miR-296-3p, miR-298-5p and their downstream networks are causally involved in the higher resistance of mammalian pancreatic α cells to cytokine-induced apoptosis as compared to β cells

Abstract

Background: The molecular bases of mammalian pancreatic α cells higher resistance than β to proinflammatory cytokines are very poorly defined. MicroRNAs are master regulators of cell networks, but only scanty data are available on their transcriptome in these cells and its alterations in diabetes mellitus.

Results: Through high-throughput real-time PCR, we analyzed the steady state microRNA transcriptome of murine pancreatic α (αTC1-6) and β (βTC1) cells: their comparison demonstrated significant differences. We also characterized the alterations of αTC1-6 cells microRNA transcriptome after treatment with proinflammatory cytokines. We focused our study on two microRNAs, miR-296-3p and miR-298-5p, which were: (1) specifically expressed at steady state in αTC1-6, but not in βTC1 or INS-1 cells; (2) significantly downregulated in αTC1-6 cells after treatment with cytokines in comparison to untreated controls. These microRNAs share more targets than expected by chance and were co-expressed in αTC1-6 during a 6-48 h time course treatment with cytokines. The genes encoding them are physically clustered in the murine and human genome. By exploiting specific microRNA mimics, we demonstrated that experimental upregulation of miR-296-3p and miR-298-5p raised the propensity to apoptosis of transfected and cytokine-treated αTC1-6 cells with respect to αTC1-6 cells, treated with cytokines after transfection with scramble molecules. Both microRNAs control the expression of IGF1Rβ, its downstream targets phospho-IRS-1 and phospho-ERK, and TNFα. Our computational analysis suggests that MAFB (a transcription factor exclusively expressed in pancreatic α cells within adult rodent islets of Langerhans) controls the expression of miR-296-3p and miR-298-5p.

Conclusions: Altogether, high-throughput microRNA profiling, functional analysis with synthetic mimics and molecular characterization of modulated pathways strongly suggest that specific downregulation of miR-296-3p and miR-298-5p, coupled to upregulation of their targets as IGF1Rβ and TNFα, is a major determinant of mammalian pancreatic α cells resistance to apoptosis induction by cytokines.

Figure 1

Apoptosis of αTC1-6 and βTC1 after treatment with cytokines. (A) Annexin V flow cytometric analysis of apoptosis in αTC1-6 treated with IFN-γ, IL-1β, TNF-α for 6, 24, 48 h and in matched untreated controls. Values represent the percentage of apoptotic cells. Data are presented as mean ± S.D. of three independent experiments (n = 3). (B) Microphotographs of βTC1 at steady state (left) and 24 h PT (right). Cell shrinkage and irregular morphology are evident in βTC1 after 24 h of treatment. Representative pictures are shown from three independent experiments (n = 3).

Figure 2

Expression of miR-296-3p and miR-298-5p in αTC1-6 and βTC1. (A) Real-time PCR amplification plot of miR-296-3p and miR-298-5p in αTC1-6 and βTC1 at steady state. Note that both miRNAs are not expressed in βTC1. Plot image is representative of three independent experiments (n = 3). (B) Gene expression fold changes of miR-296-3p and miR-298-5p in αTC1-6 at 6, 12, 24, 48 h PT, relative to matched untreated cells. MiRNAs expression was measured by quantitative real-time PCR. MiR-26a was used as endogenous control. Data are reported as mean ± S.D. of three independent experiments (n = 3). * p-value < 0.05; ** p-value < 0.01 (Student’s t-test, Bonferroni correction).

Figure 3

Upregulation of miR-296-3p and miR-298-5p reduces αTC1-6 resistance to apoptosis induced by cytokines. Annexin V flow cytometric analysis of apoptosis in αTC1-6 transiently transfected with scrambled oligonucleotides (NC, Negative Control), mimics of miR-296-3p, miR-298-5p, a mix of both, and treated with IFN-γ, IL-1β, TNF-α for 6, 24, 48 h. The y-axis represents the percentage of apoptotic cells; the x-axis represents the four experimental conditions [(i)scramble-transfected cells; (ii) cells transfected with mimics of miR-296-3p; (iii) cells transfected with mimics of miR-298-5p; (iv) cells transfected with a mix of both] assayed at the three time-points. Data are presented as mean ± S.D. of three independent experiments (n = 3). All the possible pairwise comparisons were performed among the four different experimental conditions within each time point: significant differences have been assessed through Tukey HSD post-hoc one-way ANOVA test (** p-value < 0.01). At 24 h PT, αTC1-6 transfected with mimics of miR-298-5p show a highly significant increase of the number of apoptotic cells with respect to scramble-transfected control; at the same time point, in αTC1-6 transfected with mimics of both miR-296-3p and miR-298-5p a highly significant increase of the number of apoptotic cells is detected with respect to all the other experimental conditions.

Figure 4

Modulation of miR-296-3p and miR-298-5p alters expression of their targets. Bar graph showing changes in gene expression of a selected set of miR-296-3p and miR-298-5p targets for each of three different experimental conditions: (i) αTC1-6 treated with cytokines with respect to matched untreated control cells at the time points 24 and 48 h; (ii) untreated αTC1-6 transfected with mimics of miR-296-3p with respect to scramble-transfected control cells at the time points 24 and 48 h; (iii) untreated αTC1-6 transfected with mimics of miR-298-5p with respect to scramble-transfected control cells at the time points 24 and 48 h. Data are reported as LOG of 2^^-ΔΔCt values. Hprt and Ppia were used as endogenous controls to normalize real-time PCR data.

Figure 5

Expression of IGF1Rβ and TNFα proteins is regulated by miR-296-3p and miR-298-5p in αTC1-6. (A) Western analysis of IGF1Rβ in (1) untreated αTC1-6 transfected for 24 h with (i) scramble molecules (NC); (ii) mimics of miR-296-3p; (iii) mimics of miR-298-5p; (iv) mimics of both miR-296-3p and miR-298-5p (left); (2) αTC1-6 transfected for 24 h with (i) scramble molecules (NC); (ii) mimics of miR-296-3p; (iii) mimics of miR-298-5p; (iv) mimics of both miR-296-3p and miR-298-5p and treated with cytokines for further 24 h (middle); (3) αTC1-6 treated with cytokines for 24 h and their matched untreated controls (right). (B) Western analysis of TNF-α performed in the same experimental conditions as (A). β-Actin signal was used to normalize the data. Numbers below Actin blots represent fold change expression values relative to matched controls.

Figure 6

Activation of IRS-1 and ERK-1 is under control of miR-296-3p and miR-298-5p in αTC1-6. (A) Western analysis of phospho-IRS-1 (Tyr612) in (1) untreated αTC1-6 transfected for 24 h with (i) scramble molecules (NC); (ii) mimics of miR-296-3p; (iii) mimics of miR-298-5p; (iv) mimics of both miR-296-3p and miR-298-5p (left); (2) αTC1-6 transfected for 24 h with (i) scramble molecules (NC); (ii) mimics of miR-296-3p; (iii) mimics of miR-298-5p; (iv) mimics of both miR-296-3p and miR-298-5p and treated with cytokines for further 24 h (right). (B) Western analysis of phospho-ERK-1/2 (Thr202/Tyr204) performed in the same experimental conditions as in (A). Quantification of immunoblot signals was made by equalizing phospho-specific IRS-1 or Erk1/2 band intensities to total IRS-1 or Erk1/2, respectively. The decrease in phosphorylation was normalized to the basal level of the control and reported in arbitrary units as fold decrease over basal value. Numbers below Actin blots represent fold change expression values relative to matched controls.