In Situ Generated Novel 1H MRI Reporter for β-Galactosidase Activity Detection and Visualization in Living Tumor Cells

For wide applications of the lacZ gene in cellular/molecular biology, small animal investigations, and clinical assessments, the improvement of noninvasive imaging approaches to precisely assay gene expression has garnered much attention. In this study, we investigate a novel molecular platform in which alizarin 2-O-β-d-galactopyranoside AZ-1 acts as a lacZ gene/β-gal responsive 1H-MRI probe to induce significant 1H-MRI contrast changes in relaxation times T 1 and T 2 in situ as a concerted effect for the discovery of β-gal activity with the exposure of Fe3+. We also demonstrate the capability of this strategy for detecting β-gal activity with lacZ-transfected human MCF7 breast and PC3 prostate cancer cells by reaction-enhanced 1H-MRI T 1 and T 2 relaxation mapping.

β-Galactosidase prompts the hydrolysis of β-D-galactopyranoside by cleavage of its β-anomeric C-O linkage between β-Dgalactopyranose and aglycone; the hydrolysis reactivity of β-Dgalactopyranosides to β-gal is completely dependent upon the aglycone structure. However, the structure activity relationship of the aglycones in β-D-galactopyranosides vs. β-gal is not yet clear (Juers et al., 2012;Duo et al., 2017). Therefore, further exploration is still highly needed to discover effective β-gal substrates for functional molecular imaging probes. We believe that the traditional histopathological methods of assaying β-gal activity might be the fruitful resources for developing novel imaging agents for the assessment of lacZ gene expression. In reviewing the histopathological literature, we noticed that the well-established β-gal substrate alizarin 2-O-β-D-galactopyranoside AZ-1 ( Figure 1) is readily hydrolyzed by β-gal to release aglycone alizarin, which chelates with ferric iron Fe 3+ to form an intense dark violet Fe complex (James et al., 2000). By comparison of the structural characteristics of the Fe 3+ -alizarin complex with Fe 3+based 1 H-MRI contrast agents (Davies et al., 1996;Richardson et al., 1999;Schwert et al., 2002;Schwert et al., 2005;Haas and Franz, 2009;Yu et al., 2012a;Yu et al., 2012b;Gulaka et al., 2013;Li et al., 2013;Yu et al., 2013;Kuznik and Wyskocka, 2016), we speculated that the Fe 3+ -alizarin complex could function as an Fe 3+ -based 1 H-MRI contrast agent. If so, the well-established β-gal substrate AZ-1 could work as a lacZ gene or β-gal 1 H-MRI reporter. Upon delivery and cleavage at lacZ-transfected or β-gal-overexpressed tumor cells with the presence of Fe 3+ , the paramagnetic Fe complex could be spontaneously formed in situ and specifically produced the 1 H MRI contrast effect while localizing and accumulating 1 H-MRI signals at the β-gal activity site. Figure 1 depicts the Fe 3+ -alizarin complex generated in situ for the 1 H-MRI detection of β-gal activity. We now demonstrate the use of exploiting AZ-1 to assess β-gal activity in vitro with lacZ-transfected human MCF7 breast and PC3 prostate cancer cells by 1 H MRI T 1 and T 2 relaxation mapping. FIGURE 1 | The proposed mechanism of an in situ generated stable Fe 3+ -alizarin complex for 1 H MRI detection of β-gal activity.

