In Vivo Imaging of Hypoxia and Neoangiogenesis in Experimental Syngeneic Hepatocellular Carcinoma Tumor Model Using Positron Emission Tomography

Introduction Hypoxia-induced ανβ3 integrin and aminopeptidase N (APN/CD13) receptor expression play an important role in tumor neoangiogenesis. APN/CD13-specific 68Ga-NOTA-c(NGR), ανβ3 integrin-specific 68Ga-NODAGA-[c(RGD)]2, and hypoxia-specific 68Ga-DOTA-nitroimidazole enable the in vivo detection of the neoangiogenic process and the hypoxic regions in the tumor mass using positron emission tomography (PET) imaging. The aim of this study was to evaluate whether 68Ga-NOTA-c(NGR) and 68Ga-DOTA-nitroimidazole allow the in vivo noninvasive detection of the temporal changes of APN/CD13 expression and hypoxia in experimental He/De tumors using positron emission tomography. Materials and Methods 5 × 106 hepatocellular carcinoma (He/De) cells were used for the induction of a subcutaneous tumor model in Fischer-344 rats. He/De tumor-bearing animals were anaesthetized, and 90 min after intravenous injection of 10.2 ± 1.1 MBq 68Ga-NOTA-c(NGR) or 68Ga-NODAGA-[c(RGD)]2 (as angiogenesis tracers) or 68Ga-DOTA-nitroimidazole (for hypoxia imaging), whole-body PET/MRI scans were performed. Results Hypoxic regions and angiogenic markers (αvβ3 integrin and APN/CD13) were determined using 68Ga-NOTA-c(NGR), 68Ga-DOTA-nitroimidazole, and 68Ga-NODAGA-[c(RGD)]2 in subcutaneously growing He/De tumors in rats. 68Ga-NOTA-c(NGR) showed the strong APN/CD13 positivity of He/De tumors in vivo, by which observation was confirmed by western blot analysis. By the qualitative analysis of PET images, heterogenous accumulation was found inside He/De tumors using all radiotracers. Significantly (p ≤ 0.01) higher SUVmean and SUVmax values were found in the radiotracer avid regions of the tumors than those of the nonavid areas using hypoxia and angiogenesis-specific radiopharmaceuticals. Furthermore, a strong correlation was found between the presence of angiogenic markers, the appearance of hypoxic regions, and the tumor volume using noninvasive in vivo PET imaging. Conclusion 68Ga-DOTA-nitroimidazole and 68Ga-NOTA-c(NGR) are suitable diagnostic radiotracers for the detection of the temporal changes of hypoxic areas and neoangiogenic molecule (CD13) expression, which vary during tumor growth in a hepatocellular carcinoma model.


