and specificity of immunoglobulin (Ig) or antibody to its target analyte (Crivianu-Gaita and Thompson, 2015). How the immune sensor performs primarily depends on three key factors: (1) the selection of the antibody; (2) the method of antibody fixation; (3) the biological tag used to label the antibody for signal recognition.

The sensitivity and stability of immunosensors largely depend on the number of surface-fixed immune molecules, their conformational stability, and their orientation on the surface of the sensor (Makaraviciute and Ramanaviciene, 2013). These characteristics are also determined in part by the type of antibody selected. For instance, immunoglobulins are asymmetric molecules whose recognition sites occupy different spatial positions during different fixation processes, resulting in blocked interaction with the analyte (Makaraviciute and Ramanaviciene, 2013). The method of antibody fixation also plays an important role in immunosensors. There are two main approaches that can be used in antibody-based sensor surface preparation: random and site-directed antibody immobilization. Random immobilization mainly includes physical adsorption, covalent immobilization. These methods excel in easy immobilization procedure, but frequently suffer from random orientation of antibody, resulting in high steric-hindrance and poor reproducibility (Kausaite-Minkstimiene et al., 2010). Therefore, building a well oriented layer of antibodies is necessary. Site-directed antibody immobilization technique perform well in this respect, avoiding randomness that occurs when antibody is fixed. The most commonly used technique is to modify coatings on the sensing surface, such as self-assembled monolayers (Zhou, 2006), dextran or various polymers (Kyprianou et al., 2013), to immobilize biomolecules on the sensing surface. To convert the binding events into physical signals the biological tags, such as metal nanoparticles, conductive polymer, fluorescent molecule, are employed. Moreover, these biological tags also can enhance the performance of immunosensors.

According to the design points mentioned above, immunosensors often use a double-biomolecular sandwich mechanism, competition mechanism, and unlabeled electrochemistry to measure E2, which were described in this part.

An aptamer is a single-stranded deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) fragment obtained through several rounds of screening in the oligonucleotide library by Systematic Evolution of Ligands by Exponential Enrichment (SELEX) (Zheng et al., 2019). Aptamers can bind to target analyte by folding into two-dimensional or three-dimensional structures, including stem ring, hairpin, pseudoknot, bulge loop, and G-quadruplex (Cao et al., 2018; Shaban and Kim, 2021). Additionally, various functional groups are available to modify the aptamer or to immobilize the aptamer to the transducer surface, including -OH, amine, thiol, biotin (Waifalkar et al., 2022). These various functional groups also provide numerous surface immobilization strategies, such as covalent bonding, electrostatic interaction, avidin–biotin interaction, and the self-assembled monolayer (Radi and Abd-Ellatief, 2021). Generally, different transducer surface materials choose different fixation methods, and the details would be expounded upon in this section. Therefore, aptamer biosensors have higher selectivity, affinity, and stability advantages than ordinary antibody and enzyme biosensors.

The aptamer structure can be modified flexibly to increase the sensing performance (Dong et al., 2014). For example, Qiao et al. (2021) truncated the redundant bases from the E2 aptamer (original E2 aptamer 76-mer (5′-GCT​TCC​AGC​TTA​TTG​AAT​TAC​ACG​CAG​AGG​GTA​GCG​GCT​CTG​CGC​ATT​CAA​TTG​CTG​CGC​GCT​GAA​GCG​CGG​AAG​C - 3′)), as shown in Figure 4A, avoiding the uncertainty in aptamer tertiary structure and instability of affinity for targets in different solutions. This operation effectively increased the compatibility between the aptamer and E2 as well as the performance of the sensor. In Pakawat Kongpreecha team’s work (Kongpreecha et al., 2023), the E2 aptamer was required to wrap gold nanoparticles. Therefore, the original E2 aptamer (original E2 aptamer 22-mer (5′-GCC​GTT​TGG​GCC​CAA​GTT​CGG​C-3′)) structure was removed, and two to five loops were repeatedly added to determine which structure was most effective for E2 recognition. Finally, a specific structure was selected, as shown in Figures 4B, C. In addition, split aptamers are used to strengthen the binding of biometric elements to E2. The aptamer fragment is re-associated in the presence of E2 to restore the desired whole and valid part (Röthlisberger and Hollenstein, 2018). This detection mechanism provides a valuable tool for identifying critical molecular interactions, which avoids effectively generating false positive or non-specific signals (Röthlisberger and Hollenstein, 2018). For example, Rozi et al. (2022) deposited polypyrrole nanowires on carbon screen printing electrode (SPE). They then fixed polymer microspheres on the electrode. Finally, they activated carboxylic acid groups at the end of the microspheres to immobilize E2 split apt12 and apt14, as shown in Figure 4C.

