Development of Highly Efficient Estrogen Receptor β-Targeted Near-Infrared Fluorescence Probes Triggered by Endogenous Hydrogen Peroxide for Diagnostic Imaging of Prostate Cancer

Hydrogen peroxide is one of the most important reactive oxygen species, which plays a vital role in many physiological and pathological processes. A dramatic increase in H2O2 levels is a prominent feature of cancer. Therefore, rapid and sensitive detection of H2O2 in vivo is quite conducive to an early cancer diagnosis. On the other hand, the therapeutic potential of estrogen receptor beta (ERβ) has been implicated in many diseases including prostate cancer, and this target has attracted intensive attention recently. In this work, we report the development of the first H2O2-triggered ERβ-targeted near-infrared fluorescence (NIR) probe and its application in imaging of prostate cancer both in vitro and in vivo. The probe showed good ERβ selective binding affinity, excellent H2O2 responsiveness and near infrared imaging potential. Moreover, in vivo and ex vivo imaging studies indicated that the probe could selectively bind to DU-145 prostate cancer cells and rapidly visualizes H2O2 in DU-145 xenograft tumors. Mechanistic studies such as high-resolution mass spectrometry (HRMS) and density functional theory (DFT) calculations indicated that the borate ester group is vital for the H2O2 response turn-on fluorescence of the probe. Therefore, this probe might be a promising imaging tool for monitoring the H2O2 levels and early diagnosis studies in prostate cancer research.


Introduction
In recent years, the therapeutic potential of estrogen receptor beta (ERβ) in breast cancer, prostate cancer, lung cancer, the nervous system, and bone tissue has been revealed, and this target has attracted more and more attention [1][2][3][4]. However, the fundamental research of ERβ is insufficient. The sub cellular distribution, subtype expression and the role of ERβ in different diseases need to be further confirmed [5,6]. Moreover, prostate cancer (PCa) is the second most common cancer and the fifth most common cause of cancer-related deaths in men worldwide [7,8]. Prostate cancer is closely related to age, and some signaling pathways involving reactive oxygen species (ROS) play an important role in the occurrence and progression of cancer with age [9]. At the same time, ERβ is an important target for the treatment of prostate cancer, and its expression level is different in normal prostate tissue, prostatic hyperplasia tissue, benign prostate cancer tissue, and highgrade prostate cancer tissue [10,11]. Therefore, the ERβ-targeted probe can be exploited to monitor the lesions and carcinogenesis of the prostate, so as to promote the development of the early diagnosis of prostate cancer. In order to enhance the targeting of probes to prostate cancer, environment-responsive ERβ-targeted probes can also be designed in virtue of tumor-specific micro environmental information. specifically fractured by H 2 O 2 and leaves, and is subsequently triggered and shows strong priming fluorescence and achieves tumor imaging in vitro and in vivo. Our study may bring new opportunities for prostate cancer diagnosis and research, which will certainly have greater practical application value ( Figure 1).
Molecules 2023, 28, x FOR PEER REVIEW 3 of 15 sponsive group, thus trying to develop H2O2-triggered ERβ probes for prostate cancer imaging. We believe that the introduction of the borate ester group destroys the push-pull effect of the DCM-OH fluorophore and that the probe does not emit fluorescence. However, in the environment of a high concentration of H2O2, the borate ester group of the probe is specifically fractured by H2O2 and leaves, and is subsequently triggered and shows strong priming fluorescence and achieves tumor imaging in vitro and in vivo. Our study may bring new opportunities for prostate cancer diagnosis and research, which will certainly have greater practical application value ( Figure 1).

Design and Synthesis of Probes
The design and synthesis of probes was shown in Scheme 1; compound one was condensed with two in the presence of piperidine and acetic acid to obtain three. Next, the intermediate three was demethylated using boron tribromide to obtain probe P1. The probe P2, which contains an unsaturated alkene bond, was formed by reaction of P1 with 2-chloroethane sulfonyl chloride. The intermediates and target compounds were confirmed with 1 H NMR, 13 C NMR, and high-resolution mass spectrometry. These data and detailed synthesis procedures for intermediates one and two (Scheme S1) are outlined in Supplementary Materials.

Design and Synthesis of Probes
The design and synthesis of probes was shown in Scheme 1; compound one was condensed with two in the presence of piperidine and acetic acid to obtain three. Next, the intermediate three was demethylated using boron tribromide to obtain probe P1. The probe P2, which contains an unsaturated alkene bond, was formed by reaction of P1 with 2-chloroethane sulfonyl chloride. The intermediates and target compounds were confirmed with 1 H NMR, 13 C NMR, and high-resolution mass spectrometry. These data and detailed synthesis procedures for intermediates one and two (Scheme S1) are outlined in Supplementary Materials. sponsive group, thus trying to develop H2O2-triggered ERβ probes for prostate cancer imaging. We believe that the introduction of the borate ester group destroys the push-pull effect of the DCM-OH fluorophore and that the probe does not emit fluorescence. However, in the environment of a high concentration of H2O2, the borate ester group of the probe is specifically fractured by H2O2 and leaves, and is subsequently triggered and shows strong priming fluorescence and achieves tumor imaging in vitro and in vivo. Our study may bring new opportunities for prostate cancer diagnosis and research, which will certainly have greater practical application value ( Figure 1).

