CD44v8-10 as a potential theranostic biomarker for targeting disseminated cancer cells in advanced gastric cancer

Abstract

Gastric cancer is the third most common cause of cancer mortality, and the survival rate of stage IV advanced gastric cancer (AGC) patients with distant metastasis is very low. Thus, the detection and eradication of disseminated cancer cells by targeting cell surface molecules in AGC would improve patient survival. The hyaluronic acid receptor, CD44, has various isoforms generated by alternative splicing, and some isoforms are known to be correlated to gastric cancer. In this study, to find out the most appropriate CD44v for targeting AGC, we analysed the expression differences of CD44 isoforms at the mRNA level in stomach cancer cell lines as well as in 74 patients with AGC by using exon-specific qRT-PCR. Among the CD44v isoforms, CD44v8-10 was determined as the most promising biomarker for the development of theranostic agents of gastric cancer. Next, we synthesised the conjugate of anti-CD44v9 antibody with near-infrared fluorophore or photosensitiser, and then demonstrated its feasibility for target cell-specific imaging and photoimmunotherapy in gastric cancer. As a result, these conjugates have clearly demarcated the surface of CD44v8-10 expressing cancer cells and showed efficient phototoxic effects. Therefore, this study revealed that CD44v8-10 is the efficient theranostic biomarker to target disseminated cancer cells in AGC.

Figure 1

(A) Structure of human CD44 pre-mRNA and its alternative splice variants. To detect alternative splice variants, CD44s or CD44v specific TaqMan probes spanning the border of contiguous exons and primer pairs were applied. (B) The comparison of mRNA expression levels of CD44s and CD44v in the cell lines measured by absolute quantification of qRT-PCR. (C) The comparison of mRNA expression level of CD44s and CD44v between normal and tumour tissue of stomach cancer patients (n = 74).

Figure 2

(A) Immunohistochemistry of normal-tumour paired gastric cancer tissue using anti-CD44v9 antibody. Normal tissue showed no CD44v9 signals in the section and has regularly arranged gastric ducts. In comparison, the CD44v9 signal is prevalent in tumour tissue and it has dysplastic phenotypes at the ductal regions. (B) Immunofluorescence stained images with tissue sample obtained from a gastric cancer patient using anti-CD44v9-ATTO680.

Figure 3

Establishment of stably CD44s and CD44v8-10expressing cells. (A) Stable expression of GFP-tagging open reading frame was confirmed by flow cytometry. (B) Expression of the stably expressed CD44s and CD44v8-10 was confirmed at mRNA level by qRT-PCR. (C) Protein level was confirmed by western blot using pan-CD44 antibody and CD44v9 antibody. (D) The location of CD44s and CD44v8-10 on MKN-28 cells was also confirmed by immunocytochemistry.

Figure 4

Specificity of ATTO680-conjugated anti-CD44v9 antibody on CD44v8-10(−) or (+) cells confirmed by NIR fluorescence image. No fluorescent signal was detected in MKN-28 cells, mock vector transfected MKN-28 cells, CD44s transfected MKN-28 cells, and HDF cells. The CD44v8-10 expressing MKN-28 cells, SNU-638 and KATOIII cells, clearly showed fluorescence on the lining of the cell membrane. Scale bar indicates 20 μm.

Figure 5

Targeted photoimmunotherapy of CD44v8-10(+) cancer cells using anti-CD44v9-MB conjugate. Results obtained with CD44(−) HDF/CD44(−) Mock MKN-28 cell combination (A) and CD44(−) MKN-28/CD44v8-10(+) MKN-28 cell combination (B). Cells were treated with anti-CD44v9-MB conjugate (8.96 μg/0.3 mL, 0.2 μM MB equivalent) for 1 h, and then the cells were irradiated with 633-nm laser light. Scale bar indicates 20 μm.