Evaluation of Riboflavin Transporters as Targets for Drug Delivery and Theranostics

Abstract

The retention and cellular internalization of drug delivery systems and theranostics for cancer therapy can be improved by targeting molecules. Since an increased uptake of riboflavin was reported for various cancers, riboflavin and its derivatives may be promising binding moieties to trigger internalization via the riboflavin transporters (RFVT) 1, 2, and 3. Riboflavin is a vitamin with pivotal role in energy metabolism and indispensable for cellular growth. In previous preclinical studies on mice, we showed the target-specific accumulation of riboflavin-functionalized nanocarriers in cancer cells. Although the uptake mechanism of riboflavin has been studied for over a decade, little is known about the riboflavin transporters and their expression on cancer cells, tumor stroma, and healthy tissues. Furthermore, evidence is lacking concerning the representativeness of the preclinical findings to the situation in humans. In this study, we investigated the expression pattern of riboflavin transporters in human squamous cell carcinoma (SCC), melanoma and luminal A breast cancer samples, as well as in healthy skin, breast, aorta, and kidney tissues. Low constitutive expression levels of RFVT1-3 were found on all healthy tissues, while RFVT2 and 3 were significantly overexpressed in melanoma, RFVT1 and 3 in luminal A breast cancer and RFVT1-3 in SCC. Correspondingly, the SCC cell line A431 was highly positive for all RFVTs, thus qualifying as suitable in vitro model. In contrast, activated endothelial cells (HUVEC) only presented with a strong expression of RFVT2, and HK2 kidney cells only with a low constitutive expression of RFVT1-3. Functional in vitro studies on A431 and HK2 cells using confocal microscopy showed that riboflavin uptake is mostly ATP dependent and primarily driven by endocytosis. Furthermore, riboflavin is partially trafficked to the mitochondria. Riboflavin uptake and trafficking was significantly higher in A431 than in healthy kidney cells. Thus, this manuscript supports the hypothesis that addressing the riboflavin internalization pathway may be highly valuable for tumor targeted drug delivery.

Keywords: RFVT; cancer therapy; nanocarrier; riboflavin; targeting molecules; theranostic.

FIGURE 1

Histopathological stainings of RFVT1, 2, and 3 on SCC, skin melanoma, luminal A breast cancer and healthy control tissues (skin, breast, aorta, and kidney). RFVTs: green, nuclei: blue; scale bar: 50 μm. The tumor tissues show significant upregulation of RFVTs compared to their respective healthy organ tissues. (A) Healthy skin shows low expression levels of all RFVTs, with RFVT1 expression in blood vessels (white arrows) and sweat glands (yellow arrows), while lacking epithelial expression (white dotted line). In contrast, SCC presents with an elevated expression of RFVT1, 2, and 3 in cancer clusters (red dotted line). Melanoma tissues upregulate RFVT2 and 3 (red dotted line), however, only a low expression level of RFVT1 is found in the tumor vasculature (white arrows). (B) Healthy breast tissue shows low expression levels of RFVT1, 2, and 3 in lactiferous ducts. Luminal A breast cancer tissue is weakly positive for RFVT2 and presents significant overexpression of RFVT1 and RFVT3 in ductal carcinoma in situ (red dotted line). (C) The healthy aorta shows endothelial presence of mainly RFVT1 with low expression of RFVT3 (white arrows). Yellow arrows point to autofluorescence of elastin. Healthy kidneys have a low constitutive expression of all three RFVTs.

FIGURE 2

Mean fluorescence intensity quantification of histopathological staining of RFVT1, 2, and 3 in SCC, skin melanoma, luminal A breast cancer, and healthy control tissues (skin, breast, aorta, and kidney). All tumors significantly overexpress all three RFVTs compared to their respective healthy skin and breast tissues. Each tumor type presents an individual RFVT expression pattern showing the highest expression of RFVT1 in breast cancer, the highest RFVT2 expression in melanoma and the highest RFVT3 expression in SCC. Healthy aorta and kidneys significantly overexpress RFVT1 and 3 compared to healthy skin and breast tissue. Mean fluorescence intensity was quantified using ImageJ in artificial units (AU). Error bars represent standard deviation; ^∗∗∗p < 0.0001, n = 3. Data was normalized using background subtraction of the isotype control.

FIGURE 3

In vitro immunofluorescence stainings of RFTV1, 2, and 3 in A431, HK2, and HUVEC. RFVTs: green, cell membrane: red, nuclei: blue; scale bar: 30 μm. A431 cells strongly upregulate all three RFVTs, both intracellularly and on the cell membrane. HK2 cells present low expression levels of RFVT1 and 3, while RFVT2 is moderately present. HUVEC present strong expression of RFVT2 and weak RFVT1 and 2 expression. Isotype IgG antibodies were used for control samples. n = 3.

FIGURE 4

Quantitative analysis of RFVT1, 2, and 3 mRNA expression and cellular localization. (A) The quantitative PCR analysis of RFVT mRNA expression in all three selected cell lines (A431, HK2, and HUVEC) confirmed histological findings, showing significantly elevated mRNA levels for RFVT1 and 3 in A431 cells. In contrast, HK2 and HUVEC presented low mRNA levels for all three RFVTs (^∗∗∗p < 0.0001, n = 3). (B) Z-stack confocal microscopy images of RFVT stainings in A431 cells show the co-localization (white, indicated by arrows) between the apical cell membrane (red) and the RFVTs (green). Scale bar = 15 μm. (C) Three-dimensional volume co-localization analysis and determination of the Mander’s Coefficients for RFVT colocalization with the cell membrane in A431 cells was performed with the IMARIS co-localization plugin. No significant differences between the membrane presences of the three RFVTs were found (n.s., not significant).

FIGURE 5

Quantification of riboflavin and transferrin co-localization in A431 and HK2 cells based on confocal microscopy and Mander’s Coefficient. Cells were incubated with riboflavin (red) and either Transferrin-AF488 (green) or MitoTracker^TM Deep Red (green) for 120 min. Nuclear staining (blue) was achieved with DAPI. Co-localization (yellow) between riboflavin and transferrin or MitoTracker^TM Deep Red was quantified using Mander’s Coefficients with the ImageJ plugin JACoP. M1 = the overlap ratio of total riboflavin with transferrin or MitoTracker^TM Deep Red, M2 = the overlap ratio of total transferrin or MitoTracker^TM Deep Red with riboflavin. Scale bar = 10 μm. Riboflavin colocalized strongly with transferrin in both (A) A431 cells as well as (B) HK2 cells, whereas transferrin only moderately colocalized with riboflavin in HK2 cells. Similarly, riboflavin strongly colocalized with MitoTracker^TM Deep Red in both (C) A431 cells and (D) HK2 cells. In contrast, MitoTracker^TM Deep Red showed low colocalization with riboflavin.