Design and Synthesis of Probes
The design and synthesis of probes was shown in Scheme 1; compound one was condensed with two in the presence of piperidine and acetic acid to obtain three. Next, the intermediate three was demethylated using boron tribromide to obtain probe P1. The probe P2, which contains an unsaturated alkene bond, was formed by reaction of P1 with 2-chloroethane sulfonyl chloride. The intermediates and target compounds were confirmed with 1 H NMR, 13 C NMR, and high-resolution mass spectrometry. These data and detailed synthesis procedures for intermediates one and two (Scheme S1) are outlined in Supplementary Materials.

Optical Properties of Probes
In order to explore whether the target probe has animal imaging potential and verify the quenching effect of the borate ester group on fluorescence, we used PBS as the solvent to obtain the optical information of the probe ( Table 1). The experimental results showed an obvious absorption peak of P1 at 410 nm, and an NIR emission at 653 nm was observed. The emission wavelength has reached the near-infrared region, which has a good potential for animal imaging. The absorption peak of P2 is also at 413 nm, and the NIR emission at 655 nm, which also meets the requirements of in vivo imaging. At the same time, it's interesting that the Stokes shifts of P1-P2 are more than 240 nm, and the anti-background interference ability is very strong. We tested the fluorescence quantum yield of the probe, and the results showed that the fluorescence quantum yield of P1 and P2 were low, because the introduction of the borate ester group effectively quenched the fluorescence of the DCM-OH parent nucleus. The possible reason may be that the DCM-OH fluorophore produces fluorescence mainly through the push and pull interaction between the cyano-group with strong electron absorption and the hydroxyl group with strong electron donation, while the introduction of the borate ester group destroys the push and pull interaction of the fluorescence parent nucleus, so the fluorescence quantum yield decreases. To see whether the products of P1 and P2 after oxidation via H 2 O 2 have better optical properties, we treated probes P1 and P2 with H 2 O 2 ( Figure 2). It showed that the absorption spectrum and fluorescence emission spectrum of the probe solution were greatly different before and after the addition of H 2 O 2 . After treatment with H 2 O 2 , the absorption spectra of P1 and P2 were redshifted, indicating that the properties of the probe changed before and after the H 2 O 2 response (Figure 2a,d). From the fluorescence spectroscopy, the fluorescence intensity of P1 and P2 was relatively low and did not emit fluorescence in the 600-700 nm range before H 2 O 2 was added. However, after treatment with H 2 O 2 , P1 and P2 produced fluorescence emission peaks in the 600-700 nm range significantly, indicating that the oxidation product has a good potential for in vivo imaging (Figure 2b,e). Moreover, in order to show the effect of H 2 O 2 on P1 and P2 more intuitively, we also detected the relationship between the fluorescence emission spectrum of the probe and the different concentrations of H 2 O 2 (Figure 2c,f). Interestingly, the results exhibited that after P1 was oxidized by H 2 O 2 , the fluorescence intensity of the solution showed a linear relationship with the concentration in the range of 0-100 µM (R 2 = 0.9609). Meanwhile, the fluorescence intensity of the P2 solution also displayed a good linear relation with concentration in the range of 0-100 µM (R 2 = 0.9897), which indicated that the H 2 O 2 responsive probe had a high sensitivity to the concentration. In addition, we found that after the oxidation of probe P1 with H 2 O 2 , the product was the same as probe P5 in our previous studies [36]. Comparing the fluorescence intensity of P1 and P5 with the same concentration as shown in Figure S1, the fluorescence intensity of P1 is very weak before adding H 2 O 2 , which is lower than that of probe P5. It is worth noting that after adding H 2 O 2 , the fluorescence intensity of P1 is greatly enhanced, even higher than that of P5. This is consistent with the characteristics of the "turn-on" probes, which may have more imaging advantages than "inherent" probes.

Interference Resistance of Probe
There are various substances in the cell, such as anions, cations, amino acids, and reducing substances, which may all interfere with the experimental results. Therefore, P2 was used to explore the anti-interference of the probe, and the H 2 O 2 -treated group was used as the positive control group (Figure 3a). After adding positive ions (Na + , Mg + , K + ), common anions (OH − , CO 3 2− , SO 4 2− , NO 2 − , Cl − , ClO − ), and amino acids (Tyr, GSH, His, Cys, Glu, Gly, Arg) to the probe for 60 min, no fluorescence increase was detected. Moreover, the fluorescence recovered after the addition of H 2 O 2 , indicating that common anions, cations, and amino acids had no significant interference on the detection of H 2 O 2 , so the probe would not be interfered by other substances when used for imaging in the physiological environment. We detected the effect of pH on probe fluorescence by adding P2 separately to PBS solutions with a pH = 4-10 and then observing the fluorescence differences in the solutions (Figure 3b). The fluorescence intensity of P2 changes little in different pH environments, which may be because the fluorescence of the probe is still in the quenched state, although the borate esters are sensitive to acid-base environments; the product of acid-base hydrolysis is boric acid. It can be seen that probes can be employed for imaging in physiological environments and are not affected under diverse pH conditions.