Introduction
Nowadays, in clinical and experimental oncology, the tumor angiogenesis and hypoxia are one of the most intensively researched areas. At present, little is known about the temporal variation in the expression of angiogenic markers in tumors, and even less in the context of hypoxia. The ability to visualize the formation of new vessels and hypoxic areas in solid tumors using in vivo molecular imaging methods allows the noninvasive monitoring of antiangiogenic treatments and the planning of radiotherapy which are critical in patient survival [1]. In malignant tumors, the reduced blood oxygen tension, the inapt capillary system, or the distance between blood vessels and tumor cells can cause hypoxia [2]. In hypoxic cells, the activated HIF transcriptional factors (HIF-1, HIF-2) can cause increased resistance to apoptosis or to radio-or chemotherapy [3][4][5]. Furthermore, HIF transcriptional factors also promote the development of metastases and neoangiogenic processes in tumors by activating different genes [6]. In most cases, tumor growth and metastatic capacity depend on angiogenesis [7][8][9]. Tumor neoangiogenesis is the formation of new blood vessels from a preexistent capillary system. Integrins and aminopeptidase N (CD13) are two of the key molecules of targeting neoangiogenesis in the tumors, and the presence and expression rate of these molecules correlate with the intensity of angiogenesis. Among numerous integrins, the α v β 3 integrins are highly overexpressed on the surface of several cells that contribute to angiogenesis and tumor progression [10,11]. α v β 3 integrins are transmembrane receptors on the surface of endothelial cells and can be targeted with RGD (arginine-glycine-asparagine) motif-containing proteins [12,13]. Aminopeptidase N (CD13) is a zinc-dependent transmembrane exopeptidase [14]. It can be found in high expression on several tumor cells, for example, melanoma and prostate, ovarian, renal, colon, and pancreas cancers, furthermore on the endothelial cell surface [15][16][17]. It plays an important role in angiogenesis and enzyme-catalysed degradation of the extracellular matrix, which facilitates the tumor cell invasion through the blood stream, hereby causing metastasis formation [18,19].
In vivo imaging of tumor hypoxia and angiogenesis with positron emission tomography (PET) is playing an increasingly important role in the diagnosis of tumors. In addition, more effective antitumor treatments can be planned by using new, specific radiopharmaceuticals that detect angiogenesis and hypoxia in malignant tumors. Radiopharmacons that are labelled with positron emitting radionuclides ( 11 C, 18 F, and 68 Ga) are used in PET imaging wherewith the uptake and the biodistribution of the labelled molecule can be detected and quantificated in vivo [20]. The most commonly used PET radiopharmaceuticals are 18 F-FDG, 11 Cmethionine, and 18 F-FLT which give information about cell metabolism, but they are not specific for hypoxia or proteins and receptors that are overexpressed in tumorassociated neoangiogenesis.
For hypoxia imaging, radiolabelled (e.g., 18 F and 68 Ga) nitroimidazoles and its derivatives are widely used in PET imaging. Nitroimidazoles go through bioreduction in cells; furthermore, this process causes anion radicals. In the case of normal oxygen supply, these anion radicals are oxidized and it generates diffusible products. But in lower oxygen supply, this procedure does not take place and the free radicals are bound irreversibly to the intracellular macromolecules [21,22]. Radiolabelled nitroimidazoles (e.g., 68 Ga-DOTA-nitroimidazole) accumulate in hypoxic cells, thereby enabling the detection of the hypoxic regions in tumor mass [23].
For PET imaging of tumor-associated neoangiogenesis, radiolabelled NGR and RGD peptide-based radiopharmacons are used widespread. NGR motif (asparagine-glycyl-arginine) is a specific ligand of APN/CD13 receptors [24], while RGD's (arginine-glycine-asparagine) double-ring conformation is a selective epitope of α v β 3 and α v β 5 integrin receptors [17]. The radiolabelling of NGR and RGD with 68 Ga (e.g., 68 Ga-NOTA-c(NGR) and 68 Ga-NODAGA-[c(RGD)] 2 ) permits the specific imaging of the molecules of neoangiogenesis in the tumors in vivo [25]. Previous studies have shown that NGR peptide has high selectivity and specificity for APN/ CD13, three times more efficient in the detection of neoangiogenic vessels than RGD [26,27]; moreover, the cyclic form of NGR is ten times as effective in target detection as the linear form [28,29].
In this present study, we hypothesized that the expression of APN/CD13 and the development of hypoxia vary during the growth of subcutaneous hepatocellular carcinoma (He/De) in rats. The aim of this study was to evaluate whether 68 Ga-NOTA-c(NGR) and 68 Ga-DOTA-nitroimidazole allow the in vivo noninvasive detection of the temporal changes of APN/CD13 expression and hypoxia in experimental He/De tumors using positron emission tomography.  [29,30] at the University of Debrecen, Department of Medical Imaging. The products were used for PET imaging in sterile form after the quality control.

Cell Cultures.
He/De (chemically induced rat hepatocellular carcinoma) cells [31] were cultured in T-75 culture flasks (Sarstedt Ltd., Budapest) with 12 ml of IMDM (Thermo Fisher Scientific Inc., USA) supplemented with 1% antibiotic-antimycotic solution (Thermo Fisher Scientific Inc., USA) and 10% foetal bovine serum (FBS, Thermo Fisher Scientific Inc., USA) at 37°C, in 5% CO 2 atmosphere and 95% humidity. The cells were used for tumor induction after five passages. The cell viability was determined with a trypan blue exclusion test.  [29], for western blot analysis, frozen tissue samples were pulverized under liquid nitrogen and tissue homogenization was performed with TissueLyser II (QIAGEN). Cells were lysed in RIPA buffer (50 mM Tris, 150 mM NaCl, 0.1% SDS, 1% TritonX 100, 0.5% sodium deoxycolate, 1 mM EDTA, 1 mM Na 3 VO 4 , 1 mM NaF, 1 mM PMSF, and protease inhibitor cocktail). After tissue homogenization, the samples were subjected to protein isolation. Protein samples (10-40 μg) were separated on 10% SDS polyacrylamide gels and electrotransferred onto nitrocellulose membranes. After blocking for 1 h with TBST containing 5% BSA, the membranes were incubated with primary antibodies (dilution: 1 : 1000; mouseanti-rat CD13 and integrin alphaV/beta3 (23C6) (from Santa-Cruz Biotechnology Inc., USA)) overnight at 4°C. After washing with 1x TBST solution, the membranes were probed with IgG HRP-conjugated secondary antibody (Cell Signaling Technology, Inc., Beverly, MA, 1 : 2000). Bands were visualized by enhanced chemiluminescence reaction (SuperSignal West Pico Solutions, Thermo Fisher Scientific Inc., Rockford, USA). Densitometry was performed using ImageJ software. Upon densitometry, negative control samples (rat large intestine) were considered to be 1 and values are expressed as fold change relative to controls to reduce unwanted variations. Beta-actin was used as a loading control.

