Rat embryonic fibroblasts immortalized by MRPS18-2 protein are target for NK-cells

Abstract

Overexpression of the human mitochondrial ribosomal protein MRPS18-2 (S18-2) led to immortalization of primary rat embryonic fibroblasts (REFs). The derived cells (18IM) expressed embryonic stem cell markers. Noteworthy, genes encoding the COX family proteins were up-regulated significantly. It is known that the COX family proteins are involved in the regulation of immune response. In the present work we demonstrate that 18IM cells behave like stem cells when subjected to directed differentiation in vitro. However, unlike stem cells, 18IM cells do not develop tumors in vivo, in SCID mice. This phenomenon is observed due to the strong natural killer (NK) cell immunogenicity. 18IM cells were better recognized by NK cells, compared with primary REFs, as was shown by a standard NK killing assay. Our data explain asymmetry in behavior of stem-like cells in vivo and in vitro, and this support the notion that stem and/or cancer-initiating cells are preferred targets for NK-cells. Concluding, the S18-2 protein is a putative target for cancer vaccines.

Keywords: NK killing; cell immortalization; cytokine expression profile; mitochondrial ribosomal protein MRPS18-2; stem cells.

Figure 1. Directed differentiation of 18IM cells in vitro

A. - 18IM cells grown on coverslips were treated for 1 week with 20 µM of RA in culture medium; a - control 18IM cells; b-d, differentiated cells. Notice the change of morphology after treatment with RA. B. - 18IM cells were differentiated as in (A). Upper panel - immunostaining with anti-MAP-2 antibody (green); bottom panel - with anti-nestin antibody (green). DNA is shown in blue. Objective was x63. C. - 18IM cells were differentiated with a cocktail of dexamethasone, ascorbic acid-2-phosphate, and glycerol-2-phosphate in culture medium for 4 weeks: a - untreated 18IM cells; b-c - differentiated cells, both stained with Alizarin Red S. The strong red signal suggests mineralization of the osteogenic cells. Objective was x40; d - Runx2 expression at the mRNA level, assessed by q-PCR. D. - 18IM cells were differentiated for 4 weeks upon treatment with 0.5 µM and 1 µM of the H-89 in culture medium. The untreated 18IM cells are shown in panel a, and differentiated cells are shown in panels b-d; both are stained with Alcian Blue. Bright blue staining and a change in cell morphology are seen in differentiated cells (panels b-d). Objective was x63.

Figure 2. Expression of chemokines and their receptors in 18IM cells, compared with REFs

A. - The heat map of the fold changes in gene expression between 18IM cells and REFs for every gene in the array, in the context of the array layout (see Figure S1 for the genes). B. - 18IM cells and primary REFs were incubated with either anti-MHC-I or unrelated isotype control antibodies. A representative FACS analysis showing the difference in MHC-I molecule expression between REFs and 18IM cells; the inserted panel shows statistical analysis of MHC-I expression from three independent experiments, using two tailed paired T test on the combined mean values. MFI - the mean fluorescence intensity; a. u. - arbitrary units.

Figure 3. The cytotoxic recognition of REFs and 18IM cells by splenocytes, studied at various splenocyte-to-target-cell (E:T) ratios

Rat splenocytes were used as effector (E) cells, and REFs and 18IM cells as targets (T) in the assay. Non-parametric t-test (panels A and B) and Wilcoxon signed rank test (C and D) were used to compare a median of three different experiments, performed in triplicates for all the E:T ratios. A. - No differences were observed between REFs and 18IM cells for naïve splenocytes (p = 0.0747). YAC - control mouse lymphoma YAC-1 cell line. B. - The IL-2-activated rat splenocytes preferentially recognize 18IM cells, compared with primary REFs (p = 0.0305). C. - NKp46 blocking assay for REF recognition by activated splenocytes: SPL treated with anti-NKp46 antibody (NKp46, p = 0.0938), and SPL treated with the isotype control antibody (isotype control, p = 0.0625). D. - NKp46 blocking assay for 18IM cell recognition by activated splenocytes: SPL treated with anti-NKp46 antibody (NKp46, p = 0.0320), and SPL treated with the isotype control antibody (isotype control, p = 0.0625).

Figure 4. Recognition of 18IM cells by NK-cells of SCID mice in vitro and in vivo

A. - Splenocytes of SCID mice were used as effector (E), and 18IM cells and REFs as targets (T) in the in vitro NK cell cytotoxicity assay. The cytotoxic recognition was studied at E:T ratios 100:1 and 50:1. Non-parametric t-test was used to compare a median of three different experiments, and for one experiment a pool of the activated NK cells from two spleens was used. 18IM cells were killed by the IL-2 activated NK cells more efficiently than REFs (p = 0.0313 for E:T = 100:1 and p = 0.0303 for E:T = 50:1), while no significant differences were observed between 18IM and the control mouse lymphoma YAC-1 (YAC) cells killing. B. - Tissue sections show a bulk of 18IM cells and NKp46-positive cells in the proximity of the former (black arrows on the panels a, c, d, and e). Panels a and b are merged from the two sequentially captured images with objective x10. Black arrows with numbers on the panel a correspond to those on the panel c (the image on c was captured with objective x40). Notice the cytoplasmic NKp46 signal (brown) on the panel d (objective was x100). 18IM cells that show S18-2 signal (pink) on the panel e were captured with objective x40 (the image corresponds to cells framed in the square on the panel a).

Figure 5. 18IM cells are immortalized but do not produce tumors in experimental animals, showing the asymmetric behavior in vitro and in vivo

18IM cells are recognized (and killed, eventually) by NK cells of SCID mice. They can be trans-differentiated after inoculation as well. AR - activating receptors on NK cells; IR - inhibitory receptors on NK cells.