Interference Resistance of Probe
There are various substances in the cell, such as anions, cations, amino acids, and reducing substances, which may all interfere with the experimental results. Therefore, P2 was used to explore the anti-interference of the probe, and the H2O2-treated group was used as the positive control group (Figure 3a). After adding positive ions (Na + , Mg + , K + ), common anions (OH − , CO3 2− , SO4 2− , NO2 − , Cl − , ClO − ), and amino acids (Tyr, GSH, His, Cys, Glu, Gly, Arg) to the probe for 60 min, no fluorescence increase was detected. Moreover, the fluorescence recovered after the addition of H2O2, indicating that common anions, cations, and amino acids had no significant interference on the detection of H2O2, so the probe would not be interfered by other substances when used for imaging in the physiological environment. We detected the effect of pH on probe fluorescence by adding P2 separately to PBS solutions with a pH = 4-10 and then observing the fluorescence differences in the solutions (Figure 3b). The fluorescence intensity of P2 changes little in different pH environments, which may be because the fluorescence of the probe is still in the quenched state, although the borate esters are sensitive to acid-base environments; the product of acid-base hydrolysis is boric acid. It can be seen that probes can be employed for imaging in physiological environments and are not affected under diverse pH conditions.

Interference Resistance of Probe
There are various substances in the cell, such as anions, cations, amino acids, and reducing substances, which may all interfere with the experimental results. Therefore, P2 was used to explore the anti-interference of the probe, and the H2O2-treated group was used as the positive control group (Figure 3a). After adding positive ions (Na + , Mg + , K + ), common anions (OH − , CO3 2− , SO4 2− , NO2 − , Cl − , ClO − ), and amino acids (Tyr, GSH, His, Cys, Glu, Gly, Arg) to the probe for 60 min, no fluorescence increase was detected. Moreover, the fluorescence recovered after the addition of H2O2, indicating that common anions, cations, and amino acids had no significant interference on the detection of H2O2, so the probe would not be interfered by other substances when used for imaging in the physiological environment. We detected the effect of pH on probe fluorescence by adding P2 separately to PBS solutions with a pH = 4-10 and then observing the fluorescence differences in the solutions ( Figure 3b). The fluorescence intensity of P2 changes little in different pH environments, which may be because the fluorescence of the probe is still in the quenched state, although the borate esters are sensitive to acid-base environments; the product of acid-base hydrolysis is boric acid. It can be seen that probes can be employed for imaging in physiological environments and are not affected under diverse pH conditions.  Groups were set as follows: 1, blank; 2, NaCl (100 mM); 3, MgCl 2 (100 mM); 4, KCl (2.5 mM); 5, Tyr (1 mM); 6, GSH (1 mM); 7, His (1 mM); 8, Cys (1 mM); 9, Glu (1 mM); 10, Gly (1 mM); 11, Arg (1 mM); 12, NaOH (100 µM); 13, NaCO 3 (100 µM); 14, Na 2 SO 4 (100 µM); 15, NaNO 2 (100 µM); 16, NaClO (100 µM); 17, NaCl (100 µM); 18, H 2 O 2 (100 µM). Incubation time: 10 min.

ERβ Selectivity of Probes
We first tested the Ki values for ERα and ERβ of the probe ( Table 2). The experimental results showed that the two probes displayed the same subtype selectivity. Due to the large borate ester group, they had a very weak affinity for ERα but had a certain affinity for ERβ. Compared with probe P2, probe P1 has a lower binding affinity to ERβ. The ERβ Molecules 2023, 28, 2309 6 of 15 selectivity of P1 is three times higher than ERα. The addition of an unsaturated alkene double bond to the left side of probe P1 increased the binding affinity of P2 for ERβ with the corresponding Ki value of 1.21 µM. Simultaneously, the ERβ selectivity of P2 increased 7.75-fold more than ERα. This result is consistent with our expected assumption, indicating that probe P2 is a potential ERβ-targeting fluorescent probe.

ERβ Selectivity of Probes
We first tested the Ki values for ERα and ERβ of the probe ( Table 2). The experimental results showed that the two probes displayed the same subtype selectivity. Due to the large borate ester group, they had a very weak affinity for ERα but had a certain affinity for ERβ. Compared with probe P2, probe P1 has a lower binding affinity to ERβ. The ERβ selectivity of P1 is three times higher than ERα. The addition of an unsaturated alkene double bond to the left side of probe P1 increased the binding affinity of P2 for ERβ with the corresponding Ki value of 1.21 μM. Simultaneously, the ERβ selectivity of P2 increased 7.75-fold more than ERα. This result is consistent with our expected assumption, indicating that probe P2 is a potential ERβ-targeting fluorescent probe.

