TPM1 mediates inflammation downstream of TREM2 via the PKA/CREB signaling pathway

TPM1 regulates inflammation via the PKA/CREB pathway in vitro

We have recently shown that systemic TPM1 induces endogenous TPM1 upregulation and pro-inflammatory responses in the aging retina [7]. However, it remains unclear how endogenous TPM1 triggers neuroinflammation in the retina. To this end, we transfected BV2 cells, a murine microglia cell line, with TPM1 plasmids (Additional file 1: Fig. S1A). After 24 h incubation, we found that TPM1 overexpression activated microglia by increasing CD68 immunoreactivity in microglia (Fig. 1A; Additional file 1: Fig. S1B, C), and releasing more pro-inflammatory cytokines (TNF-α, IL-1β and IL-6) and chemokines (COX-2 and iNOS) relative to BV2 controls (Fig. 1B–F), suggesting that TPM1 overexpression induces inflammation. In addition, we found that TPM1 overexpression significantly downregulated the phosphorylation of both PKA and CREB in BV2 cells (Fig. 1G–L), suggesting that TPM1 potentially regulates inflammation via the PKA/CREB signaling pathway. Indeed, we observed that TPM1 overexpression failed to downregulate phosphorylated CREB (p-CREB) in BV2 cells after treatment with dbcAMP, an activator of PKA (Fig. 1M–R), confirming that TPM1 mediates neuroinflammation by regulating the PKA/CREB signaling pathway. Consistently, the PKA/CREB signaling pathway is previously reported to regulate neuroinflammation in the CNS [15,16,17].

Fig. 1figure 1

TPM1 induces inflammation by regulating the PKA/CREB pathway in BV2 cells. A Immunostaining of BV2 cells with antibodies against Iba-1 and CD68 after treatment with TPM1 plasmid or control plasmid. Arrowheads show the colocalization of Iba-1-positive microglial cells with CD68 signal. Scale bar, 20 µm. BF qPCR analysis of TNF-α, IL-1β, IL-6, COX-2 and iNOS in BV2 cells treated with TPM1 plasmid or control plasmid. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to TPM1 plasmid, *p < 0.05, ****p < 0.001). Five independent experiments were performed. GL Western blot analysis (G) and quantification of p-PKA, PKA, p-CREB, CREB and TPM1 (HL) in BV2 cells following TPM1 plasmid or control plasmid treatment. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to TPM1 plasmid, *p < 0.05, **p < 0.01, ***p < 0.001). Five independent experiments were performed. MR Western blot analysis (M) and quantification of p-PKA, PKA, p-CREB, CREB and TPM1 (NR) in BV2 cells treated by TPM1 plasmid or control plasmid and followed by dbcAMP treatment. Data are presented as mean ± SEM analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to TPM1 plasmid, *p < 0.05, **p < 0.01, ***p < 0.001 ****p < 0.001)

To further study the role of TPM1 in regulating neuroinflammation, we treated BV2 cells with lipopolysaccharides (LPS), which is an endotoxin and is responsible for various chronic peripheral neuroinflammation [18]. We found that LPS significantly stimulated endogenous TPM1 expression (Additional file 2: Fig. S2A) and the release of pro-inflammatory cytokines and chemokines in BV2 cells (Additional file 2: Fig. S2B–F), which were then counteracted by TPM1 knockdown using TPM1-specific siRNA (Additional file 1: Fig. S1D–F, Additional file 2: Fig. S2A–F), suggesting that TPM1 is involved in LPS-induced inflammation. Indeed, we found that LPS treatment increased the number of activated microglia with CD68 immunoreactivity, which was then suppressed by TPM1 knockdown (Additional file 1: Fig. S1G, H; Additional file 2: Fig. S2G), confirming the involvement of TPM1 in LPS-mediated inflammation. In addition, we observed that LPS treatment deactivated the PKA/CREB signaling pathway in BV2 cells by downregulating phosphorylation of both PKA and CREB (Additional file 2: Fig. S2H–M). However, the downregulation of p-CREB induced by LPS was reversed following TPM1 knockdown in BV2 cells (Additional file 2: Fig. S2H, K and M), suggesting that TPM1 is involved in LPS-induced neuroinflammation through the regulation of CREB phosphorylation. Furthermore, we found that TPM1 knockdown failed to upregulate p-CREB in LPS-treated BV2 cells after application of H89, an inhibitor of PKA (Additional file 1: Fig. S1I, J; Additional file 2: Fig. S2N–Q), confirming that PKA is involved in TPM1-regulated inflammation. Taken together, these data indicate that TPM1 potentially regulates neuroinflammation via the PKA/CREB signaling pathway.

TPM1 knockdown reduces LPS-induced inflammation and function decline via the PKA/CREB pathway in vivo