Experimental Animals
2.8. Statistical Analysis. Experimental data was presented as the mean ± SD of at least three independent experiments. The significance was calculated by Student's t-test (twotailed), two-way ANOVA, and Mann-Whitney U test. The significance level was set at p ≤ 0:05 unless otherwise indicated.  Table 1).    (Figure 4(h)). By analyzing the accumulation of the two radiopharmaceuticals during tumor growth, we found that they show an increasing uptake in close correlation with each other (Figure 4(i)). Data obtained from in vivo PET observations was confirmed by ex vivo measurements (Supplementary data 3: Table 1).

Western Blot Analysis.
The expression of α v β 3 integrin and CD13 was verified by western blot analysis in subcutaneously transplanted He/De tumors ( Figure 5). We found that among the investigated neoangiogenic markers, the expression of APN/CD13 showed strong positivity.

Discussion
Confirming the presence of hypoxia and of neoangiogenic markers in tumors by in vivo PET imaging using specific radiopharmaceuticals can greatly assist in the selection of appropriate antitumor therapy. In vivo imaging of tumor hypoxia and neoangiogenesis is an intensively researched     BioMed Research International area in the field of nuclear medicine and radiotracer development. Hypoxia influences tumor resistance to radio-or chemotherapy; moreover, hypoxia is a major stimulator of expression of different neoangiogenic molecules, e.g., VEGF and CD13, in tumors [32]. Neovascularization of malignant tumors plays an important role in tumor growth, tumor progression, and the efficacy of antitumor therapies based on antiangiogenic molecules [33][34][35].
In the human clinical routine tumor imaging, e.g., 18 F-FDG, 18 F-FLT, or 11 C-methionine [36,37], radiotracers are widely used for the detection of tumors and metastases; however, these radiotracers are not specific for hypoxia or angiogenic molecules. Furthermore, it is known that 18 F-FDG has low accumulation and poor diagnostic efficiency in well-differentiated hepatocellular carcinoma. Based on this property, it is less suitable to demonstrate the efficacy of an antitumor therapy [38]. Due to this, in our study, 68 Ga-DOTA-nitroimidazole was used for specific imaging of tumor hypoxia and 68 Ga-NODAGA-[c(RGD)] 2 and 68 Ga-NOTA-c(NGR) for imaging of neoangiogenesis in experimental hepatocellular carcinoma tumors in correspondence with tumor volume enlargement.
The first step in our study was the in vivo identification of hypoxia regions in He/De tumors with PET/MRI imaging. It is known that partial pressure of oxygen (pO 2 ) is reduced in hypoxic regions of tumors which results in the accumulation of the nitroimidazole molecule and its derivatives in hypoxic tumor cells. In our experiments, 10 days after tumor cell inoculation, high 68 Ga-DOTA-nitroimidazole uptake was already observed in subcutaneously transplanted He/De tumors at the tumor size of 125 mm 3 . Furthermore, from 11 days after the implantation of He/De cells, hypoxic ( 68 Ga-DOTA-nitroimidazole avid) regions were clearly discerned in the tumor mass with significantly higher (p ≤ 0:01) SUV values than that of the normoxic or necrotic regions ( Figure 2). This heterogeneity in the tumor mass was also described by other research groups, and they found that its rate can be 50-60% of the whole tumor [39,40]. Henceforward, in our in vivo experiments, the expression of APN/CD13 and α v β 3 integrin receptors as angiogenesis markers was investigated with PET/MRI in the He/De tumor model. It is known from other papers that these two molecules are functioning as receptors and are overexpressed on the surface of endothelial cells. Previous studies have shown that NGR peptides specifically bind to APN/ CD13, and RGD molecules are the ligands of α v β 3 integrin receptors; in addition, the 68 Ga-labelled NGR and RGD peptides are useful radiotracers for the in vivo PET imaging of neoangiogenic processes in tumors [25,27,29,41]. However, our research group previously reported that the uptake of 68 Ga-NOTAc(NGR) of the primary tumors was significantly higher than that of the accumulation of the commercially available 68 Ga-NODAGA-[c(RGD)] 2 in the same tumor, when experimental renal (Ne/De) tumors were investigated [29]. In our hepatocellular carcinoma model, this difference between the uptake of 68 Ga-NODAGA-[c(RGD)] 2 and 68 Ga-NOTA-c(NGR) in He/De tumors was also found. This observation is due to the higher expression of APN/CD13 in He/De tumors which was confirmed by western blot analysis ( Figure 5). However, in this present study, in vivo PET images of He/De tumors showed strong heterogeneity in the accumulation of 68 Ga-NODAGA-[c(RGD)] 2 and 68 Ga-NOTA-c(NGR) in subcutaneously transplanted He/De tumors. Similar to the results of our hypoxia imaging studies, significantly higher (p ≤ 0:01) SUV values were observed in the APN/CD13 and α v β 3 integrin positive regions of the He/De tumors than those of the negative areas using both angiogenesis-specific radiotracers. This observation of tumor heterogeneity is known since the angiogenic phenotype can be extraordinarily diverse within the same tumor due to hypoxia or tumorassociated inflammatory processes (TNFα, TGFβ, and IL-6) taking effect on angiogenesis. All these effects with genetic instability result in abnormal, heterogenous angiogenesis [42].