Imaging of Probe in Living Cells
To investigate the toxicity of P1 and P2 on normal cells and their inhibitory activity on cancer cells before performing cellular imaging, we tested the effects of P1 and P2 on MCF-10A normal breast cells, MCF-7 breast cancer cells with high ERα expression, and DU-145 prostate cancer cells with high ERβ expression by using CCK-8 assays (Table S1). The results showed that probe P2 had no cytotoxicity to the MCF-10A normal cell line and DU-145 prostate cancer cell line, but it showed a certain inhibitory effect on the MCF-7 cancer cell line with an IC50 value of 16.34 ± 4.42 μM. However, P1 was not toxic to any of these three cell lines. This indicated that our probe does not damage normal cells and tissues when used for imaging and can truly and accurately reflect the expression level of ERβ in the lesion area.
In live cell imaging, we first explored their selectivity to ER isoforms and responsiveness to H2O2 with PMA as the H2O2 inducer and DPI as the H2O2 inhibitor ( Figure 4). When the probe, probe + PMA, and probe + DPI were added to live cells 15 min later, weak fluorescence was observed in MCF-10A, MCF-7, and DU-145 cells without PMA, and more obvious fluorescence was observed in these two cells after added PMA, which indicated that P1 and P2 had a strong response to H2O2. In the H2O2 environment, the borate ester group was specifically removed, thus releasing fluorescence. At the same time, the inhibition of H2O2 production by DPI completely eliminated the fluorescence signal in MCF-10A, MCF-7, and DU-145 cells, and the cells shrunk and deformed, indicating that the consumption of H2O2 not only led to a decrease in the fluorescence intensity of the probe, but also had a great damage effect on cancer cells. This has reference significance for the development of bifunctional probes in the future. In addition, P1 and P2 are both 4.36 ± 0.01 1.42 ± 0.02 3 P2 mM); 12, NaOH (100 μM); 13, NaCO3 (100 μM); 14, Na2SO4 (100 μM); 15, NaNO2 (100 μM); 16, NaClO (100 μM); 17, NaCl (100 μM); 18, H2O2 (100 μM). Incubation time: 10 min.

ERβ Selectivity of Probes
We first tested the Ki values for ERα and ERβ of the probe ( Table 2). The experimental results showed that the two probes displayed the same subtype selectivity. Due to the large borate ester group, they had a very weak affinity for ERα but had a certain affinity for ERβ. Compared with probe P2, probe P1 has a lower binding affinity to ERβ. The ERβ selectivity of P1 is three times higher than ERα. The addition of an unsaturated alkene double bond to the left side of probe P1 increased the binding affinity of P2 for ERβ with the corresponding Ki value of 1.21 μM. Simultaneously, the ERβ selectivity of P2 increased 7.75-fold more than ERα. This result is consistent with our expected assumption, indicating that probe P2 is a potential ERβ-targeting fluorescent probe.

Imaging of Probe in Living Cells
To investigate the toxicity of P1 and P2 on normal cells and their inhibitory activity on cancer cells before performing cellular imaging, we tested the effects of P1 and P2 on MCF-10A normal breast cells, MCF-7 breast cancer cells with high ERα expression, and DU-145 prostate cancer cells with high ERβ expression by using CCK-8 assays (Table S1). The results showed that probe P2 had no cytotoxicity to the MCF-10A normal cell line and DU-145 prostate cancer cell line, but it showed a certain inhibitory effect on the MCF-7 cancer cell line with an IC50 value of 16.34 ± 4.42 μM. However, P1 was not toxic to any of these three cell lines. This indicated that our probe does not damage normal cells and tissues when used for imaging and can truly and accurately reflect the expression level of ERβ in the lesion area.
In live cell imaging, we first explored their selectivity to ER isoforms and responsiveness to H2O2 with PMA as the H2O2 inducer and DPI as the H2O2 inhibitor ( Figure 4). When the probe, probe + PMA, and probe + DPI were added to live cells 15 min later, weak fluorescence was observed in MCF-10A, MCF-7, and DU-145 cells without PMA, and more obvious fluorescence was observed in these two cells after added PMA, which indicated that P1 and P2 had a strong response to H2O2. In the H2O2 environment, the borate ester group was specifically removed, thus releasing fluorescence. At the same time, the inhibition of H2O2 production by DPI completely eliminated the fluorescence signal in MCF-10A, MCF-7, and DU-145 cells, and the cells shrunk and deformed, indicating that the consumption of H2O2 not only led to a decrease in the fluorescence intensity of the probe, but also had a great damage effect on cancer cells. This has reference significance for the development of bifunctional probes in the future. In addition, P1 and P2 are both 8