RESULTS AND DISCUSSION
Verification of the Fe 3+ -Alizarin Complex as an 1 H-MRI Contrast Agent Alizarin is 1,2-dihydroxy-9,10-anthraquinone with a tricyclic aromatic planar structure and chelates with Fe 3+ to form a thermodynamically stable octahedral Fe 3+ -alizarin (1:3) complex at physiological pH conditions with the stability constant logβ 32.21 (Das et al., 1995;Das et al., 2002). To explore the MRI signal-enhancing capability of the Fe 3+ -alizarin complex, the spin-lattice relaxation time T 1 and spin-spin relaxation time T 2 of the Fe 3+ -alizarin complex were measured with a 4.7 T MR scanner using a saturation recovery spin echo sequence and multi-spin echo sequence with varying repetition times (TRs) and echo times (TEs), respectively. The images were acquired using a 3-cm diameter solenoid coil (home-built) with 4 × 4 cut section of a 96-well plate containing the different concentrations of alizarin and ferric ammonium citrate (FAC) mixed solutions in 1:1 (V/V') DMSO/PBS (0.1 M, pH 7.4) at 37°C. Figure 2 displays the significant changes as expected on the T 1 and T 2 maps and relaxation time values of the Fe 3+ -alizarin complex at T 1 254 ± 3, 131 ± 3, and 92 ± 8 ms, and T 2 106 ± 1, 59 ± 1, and 48 ± 1 ms, corresponding to the concentrations of alizarin at 2.5, 6.0, and 9.0 mM, respectively. The comparison with the control FAC of T 1 389 ± 6 ms and T 2 143 ± 1 ms showed that the Fe 3+ -alizarin complex formed in situ resulted in substantial signal enhancement on either T 1 -or T 2 -weighted 1 H-MRI, confirming the Fe 3+ -alizarin complex generated in situ to function as an 1 H-MRI contrast agent. Notably, the significantly different T 1 and T 2 values of the Fe 3+ -alizarin complex suggested the potential to combine T 1 and T 2 data for additional information of imaging evaluation and detection reliability, specifically where there is possibility for misinterpretation in tissue heterogeneity .

β-Gal Reactivity
AZ-1 has been identified as a highly sensitive substrate for the demonstration of β-gal in a range of Gram-negative bacteria under incubation at 37°C in air for 18 h (James et al., 2000). However, none of the existing data have shown the kinetics of AZ-1 vs. β-gal, which is crucial for further in vivo imaging applications. The absorption spectra of AZ-1 and AZ-2 solutions in 1:1 (V/V′) DMSO/PBS (0.1 M, pH 7.4) with and without β-gal (E801A) at 20-22°C indicated that upon reactions of AZ-1 and AZ-2 with β-gal, a new absorption around 520 nm, corresponding to the in situ released alizarin mono-/dianions, appeared and increased gradually. Hence, the absorbance measurements at 520 nm following the enzymatic  Frontiers in Chemistry | www.frontiersin.org July 2021 | Volume 9 | Article 709581 5 reaction of AZ-1 and AZ-2 with β-gal (E801A) at 20-22°C in different time points showed that both AZ-1 and AZ-2 are very reactive to β-gal (E801A) with varying hydrolytic rates at ν (AZ-1) 93.3 and ν (AZ-2) 133.3 μM/min/unit, respectively ( Figure 4). Also, the absorption spectra of AZ-1 and AZ-2 by reaction with other enzymes α-galactosidase (Sigma G7163) and β-glucuronidase (Sigma G8295) at 20-22°C; showed that both AZ-1 and AZ-2 remained essentially stable over the period of 60 min, verifying their specificity for reaction to β-gal.

