In Vivo Near-infrared Optical Imaging Of Esophageal Squamous Cell Carcinoma Based On Cathepsin B Activity In Nude Mice

BackgroundIn China, esophageal cancer ranks fourth in terms of incidence and mortality among all cancers. There are two major histological types of esophageal carcinoma: esophageal squamous cell carcinoma (ESCC) and adenocarcinoma with considerably varied epidemiological features. In China and other East Asian countries, more than90%of cases are ESCC, whereas adenocarcinoma is more common in the United States and European countries. The prognosis for patients with ESCC remains poor, which mainly result from late diagnosis in advanced stages to most patients. As a result, early diagnosis and early treatment is the key to improve the prognosis of patients with ESCC. Early diagnosis mainly depends on health screenings. Census methods include endoscopy and esophageal balloon cytology, which sensitivity and specificity are not ideal, or even cause esophageal injury. Therefore, highly sensitive, highly specific, and non-invasive diagnostic methods are urgently needed to improve the diagnostic capabilities of ESCC in clinic.Tumor molecular imaging is based on metabolism, cellular and molecular level changes in early tumors, rather than the later morphological changes in process of disease development which generally shows in common imaging methods such as CT, MRI, and so on. Specific imaging of target molecules may be earlier and more specific to detect early tumors, which has additional advantages compared with conventional imaging means and methods. This brings new hope for early diagnosis of ESCC. Cathepsin B (CB) is highly expressed in several cancers, closely related to occurrence and development of tumors, and in line with the general selection criteria of "target" in molecular imaging. Some scholars have chosen CB as a molecular target for tumor diagnostic studies and achieved encouraging results. In this study, we will evaluate and discuss the feasibility of using CB as a novel imaging target for the detection of human ESCC by near-infrared optical imaging in nude mice, which will provide experimental data and theoretical guidance for the follow-up clinical studies.Objective1. By examining CB expression in different esophageal tissue (including normal esophagus, esophageal intraepithelial neoplasia lesions and ESCC) and ESCC tumor cells, we preliminarily judged whether CB could act as an alternative target for diagnostic imaging of ESCC.2. In vivo detection of tumor bearing mice by optical imaging combined with a novel CB activated near-infrared fluorescent (NIRF) probe was performed, observing whether the CB probe could allow tumor localization and imaging of ESCC xenografts. We will analyze and discuss the feasibility of using CB as a novel imaging target for in vivo detection of human ESCC.Methods1. We examined CB expression in80tissue specimens from normal human esophagus (n=20), intraepithelial neoplasia (IN) lesions (n=20), tumors in situ (Tis)(n=20) and invasive cancer (IC)(n=20) by immunohistochemistry (IHC).2. Eca-109tumor cells (human ESCC cell line) and HET-1A cells (normal human esophageal epithelial cell line) were cultured and detected for CB expression by Western Blot.3. To verify whether Eca-109tumor cells could secret CB into extracellular space, cell culture fluid was retrieved and reacted with different concentrations of CB probe (12.5nM,25nM,50nM,100nM and200nM). The fluorescent signals generated by different concentrations of CB probe were then compared and analyzed between experimental group and control group. As a negative control group, the same amount of fresh serum-free culture fluid was also reacted with identical concentrations of CB probe.4. In order to understand if the novel CB activatable NIRF probe can easily enter Eca-109cells and detect CB activity presented in cytoplasm, Eca-109cells were pretreated with the CB probe. Then we observed CB expression and the fluorescence signals generated by the CB probe using immunocytochemistry (ICC).5. A human-mouse ESCC xenograft tumor model was established by subcutaneous implantation of Eca-109cells. When the diameter of xenografts reached about6mm, in vivo imaging of tumor bearing mice (n=6) were performed every2h,36h post the CB probe injection by a living imaging system. Image acquisition parameters were as follows:emission filter Cy5.5, excitation filter675run, emission spectra680to720nm, exposure time0.5s, f/stop2, binning medium, field of view12.5cm2. The acquired images were analyzed using living image software. Fluorescence intensity, defined as total radiant efficiency [p/s]/[μW/cm2], was quantified using identical size regions of interest (ROI). Normal nude mice (n=3) acted as control group.6. After in vivo imaging, tumor xenografts and main visceral organs from BALB/c nude mice (n=6) were resected out in darkness, rinsed with normal saline, and performed ex vivo imaging together. Image acquisition parameters were same as in vivo imaging. Due to there are differences in size and shape of xenografts and visceral organs, fluorescence intensity, defined as average radiant efficiency [p/s/cm2/sr]/[μW/cm2], was quantified using various size ROIs which illustrated the outlines of different organs. Normal visceral organs (n=3) acted as control group.7. Making frozen sections, the sources of fluorescence signals in tumor xenografts imaged in vivo and ex vivo were identified by NIRF histology.8. CB expression in main visceral organs of BALB/C nude mice including ESCC xenografts was examined by IHC.Results1. IHC results showed that CB was absent in normal esophageal tissue; in contrast, it was found to be highly upregulated in INs, Tis and IC. There was a higher level of CB expression with the progress of ESCC. The percentages of CB positive expression in IN lesions, Tis and IC were95%(19/20),95%(19/20) and100%(20/20),respectively. CB