Imaging of Probe in Living Cells
To investigate the toxicity of P1 and P2 on normal cells and their inhibitory activity on cancer cells before performing cellular imaging, we tested the effects of P1 and P2 on MCF-10A normal breast cells, MCF-7 breast cancer cells with high ERα expression, and DU-145 prostate cancer cells with high ERβ expression by using CCK-8 assays (Table S1). The results showed that probe P2 had no cytotoxicity to the MCF-10A normal cell line and DU-145 prostate cancer cell line, but it showed a certain inhibitory effect on the MCF-7 cancer cell line with an IC 50 value of 16.34 ± 4.42 µM. However, P1 was not toxic to any of these three cell lines. This indicated that our probe does not damage normal cells and tissues when used for imaging and can truly and accurately reflect the expression level of ERβ in the lesion area.
In live cell imaging, we first explored their selectivity to ER isoforms and responsiveness to H 2 O 2 with PMA as the H 2 O 2 inducer and DPI as the H 2 O 2 inhibitor (Figure 4). When the probe, probe + PMA, and probe + DPI were added to live cells 15 min later, weak fluorescence was observed in MCF-10A, MCF-7, and DU-145 cells without PMA, and more obvious fluorescence was observed in these two cells after added PMA, which indicated that P1 and P2 had a strong response to H 2 O 2 . In the H 2 O 2 environment, the borate ester group was specifically removed, thus releasing fluorescence. At the same time, the inhibition of H 2 O 2 production by DPI completely eliminated the fluorescence signal in MCF-10A, MCF-7, and DU-145 cells, and the cells shrunk and deformed, indicating that the consumption of H 2 O 2 not only led to a decrease in the fluorescence intensity of the probe, but also had a great damage effect on cancer cells. This has reference significance for the development of bifunctional probes in the future. In addition, P1 and P2 are both targeted to ERβ; this targeting was more pronounced in cell imaging. The fluorescence signals of P1 and P2 were stronger in DU-145 cells with high ERβ expression, especially for probe P2. But they showed a weaker image in MCF-7 cells. In MCF-10A normal mammary epithelial cells, the fluorescence signal is particularly weak and almost invisible. Therefore, P1 and P2 have good selectivity towards ERβ and excellent H 2 O 2 responsiveness, so it is expectable to obtain high ERβ subtype-targeted H 2 O 2 -responsive probes through structural modification. targeted to ERβ; this targeting was more pronounced in cell imaging. The fluorescence signals of P1 and P2 were stronger in DU-145 cells with high ERβ expression, especially for probe P2. But they showed a weaker image in MCF-7 cells. In MCF-10A normal mammary epithelial cells, the fluorescence signal is particularly weak and almost invisible. Therefore, P1 and P2 have good selectivity towards ERβ and excellent H2O2 responsiveness, so it is expectable to obtain high ERβ subtype-targeted H2O2-responsive probes through structural modification. Subsequently, we investigated the colocalization of the probes and the nuclear dye DAPI in DU-145 cells ( Figure 5A). The positions of P1, P2, and DAPI did not coincide in DU-145 cells which a high expression of ERβ; instead, they distributed in the extracellular region with a small overlap with the nuclear dye DAPI. We used Mito-Tracker Green, a mitochondrial dye, to study the distribution of probes in the cytoplasm ( Figure 5B). Imaging results showed that the two probes were highly colocalized with mitochondria, and the co-location analysis diagrams were shown in Figure S2, which indicated that they were targeted to mitochondria. The ERβ labeling ability of the probes was further confirmed via immunofluorescence staining ( Figure 5C). P1 and P2 were partially colocalized with ERβ, and the fluorescence signal mainly existed in the outer circle of the cell, which was similar to the colocalization result of mitochondria. Comparing the mitochondrial colocalization and immunofluorescence imaging results, it was proved that probes were targeting the mitochondrial ERβ (mtERβ), which has been identified by studies before [38,39]. Although it is reported that the mtERβ might be associated with apoptosis [40], the function of mtERβ in tumorigenesis remains unclear. At this juncture, probes P1 and P2 with a mtERβ targeting ability can be used as a new tool for the early diagnosis of prostate cancer. Subsequently, we investigated the colocalization of the probes and the nuclear dye DAPI in DU-145 cells ( Figure 5A). The positions of P1, P2, and DAPI did not coincide in DU-145 cells which a high expression of ERβ; instead, they distributed in the extracellular region with a small overlap with the nuclear dye DAPI. We used Mito-Tracker Green, a mitochondrial dye, to study the distribution of probes in the cytoplasm ( Figure 5B). Imaging results showed that the two probes were highly colocalized with mitochondria, and the co-location analysis diagrams were shown in Figure S2, which indicated that they were targeted to mitochondria. The ERβ labeling ability of the probes was further confirmed via immunofluorescence staining ( Figure 5C). P1 and P2 were partially colocalized with ERβ, and the fluorescence signal mainly existed in the outer circle of the cell, which was similar to the colocalization result of mitochondria. Comparing the mitochondrial colocalization and immunofluorescence imaging results, it was proved that probes were targeting the mitochondrial ERβ (mtERβ), which has been identified by studies before [38,39]. Although it is reported that the mtERβ might be associated with apoptosis [40], the function of mtERβ in tumorigenesis remains unclear. At this juncture, probes P1 and P2 with a mtERβ targeting ability can be used as a new tool for the early diagnosis of prostate cancer.

In Vivo Imaging
As shown above, the good imaging ability of P2-targeting prostate cancer cells was more pronounced in vitro (Figure 4). Compared with P1, the fluorescence signal of P2 was stronger in DU-145 prostate cancer cells with high ERβ expression; however, P2 showed a weaker image in MCF-7 cells. Moreover, in normal MCF-10A mammary epithelial cells, the fluorescence signal of P2 is particularly weak and almost invisible. In addition, according to colocalized imaging results in DU-145 prostate cancer cells, the ERβ labeling ability of the probe P2 was further confirmed via immunofluorescence staining ( Figure 5C). Considering the good imaging ability of probe P2 to target prostate cancer cells in vitro, we further performed fluorescence imaging in vivo and then tracked the imaging process to determine whether it could be accurately visualized in a DU-145 xenograft mice model. After the injection of probe P2 into the tail vein, representative fluorescence images and corresponding fluorescence intensity ratios of normal tissues to tumors at different time points (0, 3, 6, 15, and 36 h) are shown in Figure 6A. The probe did not show obvious fluorescence signal in nude mice at the beginning. Due to P2 having the ability to target ERβ protein highly expressed in DU-145 xenograft tumors, the probe will gradually enrich in the tumor site. The fluorescence signal in the tumor area is stronger than the background in 15 h, showing good tumor targeting. The mild fluorescence signals were also observed in the liver region. It seems that there is a fluorescence signal in the mice brain. Chen's group also found this signal in a related study, which might be because ERβ is also expressed in brain tissue [41]. Subsequently, we performed ex vivo animal imaging of mice to further determine the exact tissue distribution of probe P2. The fluorescence signal of tumor tissues was significantly stronger than that of the heart, liver, spleen, lung, kidney, and other tissues, but weaker fluorescence was also observed in the liver and kidney ( Figure 6B,C), indicating that the probe may be metabolized by these two organs. The fluorescence signal of the kidney tissue could be observed to be significantly enhanced at 18 h, the fluorescence signal of tumor tissue and various organ tissues was basically weak at 24 h, and the probe was completely metabolized in mice at 36 h ( Figure S3). P2 exhibited excellent ERβ selectivity and good tumor targeting ability in vivo, which has important implication for imaging studies of ERβ prostate cancer.