In Vitro 1 H-MRI Detection of lacZ Transfection in Human Tumor Cells
The recombinant vector phCMVlacZ has been successfully created and used to stably transfect human prostate cancer PC3-lacZ cells from PC3-wild-type (WT) cells . Accordingly, human breast cancer MCF7-lacZ cells were stably transfected from MCF7 wild-type (WT) cells: the β-gal activity and quantification in MCF7-lacZ cells were verified on the basis of Western blot, X-gal and S-gal staining, and the β-gal assay (Figure 7). Given AZ-2 showed much better aqueous solubility and reactivity to β-gal, the stabilized molecular tweezer complexation AZ-2/Fe 3+ obstructed its implementation spreading to effective 1 H-MRI assessment of β-gal. So, AZ-1 with a significant signal loss either on T 1 or T 2 upon β-gal hydrolysis was prompted for the further in vitro 1 H-MRI evaluation. As an initial demonstration for in vitro 1 H-MRI detection of β-gal with lacZ-transfected human cancer cells, we first acquired T 2 * maps on pair mixtures of AZ-1 (10.0 mM) with PC3-WT cells (5 × 10 5 ) and PC3-lacZ cells (5 × 10 5 ), respectively, in the presence of FAC (10.0 mM) layered between agarose after incubation 4 h at 37°C under 5% CO 2 /air with 95% humidity. Significant differences confined within the layers were observed between PC3-WT and PC3-lacZ cells at different echo times ( Figure 8A), in which there was essentially no signal loss with PC3-WT cells but a remarkable signal decrease with PC3-lacZ cells upon increasing echo times (TEs) ( Figure 8B). The relaxation time T 2 * was determined to be T 2 * (AZ-1/PC3/FAC) 96 ± 23 ms in PC3-WT cells, while T 2 * (AZ-1/PC3-lacZ/FAC) 26 ± 14 ms in PC3-lacZ cells. Again, the β-gal activity was verified based on X-gal, S-gal, and AZ-1 staining (dark violet) ( Figure 8C), with each staining method consistently showing intense lacZ expression in PC3-lacZ cells with essentially no β-gal activity in PC3-WT cells.
Currently, a Gd-based contrast agent-enhanced 1 H-MRI has been widely applied for medical diagnosis, offering a noninvasive way to generate anatomical and physiological information while maintaining high spatial and temporal resolution (Terreno et al., 2010;Haris et al., 2015;Wahsner et al., 2019). An Fe-based 1 H MRI contrast agent, different from the Gd 3+ -based 1H MRI contrast agent with very strong relaxivity, exhibited much shorter relaxation times because of the formation of Fe complexes with the complete coordination of Fe 3+ , eliminating the possibility of inner-sphere to directly coordinate water, leaving outer-sphere and second-sphere coordination water molecules as the only pathways for relaxation (Davies et al., 1996;Richardson et al., 1999;Schwert et al., 2002;Schwert et al., 2005;Haas and Franz, 2009;Kuznik and Wyskocka, 2016). However, an Fe-based contrast agent enhanced 1 H-MRI has now become a viable alternative because Fe 3+ is extensively present in the tissues of the human body and is involved in transport, storage, compartmentalization, and excretion mechanisms, while Gd 3+ is not naturally present in human biochemistry (Beutler, 2004;Weber et al., 2006;Kaplan and Kaplan, 2009;Theil and Goss, 2009). Particularly, cancer cells need a significant amount of Fe 3+ for rapid replication, so endogenously abundant Fe 3+ in tumors has been recognized as a molecular target for chemotherapeutic treatments through depleting cancer cellular Fe 3+ to disrupt cancer cell proliferation and inhibit tumor growth (Fe 3+ -chelation therapy) (Buss et al., 2003;Richardson, 2005). In this study, we introduced exogenous Fe 3+ with the ultimate goal of developing this approach to hunt the elevated Fe 3+ level in tumors for the 1 H-MRI signal generation. Indeed, alizarin has a very high thermodynamic stability constant logβ 32.21 (Das et al., 1995;Das et al., 2002), indicating its capability of capturing Fe 3+ from tumor to produce the Fe 3+ -alizarin complex in situ while simultaneously generating the 1 H-MRI signal enhancement (Richardson et al., 1999;Davies et al., 1996;Schwert et al., 2002;Schwert et al., 2005;Haas and Franz, 2009;Kuznik and Wyskocka, 2016). Moreover, alizarin has been known to inhibit human cytomegalovirus replication, HIV-1 RTassociated RDDP, and integrase activities (Esposito et al., 2011). Furthermore, alizarin is the core part of anthraquinones, which constitute numerous antitumor drugs widely applied in the treatment of various neoplasms such as Adriamycin and daunorubicin, and their coordination with Fe 3+ was shown to diminish cardiotoxicity while improving the antitumor activity in chemotherapy and maintain sound radiosensitizing properties in radiotherapy (Lown, 1993;Nowak and Tarasiuk, 2012;Malik and Müller, 2016). Therefore, this novel molecular platform also indicates the potential for cancer therapy and imaging by utilizing the β-gal responsive turn on pathway to selectively deplete tumor Fe 3+ , resulting in cancer cell cycle arrest and apoptosis while generating 1 H-MRI contrast enhancement, thereby providing insight into the lacZ gene expression, development, location, and magnitude.