was mainly overexpressed in the cytoplasm of ESCC cells. The macrophages and inflammation cells in tumor stroma occasionally expressed CB. 2. Western blot analysis revealed a high level of CB expression in Eca-109tumor cells, but normal esophageal intraepithelial cells (Het-1A cells) had no CB expression, which was in accordance with IHC results.3. There were no significant higher in fluorescent intensity of the retrieved culture fluid than that in fresh cell culture fluid (control group) at same concentration of CB probe. This indicates Eca-109cells could not secret CB into extracellular space in vitro.4. High CB expression was also observed in cytoplasm of Eca-109tumor cells by ICC. In response to CB activity, the recovered NIRF signals of the CB probe were limited in cytoplasm of Eca-109cells and completely coincided with CB distribution. This indicates that the CB probe molecules have a certain permeability, can enter into Eca-109tumor cells, and generate fluorescence signals after CB degradation ex vivo.5. The results of in vivo imaging of tumor bearing mice (n=6):after injecting CB probe, the whole mice body initially emitted weak homogeneous autofluorescence (1.34×105±3.3×104), and there were no significant differences in mean fluorescence intensity (FI) between tumor sites and other parts of the body. Two hours later, we observed a strong specific fluorescence signal localized at the tumor sites (9.78×109±1.1×109) which gradually increased up to about32hrs (2.90×1010±1.84×109) and then steadily declined in tumor bearing mice compared to normal nude mice. This indicates that the CB probe remains relatively stable in vivo, have appropriate half-life, easily enter into tumor cells, fully react with CB, and recover NIRF signals in cytoplasm in response to CB activity. In addition, superficial thyroid gland and testis showed relatively higher fluorescence signal during imaging time, although compared with the FIs at tumor site, their FIs were significantly lower at each imaging time point (all p<0.05) and also, the signals kept changing to some extent. For the control mice, the left flank areas showed no significant NIRF signals (2.04×106±5.3×105) during whole imaging time; however, their thyroid and testis showed significant signals similar to that tumor bearing mice.6. The results of ex vivo imaging of xenografts and main visceral organs:similar to in vivo imaging, tumor xenografts showed the strongest NIRF signals in ex vivo imaging (5.81×108±4.3×107) compared to the visceral organs (all p<0.05). However, notably, the probe displayed relatively higher fluorescence signals in visceral organs including liver, kidney, thyroid gland, and testis (3.14×108±3.5×107,2.26×108±4.4×107,2.71×108±3.8×107and1.66×108±2.2×107, respectively). Among the tumor xenografts and visceral organs, the fluorescence signals of esophagus were lowest (1.99×107±2.5x106). Furthermore, heart and lung which are near to esophagus also showed lower signals. The tumor to visceral organs signal ratios were as follows: tumor/liver1.85, tumor/kidney2.57, tumor/heart8.4, tumor/lung9.57, tumor/esophagus29.20, tumor/thyroid2.14, tumor/testis5.31, tumor/spleen15.17, and tumor/small intestine9.21, which clearly suggest tumor specificity of the CB probe (ratio of2.5is generally considered to have tumor specificity for molecular probe). The visceral organs of normal nude mice (control group) had no significant fluorescence signals.7. NIRF histology showed that the fluorescence signals in tumor tissue were from cytoplasm of Eca-109tumor cells, which coincided with the location of CB expression. The result demonstrated that Eca-109tumor cells instead of other organs or cells were the sources of fluorescence signals in ESCC xenografts imaged in vivo and ex vivo. Tumor cells at the edge of ESCC xenografts decreased, but the NIRF signals diffusely enhanced, which indicate increase of CB activity at the invasive leading edg of tumor tissue.8. CB positivity was observed in BALB/c nude mouse liver, kidney, thyroid gland, and testis. Moreover, CB expression levels histologically correlated with the fluorescence signals, showing that the CB probe can reflect the number of in-vivo targets. This provides a reasonable explanation for the positive organs such as liver, kidney, thyroid and testes in vivo and ex vivo imaging. No significant CB expression was seen in mouse heart, lung, esophagus, spleen, and intestine.Conclusions1. Major conclusions (1) CB was absent in normal human esophageal tissue, but it was highly upregulated in human ESCC and its precursor lesions. Eca-109tumor cells have a high level of CB expression, but Het-1A cells had no CB expression. Therefore, CB is involved in occurrence and development of ESCC, and may be an alternative target for molecular imaging of ESCC and its precancerous lesions.(2) The elevated CB expression in ESCC allowed in vivo detection of ESCC xenografts in nude mice by optical imaging combined with a novel CB activated NIRF probe. Our results support the usefulness of CB activity as a potential imaging target for the detection of human ESCC.2. Minor conclusions(1) In ESCC tissue, CB was mainly overexpressed in the cytoplasm of ESCC cells, and small amount of CB was overexpressed in macrophages and inflammation cells of tumor stroma. CB activity increased at the invasive leading edg of ESCC xenografts. Eca-109cells could not secret CB into extracellular space in vitro.(2) A novel CB activated NIRF probe, Cat B680FASTTM, has high CB specificity, remains relatively stable in vivo, have appropriate half-life, easily enter into tumor cells in vitro and in vivo, fully react with CB, and recover NIRF signals in cytoplasm in response to CB activity.(3) CB positivity was observed in BALB/c nude mouse liver, kidney, thyroid gland, and testis. No significant CB expression was seen in mouse heart, lung, esophagus, spleen, and intestine.