In Vivo Imaging
As shown above, the good imaging ability of P2-targeting prostate cancer cells was more pronounced in vitro ( Figure 4). Compared with P1, the fluorescence signal of P2 was stronger in DU-145 prostate cancer cells with high ERβ expression; however, P2 showed a weaker image in MCF-7 cells. Moreover, in normal MCF-10A mammary epithelial cells, the fluorescence signal of P2 is particularly weak and almost invisible. In addition, according to colocalized imaging results in DU-145 prostate cancer cells, the ERβ labeling ability of the probe P2 was further confirmed via immunofluorescence staining ( Figure 5C). Considering the good imaging ability of probe P2 to target prostate cancer cells in vitro, we further performed fluorescence imaging in vivo and then tracked the imaging process to determine whether it could be accurately visualized in a DU-145 xenograft mice model. After the injection of probe P2 into the tail vein, representative fluorescence images and corresponding fluorescence intensity ratios of normal tissues to tumors at different time points (0, 3, 6, 15, and 36 h) are shown in Figure 6A. The probe did not show obvious fluorescence signal in nude mice at the beginning. Due to P2 having the ability to target ERβ protein highly expressed in DU-145 xenograft tumors, the probe will gradually enrich in the tumor site. The fluorescence signal in the tumor area is stronger than the background in 15 h, showing good tumor targeting. The mild fluorescence signals were also observed in the liver region. It seems that there is a fluorescence signal in the mice brain. Chen's group also found this signal in a related study, which might be because ERβ is also expressed in brain tissue [41]. Subsequently, we performed ex vivo animal imaging of mice to further determine the exact tissue distribution of probe P2. The fluorescence signal of tumor tissues was significantly stronger than that of the heart, liver, spleen, lung, kidney, and other tissues, but weaker fluorescence was also observed in the liver and kidney

The Mechanism of Response towards H2O2
We hypothesized that since the introduction of borate ester groups destroys the push-pull effect of the DCM-OH fluorophore, the probe does not emit fluorescence, resulting in fluorescence quenching. However, when it entered DU-145 cells with a high

The Mechanism of Response towards H 2 O 2
We hypothesized that since the introduction of borate ester groups destroys the pushpull effect of the DCM-OH fluorophore, the probe does not emit fluorescence, resulting in fluorescence quenching. However, when it entered DU-145 cells with a high concentration of reactive oxygen species, the borate ester group is specifically broken by H 2 O 2 and leaves to become a phenolic hydroxyl group (electron-donating group), resulting in a strong fluorescence enhancement. Therefore, in order to reveal the mechanism of its fluorescence response, we analyzed the cell culture medium of P2 using high-resolution mass spectrum (HRMS). The mass fragments observed at 529.1588 and 441.0522 belong to P2 ([M + H] + , 529.1599) and P2 + H 2 O 2 ([M + Na] + , 441.0516), respectively ( Figure S3), which well validated the proposed mechanism of action, the probe (Scheme 2). fluorescence imaging of the tumor and major organs at 15 h after the injection of the probe. (C) Regions of interest (ROI) analysis of signal-to-background ratios.

