Complex Interplay of Genes Underlies Invasiveness in Fibrosarcoma Progression Model

Abstract

Sarcomas are a heterogeneous group of mesenchymal tumours, with a great variability in their clinical behaviour. While our knowledge of sarcoma initiation has advanced rapidly in recent years, relatively little is known about mechanisms of sarcoma progression. JUN-murine fibrosarcoma progression series consists of four sarcoma cell lines, JUN-1, JUN-2, JUN-2fos-3, and JUN-3. JUN-1 and -2 were established from a single tumour initiated in a H2K/v-jun transgenic mouse, JUN-3 originates from a different tumour in the same animal, and JUN-2fos-3 results from a targeted in vitro transformation of the JUN-2 cell line. The JUN-1, -2, and -3 cell lines represent a linear progression from the least transformed JUN-2 to the most transformed JUN-3, with regard to all the transformation characteristics studied, while the JUN-2fos-3 cell line exhibits a unique transformation mode, with little deregulation of cell growth and proliferation, but pronounced motility and invasiveness. The invasive sarcoma sublines JUN-2fos-3 and JUN-3 show complex metabolic profiles, with activation of both mitochondrial oxidative phosphorylation and glycolysis and a significant increase in spared respiratory capacity. The specific transcriptomic profile of invasive sublines features very complex biological relationships across the identified genes and proteins, with accentuated autocrine control of motility and angiogenesis. Pharmacologic inhibition of one of the autocrine motility factors identified, Ccl8, significantly diminished both motility and invasiveness of the highly transformed fibrosarcoma cell. This progression series could be greatly valuable for deciphering crucial aspects of sarcoma progression and defining new prognostic markers and potential therapeutic targets.

Keywords: Ccl8; fibrosarcoma; invasiveness; progression series; transcriptome.

Figure 1

Fos and Jun expression in fibrosarcoma cell lines. (A,B) Comparison of jun and fos expression in different JUN-sarcoma cell lines. Results of permutational ANOVA: F_3,20 = 1531, ** p < 0.01 (jun expression) and F_3,20 = 2948, ** p < 0.01 (fos expression). See Table S1 for results of contrasts between pairs of lines, their confidence intervals, and exact p-values. Each point represents an individual experiment (n = 6). Violin plots with means ± SEM are shown.

Figure 2

Fos and Jun oncoproteins expression in fibrosarcoma cell lines. Indirect immunofluorescence analysis. Notice the high nuclear expression of both oncoproteins in JUN-2 as well as JUN-2fos-3 cells and their low expression levels in JUN-1 cells. Fos seems to be quite highly expressed in JUN-3 cells as well, with an unusual perinuclear localisation. Both JUN-1 and JUN-3 seem to be generally devoid of appreciable nuclear Jun, except some individual scattered cells (arrows). Pictures were taken by Olympus IX70 fluorescent microscope equipped with the Olympus DP71 camera system (Bar: 100 µm) Negative controls are shown in Figure S2.

Figure 3

Morphological characteristics of JUN-sarcoma cell lines. (A–C) Size analysis of single adherent cells of JUN-sarcoma cell lines. Three parameters were evaluated—the total cell area of an adherent cell (μm²; permutational ANOVA: F_3,34 = 3.4, p = 0.021), the cell perimeter (μm; permutational ANOVA: F_3,34 = 45, p < 0.001), and the cell roundness (permutational ANOVA: F_3,34 = 3.8, p = 0.018). The JUN-2-fos-3 cell line presented the largest cell size with prominent lamellipodia. * p < 0.05, ** p < 0.01, *** p < 0.001. The statistical significances are based on permutational t-test with FDR correction. See Table S2 for effect sizes, their confidence intervals, and exact p-values. Each point represents an individual cell (n = 15). Violin plots with means ± SEM are shown.

Figure 4

Proliferation and stemness-related characteristics of JUN-sarcoma cell line. (A) Proliferation of JUN-sarcoma cell lines. The doubling time and slope analysis (Table 1) were performed on linear growth phase, set arbitrarily as growth curve interval between 45 and 75 h (permutational ANOVA: F_3,34 = 33, p < 0.001). The JUN-3 cell line showed the fastest growth. The proliferative activity of the newly established derivative JUN-2fos-3 was inferior to all the other sarcoma cell lines of the series, even below the proliferative activity of its parental cell line JUN-2. (B) Growth curve of the JUN-sarcoma cell lines. The proliferative activity of the newly established derivative JUN-2fos-3 was inferior to all the other sarcoma cell lines of the series, even below the proliferative activity of its parental cell line JUN-2. (C) Clonogenicity in semisolid media differ among JUN-sarcoma cell lines (permutational ANOVA: F_3,53 = 252, p < 0.001). JUN-3 was the most active cell line, whereas the JUN-2fos-3 showed the weakest clonogenicity. Representative pictures of colonies formed in 15% methylcellulose are shown in (D). The pictures were taken by the Olympus IX 70 inverted microscope equipped with the Hamamatsu Orca-ER camera at 100× magnification. (E) Sarcosphere formation capacity differed among JUN-sarcoma cell lines (permutational ANOVA: F_3,20 = 14.8, p < 0.001). Both JUN-2fos-3 and JUN-3 presented rather high sarcosphere formation activity. Representative pictures of spheres are shown in (F). The pictures were taken by the Olympus IX 70 inverted microscope equipped with the Hamamatsu Orca-ER camera at 40× magnification. * p < 0.05, ** p < 0.01, *** p <0.001 (A,C,E, respectively). The statistical significances are based on a permutational t-test with FDR correction. See Table S3 for effect sizes, their confidence intervals, and exact p-values. Each point represents an individual well. Violin plots with means ± SEM are shown.

