Results

3.1 Confirmation of constructions

HT-1080, HT-29, and PANC-1 were transduced with OATP1B3, NTCP, or ASBT membrane transporters. In HT-1080, the plasmids used for overexpression were pWPXL-OATP1B3-ires-Puro, RC210241L2, and RC221202L2 (Fig. 1C, 1F, and 1G).

RC210241L2, and RC221202L2 contain GFP. The overexpressions of OATP1B3, NTCP-GFP, ASBT-GFP, and GFP in HT-1080 were validated by western blotting (Fig.

2). Further, the cellular location of OATP1B3 was visualized by OATP1B3 antibody in green color, and others were using GFP protein to locate the NCTP and ASBT (Fig. 3).

In HT-29, the plasmids used for overexpression were pWPXL NTCP-ires-Puro and pWPXL ASBT-ires-Puro (Fig. 1E and 1G). The overexpressed OATP1B3 and NTCP in HT-29 were confirmed by western blotting and immuno-fluorescent staining (Fig. 2 and 3). In PANC-1, the overexpression of OATP1B3 also confirmed by Q-PCR, western blotting, and immuno-fluorescent staining (Fig. 2 and 3).

3.2 Evaluation of the intake capacity of ICG in vitro

OATP1B3, NTCP, and ABST expressing HT-1080 had higher ICG intensity in confocal imaging, flow cytometry, and multimode detection platform data. Among them, the order of increased intake capacity of ICG was NTCP, ASBT, and OATP1B3 expressing HT-1080 (Fig. 4). Only OATP1B3 and NCTP expressing HT-29 remained a higher intake of ICG. Still, the intake of ICG was better in NTCP than OATP1B3 (Fig. 5 and 6).

The longer time for treating ICG, the higher ICG intensity was detected by the multimode detection platform. The difference of ICG intensity between NTCP and OATP1B3 expressing HT-29 was differentiated in 30 and 60 min treatments (Fig. 5A).

Moreover, OATP1B3 and NCTP expressing HT-29 could remain ICG in the cell at least for 72 h (Fig. 5B). The visualization of ICG in NTCP and OATP1B3 expressing HT-29 was using IVIS, and its detecting limitation was 3.13 x 10³ and 1.25 x 10⁴ in NTCP and OATP1B3 expressing HT-29, respectively (Fig. 6B).

3.3 Evaluation of the intake capacity of fluorescein isothiocyanate (FITC) in vitro

NaFluo, transported through OATP1B3⁶, and fluorescein isothiocyanate (FITC) have a similar structure. We tested whether OATP1B3 was a FITC transporter. After treating 100μM FITC, the intake of FITC was increased in ASBT, NTCP, and OATP1B3 expressing HT-29 under fluorescent microscopy examining at 1 h after washing (Fig. 7) and showing in flow cytometry data (Fig. 5D).

3.4 Evaluation of the intake capacity of Primovist and other contrasts in vitro

OATP1B3 had high specificity to Primovist instead of Gadovist, Magnevist, and Omniscan. Furthermore, ASBT and NTCP couldn’t serve as the transporters for these four Gd containing MR contrasts (Fig. 8).

3.5 Evaluation of the intake of ICG in vivo and ex vivo

After inoculating NCTP and OATP1B3 expressing HT-29 into nude mice and injecting ICG, the ICG signals were traced in a time-dependent manner. NTCP and OATP1B3 expressing HT-29 could retain the ICG for at least 72 h. Moreover, OATP1B3 expressing HT-29 had better retain the ability of ICG (Fig. 9). After 2 d ICG administration, all organs were examined using IVIS for ICG bio-distribution.

OATP1B3 expressing HT-29 tumor had the strongest ICG intensity, and the followings were liver, NTCP expressing HT-29 tumor, and the others (Fig. 10A and 10B). These HT-29 tumors have further confirmed the overexpression of NTCP and OATP1B3 using immunohistochemistry (Fig. 10C).