The Mechanism of Response towards H2O2
We hypothesized that since the introduction of borate ester groups destroys the push-pull effect of the DCM-OH fluorophore, the probe does not emit fluorescence, resulting in fluorescence quenching. However, when it entered DU-145 cells with a high concentration of reactive oxygen species, the borate ester group is specifically broken by H2O2 and leaves to become a phenolic hydroxyl group (electron-donating group), resulting in a strong fluorescence enhancement. Therefore, in order to reveal the mechanism of its fluorescence response, we analyzed the cell culture medium of P2 using high-resolution mass spectrum (HRMS). The mass fragments observed at 529.1588 and 441.0522 belong to P2 ([M + H] + , 529.1599) and P2 + H2O2 ([M + Na] + , 441.0516), respectively ( Figure  S3), which well validated the proposed mechanism of action, the probe (Scheme 2). Scheme 2. Proposal mechanism of the probe P2 response to H2O2.
Finally, a DFT calculation was performed to analyze its turn-on mechanism from a theoretical perspective (Figure 7). According to the results of the lowest unoccupied molecular orbital (LUMO) energy and the highest occupied molecular orbital (HOMO) energy of probe P2 and its H2O2 oxidating product, the electrons on the fluorophore are transferred to the borate ester group via a photoinduced electron transfer (PET) upon excitation, leading to fluorescence quenching. However, when the borate ester group is ox-Scheme 2. Proposal mechanism of the probe P2 response to H 2 O 2 .
Finally, a DFT calculation was performed to analyze its turn-on mechanism from a theoretical perspective (Figure 7). According to the results of the lowest unoccupied molecular orbital (LUMO) energy and the highest occupied molecular orbital (HOMO) energy of probe P2 and its H 2 O 2 oxidating product, the electrons on the fluorophore are transferred to the borate ester group via a photoinduced electron transfer (PET) upon excitation, leading to fluorescence quenching. However, when the borate ester group is oxidized and broken into the hydroxyl group, the probe formed a D-π-A skeleton and released fluorescence, thus blocking the PET effect. In addition, the phenolic hydroxyl group can increase the intramolecular charge transfer (ICT) effect of the fluorophore and improve the fluorescence performance, which is consistent with the electron cloud change between the LUMO and HOMO of the oxidation product. In summary, the PET effect of borate ester group is the key to fluorescence quenching; the probe forms a D-π-A skeleton via oxidation to hydroxyl, realizing the fluorescence turn-on phenomenon, blocking the PET effect, and maintaining the ICT effect of fluorophore.
can increase the intramolecular charge transfer (ICT) effect of the fluorophore and improve the fluorescence performance, which is consistent with the electron cloud change between the LUMO and HOMO of the oxidation product. In summary, the PET effect of borate ester group is the key to fluorescence quenching; the probe forms a D-π-A skeleton via oxidation to hydroxyl, realizing the fluorescence turn-on phenomenon, blocking the PET effect, and maintaining the ICT effect of fluorophore.

Materials and Measurements
The chemicals used in the experiments were purchased commercially. 1 H NMR and 13 C NMR spectra were obtained on a Bruker Biospin AV400 (400 MHz) (Billerica, MA, USA) instrument with TMS as the internal standard, and the compound charge/mass ratio was obtained using IonSpec 4.7 Tesla FTMS mass spectrometer (Irvine, CA, USA). The UV spectrum information of the probe was obtained via the SHIMADZU UV-2600 (Kyoto, Japan), and the fluorescence spectrum information of the probe was obtained via the HI-TACHI F-4600 (Hitachi, Japan). Cell imaging was performed using a LECA-LCS-SP8 laser confocal microscope (Wetzlar, Germany).

Chemical Synthesis
The synthetic route for the probes is based on the modification method described in references [42,43] Under an argon atmosphere, compound 1 (500.0 mg, 2.1 mmol) and 2 (580 mg, 2.5 mmol) were added to the acetonitrile solution, then piperidine and acetic acid (piperidine: acetic acid = 1 mL: 0.5 mL) were added in turn and the reaction was heated to 85 °C and refluxed for 12 h. Then, it was extracted with dichloromethane (50 mL × 3), washed with saturated sodium chloride solution (20 mL × 1), and dried with anhydrous sodium sulfate; the organic phase was concentrated, and the crude product was purified using column chromatography (petroleum ether/ethyl acetate = 9:1) to obtain yellow solid 3 (300 mg).

Materials and Measurements
The chemicals used in the experiments were purchased commercially. 1 H NMR and 13 C NMR spectra were obtained on a Bruker Biospin AV400 (400 MHz) (Billerica, MA, USA) instrument with TMS as the internal standard, and the compound charge/mass ratio was obtained using IonSpec 4.7 Tesla FTMS mass spectrometer (Irvine, CA, USA). The UV spectrum information of the probe was obtained via the SHIMADZU UV-2600 (Kyoto, Japan), and the fluorescence spectrum information of the probe was obtained via the HITACHI F-4600 (Hitachi, Japan). Cell imaging was performed using a LECA-LCS-SP8 laser confocal microscope (Wetzlar, Germany).

Chemical Synthesis
The synthetic route for the probes is based on the modification method described in references [42,43] Under an argon atmosphere, compound 1 (500.0 mg, 2.1 mmol) and 2 (580 mg, 2.5 mmol) were added to the acetonitrile solution, then piperidine and acetic acid (piperidine: acetic acid = 1 mL: 0.5 mL) were added in turn and the reaction was heated to 85 • C and refluxed for 12 h. Then, it was extracted with dichloromethane (50 mL × 3), washed with saturated sodium chloride solution (20 mL × 1), and dried with anhydrous sodium sulfate; the organic phase was concentrated, and the crude product was purified using column chromatography (petroleum ether/ethyl acetate = 9:1) to obtain yellow solid 3 (300 mg). Yield: 32%. 1

Optical Properties
Fluorescence performance test preparation with 10 mM in PBS (pH = 7.4) probe solution, using an ultraviolet visible light spectrophotometer and HITACHI SHIMADZU UV-2600F-4600 fluorescence spectrophotometer instrument measuring the optical properties of the probe. Fluorescein (Φ f1 = 0.85) was used as a control, and the fluorescence quantum yields of P1-P2 were calculated using the following formula: The meaning of the abbreviations: Φ: Fluorescence quantum yield; A (standard) and A (sample): absorption values of controls and probes at λ em ; and S (standard) and S (sample): emission peak-to-peak areas of controls and probes. Slit width = 10/10 nm.