Figure 5

Invasion-related characteristics of JUN-sarcoma cell line. (A–D) JUN-2fos-3 cells are highly motile and invasive. (A) Cell motility after 24 and 48 h in vitro wound-healing test. The newly established JUN-2fos-3 cell line was highly motile in the in vitro wound-healing assay compared to its mother cell line JUN-2. Representative pictures were taken by the Olympus IX 70 inverted microscope at 40× magnification (at 100× magnification in detail (B)). (C) Invasion of multicell tumour spheroids of JUN-sarcoma cell lines embedded into type I collagen at 100× magnification. JUN-3 and JUN-2fos-3 cell lines showed comparatively intensive invasion, whereas the JUN-2 cell line was completely non-invasive. JUN-1 cell line displayed minimal invasiveness in type I collagen. (D) Matrigel in vitro invasion assay differed among JUN-sarcoma cell lines (permutational ANOVA: F_3,23 = 910, p < 0.001). Both the JUN-2fos-3 and the JUN-3 cell lines showed comparatively intensive invasion in this assay. * p < 0.05, ** p < 0.01 The statistical significances are based on permutational t-test with FDR correction. See Table S4 for effect sizes, their confidence intervals, and exact p-values. Each point represents an individual well. Violin plots with means ± SEM are shown.

Figure 6

Metabolic analysis of JUN-sarcoma cell lines. (A,C) Mitochondrial oxygen consumption in JUN-sarcoma cell lines. Cell lines differed in oxygen consumptions in state R (permutational ANOVA: F_3,32 = 4, p = 0.015), L (permutational ANOVA: F_3,32 = 3.2, p < 0.032), and E (permutational ANOVA: F_3,32 = 6, p = 0.0021), and also differed in E-R capacity (permutational ANOVA: F_3,32 = 9.5, p < 0.001) but not R-L capacity (permutational ANOVA: F_3,32 = 1.3, p = 0.3). The JUN-2fos-3 cell line displayed significantly higher oxygen consumption in the states R, L, and E compared to the least transformed non-invasive, non-motile JUN-2 cells. The spare respiratory capacity (excess E-R capacity) was higher in both invasive and motile cell lines, JUN-3 and JUN-2fos-3, than in cells with limited motility and invasiveness, JUN-1 and JUN-2. (B) The production of L-lactate differed among JUN-sarcoma cell lines (µmol/1000 cells; (permutational ANOVA: F_3,20 = 25, p < 0.001). The JUN-2 cell line had a relatively high capacity of oxidative phosphorylation and the lowest production of lactate, suggesting that ATP is generated especially aerobically through the respiratory chain. (D) The consumption of glucose differed among JUN-sarcoma cell lines (µmol/1000 cells; (permutational ANOVA: F_3,23 = 9.9, p < 0.001). Both the invasive cell lines JUN-2fos-3 and JUN-3 combined relatively high oxphos parameters with the highest production of lactate and consumption of glucose, taking advantage of both pathways of energy production. * p < 0.05, ** p < 0.01, *** p < 0.001. The statistical significances are based on permutational t-test with FDR correction. See Table S5 for effect sizes, their confidence intervals, and exact p-values. Each point represents an individual experiment. Violin plots with means ± SEM are shown. (ROUT: (R)—resting respiration of intact cells, LEAK (L)—oxygen consumption essential for compensation for the proton leakage, ETS cap (E)—uncoupled respiration, i.e., maximum capacity of the electron-transporting system. R-L—ATP-linked oxygen consumption, E-R—spare respiratory capacity).

Figure 7

The gene set enrichment analysis of JUN-sarcoma cell lines. Median log (fold change) with 95% confidence interval in pathways sorted by fold change. Downregulated pathways are shown in blue, whereas upregulated pathways are shown in red colours. Analysis revealed that the downregulated genes are dominated by extracellular matrix and cell adhesion, as well as antigen presentation, whereas among upregulated pathways, those related to cell cycle regulation and DNA replication are particularly frequent (see Table S7 The gene set enrichment analysis for complete list of genes in each pathway).

Figure 8

Pharmacological inhibition of Ccl8 - Ccr5 signalling entails a significant decline in invasive capacity. Cells were treated with Bindarit (Ccl8 inhibitor, 250 µM), Maraviroc (Ccr5 inhibitor, 10 µM), or their combination and their invasiveness was analysed by xCelligence motility assay (A) and by the invasion of multicell tumour spheroids of JUN-sarcoma cell lines embedded into type I collagen identically to or. (B) Linear model revealed that both Bindarit (ẞ = −0.89 (95% CI: −1.15, −0.69), p < 0.001) and Maraviroc (ẞ = −0.54 (−0.82, −0.26), p < 0.001) significantly decreased the motility of JUN-3 cells, but we were not able to detect significant effect of their interaction (ẞ = −0.48 (−1.04, 0.08), p = 0.072). *** p < 0.001. The statistical significances are based on permutational t-test with FDR correction. See Table S8 for effect sizes, their confidence intervals, and exact p-values. Each point represents an individual well. Violin plots with means ± SEM are shown.