3.6 Evaluation of the intake of Primovist in vivo

After inoculating NCTP and OATP1B3 expressing HT-29 into nude mice for 11 d and injecting Primovist, the MR images were taken. Only OATP1B3 expressing HT-29 tumors with small size had elevated MRI intensity not interfered by tumor size (Fig.

11).

3.7 The application of OATP1B3 for tumor cell tracking using IVIS

After inoculating OATP1B3 expressing HT-1080 into nude mice, ICG was administrated through subcutaneous injection. The ICG signal was sustainable for at least 96 h (Fig. 12A and 12B). For validating the biodistribution of ICG, the mice with OATP1B3 expressing HT-1080 tumor was enthused at 2 d ICG injection. The OATP1B3 expressing HT-1080 tumor had the strongest ICG signal (Fig. 12C and 12D). At the endpoint, OATP1B3 expressing HT-1080 tumor remained higher OATP1B3 expression (Fig. 12E).

3.8 The application of OATP1B3 for tumor cell tracking using MRI

The nude mice bearing OATP1B3 expressing HT-1080 tumor was administrated Primovist through intravenous injection and detected in 7T-MRI. OATP1B3 expressing HT-1080 tumor had higher MR intensity (Fig. 13A and 13C). Moreover, the dark pattern in MR imaging was correlated to terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay and histology indicating living cells had Primovist intake ability (Fig. 13B and 13C). The more comprehensive information could be revealed by MR imaging compared with IVIS imaging (Fig. 13C).

3.9 The application of OATP1B3 for PANC-1 islet-like cell tracking in cell therapy

We chose PANC-1 for the investigation of cell tracking in cell therapy since PANC-1 is a potential cell for treating diabetes⁶⁴. After transduced OATP1B3 into PANC-1, the

Primovist intake ability was increased in vitro 3T-MRI and further confirmed the high quantity of Gd with ICP-MS (Fig. 14A and 14B). OATP1B3 overexpression doesn’t affect the cellular functions in mitochondria membrane potential, viability, and ROS reaction (Fig. 14C, 14D, and 14E). Moreover, the differentiation ability remained after OATP1B3 transduction into PANC-1, such as sphere formation, the expression of insulin (Ins) and glucagon (GCG) in RNA and protein levels (Fig. 15).

The SCID mice bearing with OATP1B3 expressing PANC-1 xenograft had a high MR signal after Primovist injection (Fig. 16A and 16C). At the endpoint, OATP1B3 expressing PANC-1 xenograft remained higher OATP1B3 expression (Fig. 16B).

Instead of tracing MR signal in vivo, the MR signal was tracked for 44 h in OATP1B3 expressing PANC-1 in vitro. The differentiation of MR signal was significant from the beginning to 2 h and OATP1B3 expressing PANC-1 still had insignificantly higher MR intensity at least for 28 h after Primovist washed out (Fig. 16D).

3.10 The utility of NTCP in a drug screening platform in vitro

To establish the drug screening platform for HBV through blocking NTCP, the candidate drugs (cyclosporin A, emodin, and erythrosin B) were treated first and added ICG for evaluating the possibility for the HBV treatment. The evaluation was confirmed by two methods; one is a multimode detection platform, and the other one is IVIS. Both methods represented similar results: cyclosporin A, emodin, and erythrosin B could

decrease ICG intensity indicating they could be the candidate drugs for HBV treatments (Fig. 17). Moreover, cyclosporin A treatment had a dose-dependent manner in ICG intensity in the multimode detection platform (Fig. 17C).

3.11 The application of OATP1B3 as a luminescent reporter

OATP1B3 and luciferase transduced into HT-1080 and the overexpression was confirmed in RNA level (Fig. 18A). Furthermore, the luminescent intensity was increased in OATP1B3 and Luciferase expressing HT-1080 compared with single Luciferase expressing HT-1080 (Fig. 18B). The luminescent intensity in OATP1B3 and Luciferase expressing HT-1080 was decreased after the inhibitor treatment (Fig. 18C).