Role of RPTPβ/ζ in neuroinflammation and microglia-neuron communication

Abstract

Pleiotrophin (PTN) is a cytokine that is upregulated in different neuroinflammatory disorders. Using mice with transgenic PTN overexpression in the brain (Ptn-Tg), we have found a positive correlation between iNos and Tnfα mRNA and Ptn mRNA levels in the prefrontal cortex (PFC) of LPS-treated mice. PTN is an inhibitor of Receptor Protein Tyrosine Phosphatase (RPTP) β/ζ, which is mainly expressed in the central nervous system. We aimed to test if RPTPβ/ζ is involved in the modulation of neuroinflammatory responses using specific inhibitors of RPTPβ/ζ (MY10 and MY33-3). Treatment with MY10 potentiated LPS-induced microglial responses in the mouse PFC. Surprisingly, MY10 caused a decrease in LPS-induced NF-κB p65 expression, suggesting that RPTPβ/ζ may be involved in a novel mechanism of potentiation of microglial activation independent of the NF-κB p65 pathway. MY33-3 and MY10 limited LPS-induced nitrites production and iNos increases in BV2 microglial cells. SH-SY5Y neuronal cells were treated with the conditioned media from MY10/LPS-treated BV2 cells. Conditioned media from non-stimulated and from LPS-stimulated BV2 cells increased the viability of SH-SY5Y cultures. RPTPβ/ζ inhibition in microglial cells disrupted this neurotrophic effect of microglia, suggesting that RPTPβ/ζ plays a role in the neurotrophic phenotype of microglia and in microglia-neuron communication.

Figure 1

Effects of LPS on STAT3 phosphorylation and gene expression in the PFC of Wt and Ptn-Tg mice. Western blot assays using phospho-STAT3 and STAT3 antisera of PFC proteins extracts prepared 16 h after lipopolysaccharide (LPS) or saline (Sal) treatment (a). Optical density measurements did not reveal relevant differences between genotypes (b). Real-time PCR analyses of Ptn mRNA in PFC from Saline-treated (black bars) and LPS-treated (grey bars) Wt and Ptn-Tg mice (c). Real-time PCR analyses of iNos in PFC from Saline-treated and LPS-treated Wt and Ptn-Tg mice (d). Correlation between the mRNA of iNos and Ptn in PFC of LPS-treated Wt mice (e) and of LPS-treated Ptn-Tg mice (f). Real-time PCR analyses of Tnfα mRNA in PFC from Saline-treated and LPS-treated Wt and Ptn-Tg mice (g). Correlation between the mRNA of Tnfα and Ptn in PFC of LPS-treated Wt mice (h) and of LPS-treated Ptn-Tg mice (i). Pearson correlation coefficient (r) and p value are shown in all correlation graphs. Fold changes in the qPCR analysis are calculated relative to the control, Wt Sal. *P < 0.05 LPS vs. Sal. **P < 0.01 LPS vs. Sal. # P < 0.05 Ptn-Tg vs. Wt. ^##P < 0.01 Ptn-Tg vs. Wt.

Figure 2

Effects of MY10 and LPS on astrocytosis in the mouse PFC. Photomicrographs are representative from GFAP-immunostained PFC sections of control (vehicle + saline) (a), LPS-treated (b) or MY10 + LPS-treated animals (c). Higher magnifications images in the lower right corner of every representative picture show hypertrophic and densely stained astrocytes in LPS-treated mice.

Figure 3

Effects of MY10 and LPS on microgliosis in the mouse PFC. Photomicrographs are representative from Iba1-immunostained PFC sections of control (vehicle + saline) (a), LPS-treated (b) or MY10 + LPS-treated animals (c). Higher magnifications images in the lower right corner of every representative picture show hypertrophic and densely stained microglia in LPS-treated mice. Graphs represent quantification of data (mean ± S.E.M) obtained from the counts of Iba-1-positive cells (d), total marked area (e), cell area (f) and circularity index (g). *P < 0.05; **P < 0.01; ***P < 0.001vs. control (C). ^#P < 0.05 MY10 + LPS vs. LPS.

Figure 4

Effects of MY10 and LPS on NF-κB p65 expression in the mouse PFC. Photomicrographs are representative from DAPI- (blue) and NF-κB p65-immunostained (green) PFC sections of control (vehicle + saline) (a), LPS-treated (b) or MY10 + LPS-treated animals (c).

Figure 5

Effects of inhibition of RPTPβ/ζ on LPS-induced nitrites production and in iNos and Tnfα mRNA levels in BV2 microglial cells. Cells were treated with the indicated concentrations of MY10 or MY33-3 (0.1, 1.0 or 10 μM) and/or with LPS (1.0 μg/ml) for 24 h. Nitrites accumulation in the BV2 culture medium was analyzed using the Griess reactive (a,d). Real-time PCR analyses of iNos (b,e) and Tnfα mRNA (c,f). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 vs. control non-stimulated cells (without LPS and MY10 or MY33-3, black bar). ^#P < 0.05; ^####P < 0.0001 vs. LPS-stimulated cells (without MY10 or MY33-3, dark grey bar).

Figure 6

Pharmacological inhibition of RPTPβ/ζ prevents the neurotrophic phenotype of microglia. MY10 significantly decreases the neurotrophic effects of media from BV2 cells when SH-SY5Y neurons are incubated with the media from MY10/LPS-treated BV2 cells, measured by the MTT test (a). High concentrations of MY10 decrease the viability of SH-SY5Y neurons in the MTT test (b). Correlation between the nitrites production of LPS-treated BV2 cells and SH-SY5Y cell viability (c). Correlation between the SH-SY5Y cell viability and the nitrites production of 0.1 μM MY10 + LPS-treated BV2 cells (d), 1 μM MY10 + LPS-treated BV2 cells (e) and 10 μM MY10 + LPS-treated BV2 cells (f). Pearson correlation coefficients (r) and p values are shown in all correlation graphs. ***P < 0.001; ****P < 0.0001 vs. non-stimulated BV2 cells conditioned media (without LPS and MY10; dark grey bar). ^#P < 0.05; ^####P < 0.0001 vs. control SH-SY5Y cell just incubated with control media (black bar). Data are shown as % viability considering 100% viability of neurons incubated with control media (black bar).

Figure 7

Inhibition of RPTPβ/ζ differentially modulates LPS-induced microglial responses in vivo and in vitro. Transgenic overexpression of the endogenous RPTPβ/ζ inhibitor PTN in the mouse brain potentiates LPS-induced microgliosis and LPS-induced increases of inflammatory mediators (a). Accordingly, we now demonstrate that systemic administration of the exogenous RPTPβ/ζ inhibitor MY10 potentiates LPS-induced microgliosis in the mouse PFC. In contrast, when acting in isolated microglial cells in vitro, MY10 limits the LPS-induced production of inflammatory mediators such as TNFα, iNOS and NO (b). We hypothesize that the overall potentiation of LPS-induced microgliosis by MY10 in vivo may reflect the actions of the inhibition of RPTPβ/ζ in different cells in the brain. RPTPβ/ζ is also expressed in oligodendrocytes, astrocytes and neurons. Inhibition of the phosphatase activity of RPTPβ/ζ in these cells could alter their communication with microglia (e.g. through changes in the secretome of these cells), potentially impacting microglial cells in different ways (c). This figure was created with BioRender's web-based software—https://biorender.com/.