Additionally, aptamers are modified by various tag molecules, including fluorophores, luminescent groups, nanoparticles, and other tags. These modifications link the molecular recognition ability of aptamers with the process of signal generation (Cho et al., 2009; Famulok and Mayer, 2011), thus generating a variety of detectable signals. Then, these detectable signals are easily measured by colorimetry, fluorescence, electrochemistry, and photoelectric chemistry. The above categories were described in this section.

Steroestrogens (17α-acetylenestradiol, 17β-estradiol, and estrone) have been included in the EU Water “watch list” of the EU Water Framework Directive (WFD) (Kase et al., 2018). Even if the content of E2 is not high in the sewage discharged from livestock farms, sewage treatment plants, and food processing plants, it will endanger human health through the accumulation of the food chain. Numerous studies have demonstrated the widespread presence of E2 in samples from sludge, sediment, rivers, and lakes (Wang et al., 2011). Therefore, timely detecting of water resources, food, and other substances that are directly ingested by the human body ensures the E2 content at a safe level. Currently, biosensors based on enzymes, antibodies, and aptamers have been successfully developed to detect E2 in water resources and food. Some methods are summarized in Table 3.

It is challenging to detect trace hormones in the environment. Single-use analysis methods with suboptimal LOD may sometimes fail to meet the detection requirements. Optimizing the signal processing circuit, biometric identification, and detection devices can enhance the signal quality. It can be known from the second part of this paper that selecting appropriate bioreceptors and modifying appropriate functional materials effectively increase the sensing performance of the E2 biosensor. Furthermore, various innovative detection devices have been designed to optimize signal quality. The work of Li et al. (2019) was ingenious. They proposed a differential graphene field-effect transistor (GFET) sensor to selectively detect E2 in the presence of nonspecific interference. Certain biosensors are susceptible to nonspecific interference from factors such as pH, ionic strength, temperature, and other parameters. Therefore, Yijun et al. designed two sensing units to address this issue. The measuring unit was sensitive to analytes and interference, and the reference unit was only sensitive to interference. This design minimized the impact of changes in environmental conditions on measurement accuracy, achieving an LOD of 34.70 pM. These results were consistent with the results of non-interference detection. De Cezaro et al. (2020) first developed a cantilever nanobiosensor modified with tyrosinase to detect E2, with a sensitivity of 0.101 V/ug and an LOD of 0.4 ng/L. This nanobiosensor significantly increased the ability to recognize nanoscale events.

As a natural estrogen, E2 plays an important role in EDCs, exhibiting the strongest estrogenic effect when introduced into the human body exogenously, despite its essentiality to human physiology. When consuming food or medicine with E2 residues or contamination, human beings are exposed to exogenous E2 unconsciously (Pu et al., 2019a). Therefore, it is of great importance to develop rapid, sensitive and selective biosensors for detecting E2 residues in food matrices. Some methods are summarized in Table 3.

Due to the majority of E2 exposure from foods come from animal-derived food, most studies focused on dairy and meat products (Hartmann et al., 1998; Malekinejad et al., 2006; Courant et al., 2007; Regal et al., 2012). Some researches found that positive associations were observed between dairy product consumption and total and free estradiol concentrations (P for trend = 0.02 and 0.03, respectively) (Brinkman et al., 2010). Therefore, although the concentration of E2 in commercial dairy products is usually lower than 100 ng/L, the low dose effect and biological amplification effect of E2 cannot be ignored (Pape-Zambito et al., 2010). The traditional method used to measure E2 in dairy products, such as milk, are hollow fiber-based stirring extraction bar liquid-liquid microextraction (HF-SEBLLME) (Xu et al., 2013), HPLC with magnetic solid phase extraction method (Wang et al., 2015), and so on. Although these methods can accurately detect the content of E2, they need more sample pretreatment and detection times before detection. Xiao et al. (2017) fabricated an activated paper with molecularly imprinted polymers (MIPs) to detect E2. The whole detection process could be finished in 10 min with a much lower cost. More importantly, the result could be directly observed by naked eye. To satisfy E2 concentrations below ng/L level, Jiang et al. (2018) selected aptamer which perform well in thermal stability as bioreceptor, and applied Poly (N-isopropylacrylamide) on the surface of the transducer as a microgel layer to complete the detection of E2 in milk. The detection limit was 0.9 pg/mL (3.2 pM).