ER Binding Affinity Assay
The binding affinities of the probes to ERα and ERβ were determined via fluorescence polarimetry (FPA). In a 384-well plate, 20 µL of potassium phosphate buffer consisting of 0.8 µM ERα or ERβ protein, 150 nM fluorescent ligand, and 2.4 µg bovine immunoglobulin was added, followed by 20 µL of the target compound solution and incubated for 2 h at 25 • C in the dark. Fluorescence polarization values were obtained with a citation 3 microplate reader (Biotek, Winooski, VT, USA) and the experimental results were analyzed and the K i value of each compound was calculated.

CCK-8 Assay
Normal breast cells MCF-10A and prostate cancer cells DU-145 (purchased from ATCC, Rockville, MD, USA), breast cancer cells MCF-7 (purchased from the Type Culture Collection of the Chinese Academy of Sciences, Shanghai, China) were cultured in a phenol red DMEM liquid medium supplemented with 10% fetal bovine serum. When the cell density reached 80% to 90%, the cells were digested, and the cell suspension was spread into 96-well cell culture plates with a phenol red-free DMEM medium containing 10% CS. After the cells were completely adherent, the original culture medium was discarded, and 100 µL of fresh compound solution prepared with the DMEM medium containing 10% CS were added to each well. The concentration gradient of the compound is: 1 × 10 −7.5 M, 1 × 10 −7 M, 1 × 10 −6.5 M, 1 × 10 −6 M, 1 × 10 −5.5 M, 1 × 10 −5 M, 1 × 10 −4.5 M, 1 × 10 −4 M. After 4 days of drug-treated culture, the culture plate was removed, the culture medium was aspirated, 100 µL CCK-8 working solution was added to each well and incubated for 1.5-2 h at 37 • C in a 5% CO 2 incubator. The plate was read on the microplate reader, the wavelength at 450 nm was selected as the reference wavelength, the experimental results were analyzed, and the IC 50 was calculated.

Cell Imaging
The DMEM medium containing the above cells was placed in a cell incubator at 37 • C, and after resuscitation, the cells were transferred to confocal small dishes and cultured for 24 h. The probes (10 µM), probe + PMA (10 µM), and probe + DPI (10 µM) were added to the small dishes to stain the viable cells for 15 min, and then the cells were carefully washed with PBS buffer 3 times. In the colocalization experiment, the cells in the confocal dishes were stained with probes for 15 min, then washed with PBS buffer 3 times, fixed with 4% paraformaldehyde, permeated with 0.2% Triton X-100, and washed with PBS after 10 min. DAPI was added to stain nuclei or the Mito-tracker Green was added to stain mitochondria. After 30 min, the free dye was washed off with PBS to observe the imaging results. In immunofluorescence staining, the DU-145 cells were stained with a probe (10 µM) for 30 min, and then fixed and permeated through the membrane. The cells were washed three times with PBS and incubated with a monoclonal anti-ERβ antibody (1:200) for 12 h at 37 • C. The cells were washed with PBS, and the secondary antibody dye DyLight 488 AffiniPure Goat Anti-Rabbit IgG (1:200) was added for ERβ staining for 1 h, after washing off the free dye to observe the image. Imaging results were obtained with a Leica-lcs-sp8 confocal laser scanning microscope.

Animal Imaging
In the animal imaging study, the mouse DU-145 tumor transplantation model was established. Firstly, the prostate cancer cell DU-145 was expanded and cultured, then injected subcutaneously into the waist of 6-week-old male Balb/c nude mice (purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China). When the tumor had grown to approximately 80-100 mm 3 in size, the probe was dissolved in PBS and injected into the mice through the tail vein (n = 3). After 30 min, the mice were anesthetized with 2% sodium pentobarbital and imaged with a living animal imager (Bruker Xtreme BI, Karlsruhe, Germany), and the imaging results were observed at 3-h intervals. Excitation: 510 nm, emission 700 nm.

DFT Calculation
All DFT theoretical calculations have been carried out using the Gaussian 09 D0.1 program package. The B3LYP density functional method with the D3(BJ) dispersion correction was employed in this work to carry out all the computations [44]. The 6-31G (d,p) basis set was used for the atoms in geometry optimizations. Vibrational frequency analyses at the same level of theory were performed on all optimized structures to characterize stationary points as local minima [45].

Conclusions
In conclusion, the first ERβ-targeting and H 2 O 2 -triggered turn-on fluorescent probes were successfully developed in this study. As expected, probes P1 and P2 possessed favorable ERβ binding affinity and high ERβ selectivity. In addition, they exhibited good optical properties and excellent imaging ability, including a quick turn-on fluorescence response to H 2 O 2 and mitochondrial ERβ imaging ability. Most importantly, P2 was able to selectively bind to prostate cancer DU-145 cells and displayed no toxicity toward normal cells. In the in vivo imaging studies, probe P2 could rapidly and precisely identify tumor tissues in DU-145 xenograft tumors. We believe that these novel ERβ-targeted H 2 O 2 -triggered fluorescent probes will be useful for early prostate cancer diagnosis and therapy. These results implied that our study might be useful for investigating the role of mitochondrial ERβ in cancer, cardiovascular diseases, and neurological diseases, etc.