Optimized functional and structural design of dual-target LMRAP, a bifunctional fusion protein with a 25-amino-acid antitumor peptide and GnRH Fc fragment

Abstract

To develop fusion protein of a GnRH Fc fragment and the integrin targeting AP25 antitumor peptide for GnRH receptor-expressing cancer therapy. The LMRAP fusion protein was constructed. A transwell invasion assay was performed. The gene mRNA and protein levels of GnRHR-I, α5β1, and αvβ3 in different cancer cell lines were assessed. Cell proliferation was measured using a cell counting kit-8. An antagonist assay was performed on GnRH receptors. Anti-tumor activity was evaluated with a mouse xenograft tumor model. Immunohistochemistry (IHC) was applied to detect CD31 and CD34 expressions. Pharmacokinetic characteristics were determined with an indirect competition ELISA. The developed bifunctional fusion protein LMRAP not only inhibited HUVEC invasion, but also inhibited proliferation of GnRHR-I, α5β1, and αvβ3 high expression cancer cells. The IC₅₀ for LMRAP in the GnRH receptor was 6.235 × 10^-4 mol/L. LMRAP significantly inhibited human prostate cancer cell line 22RV1 proliferation in vivo and in vitro. LMRAP significantly inhibited CD31 and CD34 expressions. The elimination half-life of the fusion protein LMRAP was 33 h in rats. The fusion protein made of a GnRH Fc fragment and the integrin targeting AP25 peptide retained the bifunctional biological activity of GnRHR blocking, angiogenesis inhibition, prolonged half-life and good tolerance.

Keywords: Angiogenesis; Fusion protein; GnRH; Integrin; Prostate cancer.

Graphical abstract

Figure 1

Schematic of the domain arrangements and structural identifications of LMRAP, LMRAP-A, and LMRAP-B. (A) Schematic of LMRAP, LMRAP-A, and LMRAP-B domain arrangements. (B) SDS-PAGE analyses of the final products after being purified with affinity filler Prosep Ultra Plus. Marker: molecular weight marker; Lanes A–C: LMRAP-A (reduced), LMRAP-B (reduced), LMRAP (reduced), respectively; Lanes D–F: LMRAP-A (non-reduced), LMRAP-B (non-reduced), LMRAP (non-reduced), respectively. Confirmation of proteins sequences with LC–MS of LMRAP (C), LMRAP-A (D), LMRAP-B (E).

Figure 2

LMRAP, LMRAP-A, but not LMRAP-B, inhibited the invasion activity of HUVECs. Cell invasive ability was determined with a transwell invasion assay employing Boyden chambers coated with Matrigel. HUVECs in serum-free media were put into the upper chamber of an insert. The cells were then treated with AP25, avastin, LMRAP-A, LMRAP, or LMRAP-B. The cells that had invaded through the membrane were stained with 0.1% crystal violet and methanol. The cells were then imaged and counted in random fields in each well under a microscope at 100× magnification. Data are representative of three independent experiments (n = 3) (A). The migration cell numbers (B) and the migration inhibition rate (C) are shown based on the number of invasion cells in the experiment in different groups. Data are mean±SD; *P < 0.05, **P < 0.01 compared with the control group.

Figure 3

The mRNA and protein levels of GnRHR-I, α5β1, and αvβ3 were measured with real-time PCR and Western blot in different cancer cell lines. The SKOV3 human ovarian cancer and 22RV1 human prostate cancer cells had high GnRHR-I expression (A). Human cervical cancer SiHa, human prostate cancer PC-3 and human ovarian cancer A2780 had medium GnRHR-І expression (A). Human cervical cancer SiHa and human ovarian cancer SKOV3 had high α5β1 expression. Human ovarian cancer A2780, human prostate cancer 22RV1 and PC-3 had medium α5β1 expression (B and C). Human ovarian cancer SKOV3 had high αvβ3 expression. Human prostate cancer 22RV1 had medium αvβ3 expression (D and E). Data are mean±SD, n=3. The protein expression levels of GnRHR-I, α5β1, and αvβ3 were measured by Western blot in different cancer cell lines (F).

Figure 4

The in vitro antiproliferative effect of LMRAP and LMRAP-A on GnRHR-I expression cells was determined by CCK-8. LMRAP significantly inhibited GnRHR-I positive cell viability in human prostate cancer 22RV1 (A) and PC-3 (D), human ovarian cancer SKOV3 (B), and A2780 (C) cells in vitro from 6.25 to 12.5 μmol/L. AP25 itself inhibited cell viability in 22RV1 (A), SKOV3 (B), A2780 (C), PC-3 (D) and SiHa (E) cells in vitro from 25 to 50 μmol/L. Data are mean±SD, n=6; *P<0.05, **P<0.01 vs. Control.

Figure 5

Functional characterization of LMRAP. The antagonist assay on GnRH receptors of gonadorelin and LMRAP in the CHO-K1/GnRHR/Gα15 stable cell line showed the IC₅₀ for gonadorelin was 1.641 × 10⁻⁹ mol/L (A) and LMRAP was 6.235 × 10⁻⁴ mol/L (B). Data are mean±SD, n=3.

Figure 6

In vivo anti-tumor study of LMRAP. Xenograft 22Rv1 tumors were induced by s.c. flank injection in nude mice. This model was used to assess the therapeutic efficacy of LMRAP in vivo. Tumor growth curve (A) and the tumor weight (B) of tumor bearing nude mice showed that the T/C (%) of LMRAP 12.5, 25, and 50 mg/kg for transplanted tumors of 22RV1 nude mice were 56.34%, 47.44%, and 32.16%, respectively, and the inhibition rates were 29.56%, 48.00%, and 61.97%, respectively. **P < 0.01 compared with the model group (Data are mean±SD, n = 8 per group except n = 16 in model group).

Figure 7

IHC detection of CD31 and CD34 expressions. Histological sections from FFPE xenograft tumors were used. (A) Positive signal of CD31 and CD34 in the model group, LMRAP (50 mg/kg) and Avastin (5 mg/kg). (B) CD31 and CD34 expression densities were independently assessed. LMRAP significantly inhibited both CD31 and CD34 expressions in prostate cancer. Data are mean±SD of eight independent experiments (n = 8). **P < 0.01 compared with the model group. The bar = 50 μm.

Figure 8

Plasma concentration–time curve of LMRAP following a single i.v. administration to a rat at a dose of 12.5 mg/kg. Data are mean±SD, n=6.