Meat products are another food product containing estrogens such as estrone, estradiol, and estriol. Illegal use of hormonal drugs in animal feed to promote growth rate is common, resulting in residual hormones in meat. Therefore, exogenous estrogen may accumulate in the human body through dietary estrogen intake and affect human health. Because meat products are not liquids, the use of portable biosensors also requires a series of pretreatment procedures. At present, HPLC and GC-MS techniques are commonly used to detect the E2 content in meat products.

E2 is also found in many medicines. For example, the concept of using natural E2 in combined oral contraceptives emerged in the 1970s (Greenblatt et al., 1977; Wenzl et al., 1993; Hoffmann et al., 1998). Although the introduction of E2 may reduce cardiovascular risk in women using oral contraceptives, it may also lead to other physical health conditions. In addition, it is also necessary to accurately detect the E2 level during clinical drug intervention. Hence how accurately control the E2 level during drug preparation is important. However, the current detection of E2 in pharmaceutical reagents still largely relies on large-scale equipment like HPLC. There are currently limited reports on the direct utilization of biosensors.

E2 is a primary female steroid hormone primarily produced in the ovaries, and in small amounts in the liver, heart, muscles, bones, and brain (Cui et al., 2013). E2 primarily regulates physiological functions in various tissues through the estrogen receptor (ER). The growth, development, and differentiation of various cancer cells are accompanied by abnormal E2 content (McDonnell and Norris, 2002; Burns and Stabile, 2014; Che et al., 2014; Huang et al., 2018). E2 protects the cardiovascular system and inhibits the aging of Human Vein Endothelial Cells (HUVECs) (Fang et al., 2018; Xiang et al., 2023). E2 is also believed to play neurotrophic and neuroprotective roles in the brain, including regulating reproductive behavior, spinal density in male and female forebrains (Lu et al., 2019), synaptic plasticity (Brandt et al., 2020), neuroprotection, and hippocampus-dependent cognitive function with memory (Lu et al., 2019).

In conclusion, E2 is extremely important for human health. At present, high sensitivity ELISA kit (Lu et al., 2019) and gas chromatography-mass spectrometry (GC-MS/MS) (Ankarberg-Lindgren et al., 2018) have been used to detect E2 content in human body fluids. However, they require complex preprocessing and highly skilled operations. Therefore, efficient, sensitive, and miniature biosensors are needed to detect E2 content in the human body. The reference range of E2 concentration in urine and serum is essential for clinical evaluation, and part of the detection methods are shown in Table 4. The instruments listed in the table achieved miniaturization, high efficiency, and sensitive detection of E2. Nevertheless, they all need to conduct specific pretreatment of samples, and integration of sampling and detection fails to be formed. The literature reviewed so far is all about in vitro detection of E2. Continuous and dynamic in vivo examination in clinical application is able to help to monitor the condition and provide better treatment. As previously mentioned in this context, E2 can regulate the plasticity of neural synapses and the formation of memory as a newly discovered neuromodulator in the brain (Lu et al., 2019), thus helping to treat some neurological diseases. However, current research methods are to affect the expression of E2 in the neural circuit by gene knockout or drug intervention (Spencer-Segal et al., 2012; Lu et al., 2019), or to directly remove the gonads to study the effect of E2 on the electrophysiological properties of neurons (Tuscher et al., 2016; Zhao et al., 2018). None of these methods can accurately reflect the concentration of E2 that can produce positive effects on neurological diseases. Therefore, it is also necessary to develop some new biosensors that can detect the changes of E2 concentration in specific brain regions in real time.