CONCLUSION
In this study, we present a novel responsive molecular platform for β-gal activity detection using 1 H-MRI, in which the 1 H-MRI signal enhancement is specifically generated, localized, and accumulated in situ at the β-gal activity site. In conjunction with this design, we have successfully produced and characterized alizarin 2-O-β-D-galactopyranoside AZ-1 and alizarin 1,2-di-O-β-D-galactopyranoside AZ-2. We have also demonstrated the feasibility of using AZ-1 by spontaneous in situ formation of paramagnetic Fe 3+ -alizarin complex to assess the β-gal activity in solution with Fe 3+ ions existence by 1 H-MRI T1 and T2/T2* relaxation mapping. 1 H-MRI clearly showed the significant differences in both T 1 and T 2 at WT vs. lacZ gene expressing cells in culture after incubation with AZ-1, signifying the potential of integrating T 1 and T 2 data together to gain the additional certainty in imaging evaluation and detection reliability of β-gal activity.

EXPERIMENTAL General Methods
NMR spectra were recorded on a Varian Unity INOVA 400 spectrometer (400 MHz for 1 H, 100 MHz for 13 C). 1 H and 13 C chemical shifts are referenced to TMS as an internal standard with CDCl 3 , or DMSO-d 6 as solvents, and chemical shifts are given in ppm. All compounds were characterized by NMR at 25°C. Mass spectra were obtained by positive and negative ESI-MS using a Micromass Q-TOF hybrid quadrupole/time-of-flight instrument (Micromass UK Ltd.). Absorption spectra were taken on a UV-2700 UV-Vis Shimadzu spectrophotometer.
Solutions in organic solvents were dried with anhydrous sodium sulfate and concentrated in vacuo below 45°C. 2, 3, 4, 6-Tetra-O-acetyl-α-D-galactopyranosyl bromide was purchased from the Sigma Chemical Company. β-Gal (E801A) was purchased from the Promega (Madison, WI, United States), and enzymatic reactions were performed at 37°C in the PBS solution (0.1 M, pH 7.4). Column chromatography was performed on silica gel (200-300 mesh), and silica gel GF 254 used for analytical TLC was purchased from the Aldrich Chemical Company. The detection was affected by spraying the plates with 5% ethanolic H 2 SO 4 (followed by heating at 110°C for 10 min) or by direct UV illumination of the plate. The purity of the final products was determined by HPLC with ≥95%.

MRI
MRI studies were performed using a 4.7T horizontal bore magnet or a 9.4T vertical bore magnet equipped with a Varian INOVA Unity system (Palo Alto, CA, United States). T 1 and T 2 (or T 2 *) maps were acquired using a spin echo (or gradient echo) sequence with varying repetition times (TRs) or echo times (TEs), respectively. The raw data were acquired using a centric k-space reordering scheme, followed by the phase encoding steps with higher phase encoding gradient amplitudes. Data acquisition parameters of the FLASH readout were TR/TE/Flip angle 10 ms/5 ms/10°. The standard multi-echo Carr-Purcell-Meiboom-Gill pulse sequence was used for measuring T 2 from a single echo train. The T 2 and T 2 * maps were obtained on a voxel-by-voxel basis using a nonlinear leastsquares fit equation M M 0 e −TE/T2 from the images taken at each echo time. Images were reconstructed and analyzed by using MatLab (MathWorks, Natick, MA).