Tumor enlargement is greatly influenced by the oxygen and nutrient supply. In case of reduced oxygen supply (hypoxia), HIF transcriptional factors are activated which promotes gene expression of those responsible for tumor survival and progression. One of these hypoxia and HIFinduced processes is tumor neoangiogenesis [43]. In this present study, for the in vivo assessment of the correspondence between angiogenesis and tumor growth, 68 Ga-NOTAc(NGR) was used due to the fact that the expression of APN/CD13 was higher than that of α v β 3 integrin in He/De tumors. We hypothesized that the hypoxic and angiogenic areas increase in size with tumor volume. This assumption was confirmed since 68 Ga-NOTA-c(NGR) and 68 Ga-DOTAnitroimidazole uptake increased and showed a strong correlation with the tumor volume enlargement (Figure 4 and Supplementary data 3: Table 1). Furthermore, we hypothesized that elevation of hypoxia will induce increasingly strong angiogenesis with tumor growth. This was confirmed by indirect evidence that the increasing uptake of 68 Ga-DOTAnitroimidazole was strongly correlated with the accumulation of 68 Ga-NOTA-c(NGR) (Figure 4). Interestingly, Deshpande et al. [44] found the opposite when integrin, endoglin, and VEGFR2 expression levels of different tumors were followed with targeted microbubbles by ultrasound imaging in tumors of different sizes. They found that the expression of these angiogenic markers decreased with increasing size of tumor xenografts. An explanation for this phenomenon may be the matrigel used during the injection of tumor cells, which is known to have an angiogenesisinducing effect. Furthermore, they described that the site of tumor cell transplantation and the applied preclinical model also influence the time appearance of angiogenic markers; therefore, it is necessary to follow the change in the expression of angiogenic markers over time in as many tumor models as possible.
The correlation between hypoxia and angiogenesis was already investigated by other workgroups. In general view, angiogenesis is the outcome of hypoxia because the oxygen applied is increased, and it results in new blood vessel formation [2,6,45,46]. However, by investigation of other diseases (e.g., liver fibrosis and wound repair), it was pointed out that the relationship between hypoxia and angiogenesis is more complicated than it was earlier imagined. Hypoxia can also 7 BioMed Research International be the cause or the consequence of tissue damage. Hypoxia is not necessarily the result of blood supply, but rather, it can also be the result of the proliferation and inflammatory response; furthermore, hypoxia is the result of the contradiction between oxygen supply and need [47][48][49].
In tumors, hypoxia can be developed without any connection to oxygen supply. Oxygen utilization increases fivefold during cell proliferation; consequently, hypoxiainduced angiogenesis also occurs at small tumor size. Small tumors consist of approximately 100 cells can induce angiogenesis by the reduced pO 2 level due to the increased oxygen consumption. In this present work, we also found that hypoxic regions can be observed in He/De tumors at a small size by 68 Ga-DOTA-nitroimidazole imaging (Figure 4(a)).
In summary, as our research group previously demonstrated for Ne/De renal tumors [29], 68 Ga-NOTA-c(NGR) is a suitable diagnostic agent for the detection of APN/CD13 expression in a He/De hepatocellular carcinoma tumor also. APN/CD13 and hypoxia-specific radiopharmaceuticals may contribute significantly to a better understanding of the relationship between hypoxia and neoangiogenesis in tumors. Therefore, the in vivo detection of the presence of angiogenic molecules-as potential therapeutic targets-and the determination of the changes in their expression levels during tumor growth may play an important role in the development of antitumor therapies.

Conclusion
Noninvasive in vivo PET/MRI imaging using 68 Ga-NOTAc(NGR) and 68 Ga-DOTA-nitroimidazole-as specific radiolabelled diagnostic molecules-provides an opportunity to the detection of temporal changes of hypoxic regions and neoangiogenic molecule (APN/CD13) expression in hepatocellular carcinoma tumors. These results give the possibility of the development of new molecular imaging strategies, the early detection of tumors, and the monitoring of antitumor therapies.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
We declare that we have no conflict of interest.

Acknowledgments
The research was financed by the Thematic Excellence Programme of the Ministry for Innovation and Technology in Hungary (ED_18-1-2019-0028), within the framework of the Space Sciences thematic programme of the University of Debrecen.

Supplementary Materials
Supplementary data 1: Figure