lacZ Transfection in Human Tumor Cells
The E. coli lacZ gene (from pSV-β-gal vector, Promega, Madison, WI, United States) was inserted into a high expression human cytomegalovirus (CMV) immediate early enhancer/promoter vector phCMV (Gene Therapy Systems, San Diego, CA, United States), producing a recombinant vector phCMV/lacZ. This was used to transfect wild-type MCF7 (human breast cancer) and PC3 (human prostate cancer) cells (ATCC, Manassas, VA, United States) using GenePORTER2 (Gene Therapy Systems, Genlantis, Inc., San Diego, CA, United States). The highest β-gal expressing colony was selected using the antibiotic G418 disulfate (800 μg/ml, Research Products International Corp, Mt Prospect, IL, United States), and G418 (200 μg/ml) was also included for routine culture. The cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Mediatech Inc., Herndon, VA, United States) containing 10% fetal bovine serum (FBS, 0.1 M, pH 7.4, Atlanta Biologicals, Inc., Lawrenceville, GA, United States) with 100 units/mL of penicillin and 100 units/mL streptomycin, and cultured in a humidified 5% CO 2 incubator at 37°C. The β-gal activity of lacZ-transfected tumor cells was measured using a β-Gal Assay Kit with o-nitrophenyl-β-D-galactopyranoside (Promega, Madison, WI, United States) and confirmed by X-gal or S-gal staining. Cells were fixed in PBS plus 0.5% glutaraldehyde (5 min) and rinsed in PBS prior to staining. Staining was performed using standard procedures for 2 h at 37°C in PBS plus 1 mg/ml X-gal (Sigma, St. Louis, MO, United States), 1 mM MgCl 2 , 5 mM K 3 Fe(CN) 6 , and 5 mM K 4 Fe(CN) 6 or with 1.5 mg/mL S-gal (Sigma, St. Louis, MO, United States) and 2.5 mg/ml FAC.

Western Blot
The protein extracted from the wild-type and lacZ-expressing MCF7 and PC3 cancer cells was quantified using the Bradford method by a protein assay (Bio-Rad, Hercules, CA, United States). Protein (30 μg) was added to each well, separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and transferred to a polyvinylidene fluoride membrane. A primary monoclonal anti-β-gal antibody (Promega, Madison, WI, United States) and anti-actin antibody (Sigma, St. Louis, MO, United States) were used as probes at a dilution of 1:5,000, with the reacting protein detected using a horseradish peroxidase-conjugated secondary antibody and ECL detection (Amersham, Piscataway, NJ, United States).

Cytotoxicity
The cytotoxicity for the free β-D-galactopyranoside AZ-1 and the released aglycone alizarine was assessed in both wild-type and lacZ-expressing MCF7 and PC3 cells using a colorimetric CellTiter 96 Aq ueous Non-Radioactive Cell Proliferation Assay (MTS) (Promega, Madison, WI, United States). Assays were performed in triplicate using 24-well plates seeded with 10 3 cells per well in 500 μL of RPMI-1640 without phenol red and supplemented with 10% FCS and 2 mM glutamine (Urano et al., 2005;Kamiya et al., 2007).

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material.

AUTHOR CONTRIBUTIONS
J-XY: conceived and designed the study, analyzed the data, and wrote the manuscript. SG, LZ, and ZF: synthesized, purified the compounds, and performed most part of NMR experiments. VK: conducted 1 H-MRI experiments with PC3-lacZ cells and demonstrated the feasibility for detection of the lacZ gene expression. LL: conducted lacZ transfection in tumor cells and validated the β-gal activity. HW: assisted in toxicity evaluation. HX: helped with structural characterization. MT, BH, CC, and ZZ: assisted in processing data.