To confirm our in vitro findings, we further investigated the role of TPM1 in vivo. Similarly, we observed that intraperitoneal injection of LPS induced TPM1 upregulation (Additional file 3: Fig. S3A; Fig. 2K, P), and the release of cytokines including TNF-α, IL-1β and IL-6 (Fig. 2A–C) and chemokines including COX-2 and iNOS (Fig. 2D, E) in the retina of C57BL/6J (WT) mice. After TPM1 knockdown by intravitreal administration of siTPM1, however, we found that LPS-induced release of the cytokines and chemokines was reversed in the retina of WT mice (Additional file 3: Fig. S3A, Fig. 2A–E), indicating that TPM1 knockdown counteracts LPS-induced neuroinflammation in vivo. Moreover, we found that TPM1 knockdown significantly increased the expression of anti-inflammatory cytokine IL-10 in LPS-treated WT retinas (Additional file 3: Fig. S3B), further suggesting that TPM1 mediates LPS-induced inflammation. Microglia in the WT mouse retina are positive for Iba-1 but negative for CD68, and their dendrites are restricted within the outer plexiform layer (OPL), the inner plexiform layer (IPL), or the ganglion cell layer (GCL) [3]. However, we found that LPS treatment increased the number of CD68-positive microglia, especially in the IPL of the WT mouse retina (Additional file 3: Fig. S3C–E), and some activated microglia migrated into the ONL from their normal localization within the OPL compared to control retinas (Fig. 2F, G). Interestingly, TPM1 knockdown decreased the incidence of microglia migration into the ONL induced by LPS treatment (Fig. 2F, G), even though the numbers of activated microglia remained unchanged compared to LPS-treated WT retinas (Additional file 3: Fig. S3C–E). Similarly, we observed that LPS treatment increased GFAP immunoreactivity (Fig. 2H, Additional file 3: Fig. S3F), a marker for astrocytes and activated Müller cells, in the retina of WT mice. The dendritic processes of astrocytes in the WT mouse retina are entirely restricted to the nerve fiber layer (NFL) [19]. After LPS treatment, however, we observed that some processes of activated astrocytes extended well beyond their normal strata within the NFL to the IPL (Fig. 2H). However, TPM1 knockdown partially restored the appropriate dendritic localization of astrocytes in the NFL of LPS-treated WT retinas (Fig. 2H, Additional file 3: Fig. S3F). Functionally, we found that LPS treatment significantly decreased the amplitudes of a- and b-waves of electroretinogram (ERG) under both scotopic and photopic conditions compared to WT controls (Fig. 2I, J), even though latency remained unchanged (Additional file 3: Fig. S3G, H). However, TPM1 knockdown partially rescued LPS-induced function declines on WT mice (Fig. 2I, J). Similarly, we found that LPS treatment downregulated p-PKA and p-CREB in WT retinas (Fig. 2K–O), indicating the deactivation of the PKA/CREB signaling pathway. Following TPM1 knockdown (Fig. 2K, P), however, we found that the LPS-induced downregulation of p-CREB was reversed (Fig. 2K, N), suggesting that TPM1 potentially regulates LPS-induced neuroinflammation via the PKA/CREB signaling pathway.

Fig. 2figure 2

TPM1 knockdown reduces LPS-induced inflammation and function decline via the PKA/CREB pathway in vivo. AE qPCR analysis of TNF-α, IL-1β, IL-6, COX-2 and iNOS in C57BL/6J mice after treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001), n = 4 mice in each group. F, G Representative confocal images of microglia from retinal sections stained with antibodies against Iba-1 and CD68 (F) and quantification of microglial cell density in the ONL (G) of C57BL/6J mice following treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to LPS + siCTR, *p < 0.05), n = 5 mice in each group. Arrowheads show the migration of activated microglia to the ONL. The boxed regions are highly magnified at the bottom showing migrated microglia in the ONL. Scale bars, 20 µm. H Retina sections stained with GFAP antibody. Arrowheads show activated astrocytes and Müller cells. Scale bars, 20 µm. I, J ERG recording in C57BL/6J mice after treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001; LPS + siCTR vs. LPS + siTPM1-1, ##p < 0.01), n = 10 mice in each group. KP Western blot analysis (K) and quantification of p-PKA, PKA, p-CREB, CREB and TPM1 (LP) in C57BL/6J mice following treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001). n = 4 mice in each group

Collectively, we demonstrate that TPM1 is potentially involved in LPS-induced neuroinflammation, and TPM1 knockdown ameliorates glial reactivity, neuroinflammation and function decline in WT retinas.

TPM1 knockdown exacerbates inflammation in the TREM2−/− mouse retina

Triggering receptor expressed on myeloid cells 2 (TREM2), which is exclusively expressed by microglia, regulates neuroinflammation in the brain [20,21,22,23]. Interestingly, we observed that TREM2 deficiency upregulated TPM1 (Additional file 4: Fig. S4A, B), and TPM1 knockdown did not alter TREM2 expression level in LPS-treated in WT mouse retina (Additional file 4: Fig. S4C, D), suggesting that TPM1 might mediate inflammation downstream of TREM2. To further examine the relationship between TPM1 and TREM2 in inflammation process, we used TREM2 knockout mice. We found that LPS treatment remarkably increased the transcriptomic levels of TPM1, TNF-α, IL-1β, IL-6, COX-2 and iNOS in TREM2−/− mouse retina compared to PBS-treated TREM2−/− controls (Additional file 4: Fig. S4E, Fig. 3A–E). Surprisingly, we found that TPM1 knockdown further increased the expression levels of TNF-α, IL-1β, IL-6, COX-2 and iNOS (Additional file 4: Fig. S4E, Fig. 3A–E), and enhanced microglia and astrocyte activation in LPS-treated TREM2−/− mouse retinas relative to siCTR- and LPS-treated TREM2−/− controls (Fig. 3F–J). We found that TPM1 knockdown increased the numbers of activated microglia (Iba-1+CD68+) in the OPL of LPS-treated TREM2−/− mouse retinas relative to siCTR- and LPS-treated group (Fig. 3F–H). Similarly, we observed that TPM1 knockdown promoted dendritic migration of more activated microglia into the ONL from the OPL in LPS-treated TREM2−/− mouse retinas compared to LPS-treated TREM2−/− control groups (Fig. 3I, Additional file 4: Fig. S4F). Moreover, we observed that LPS treatment increased GFAP immunoreactivity and the incidences of dendritic processes of activated astrocytes into the IPL from their normal NFL strata in TREM2−/− mouse retinas compared to PBS-treated TREM2−/− control mouse retinas (Fig. 3J, Additional file 4: Fig. S4G). Remarkably, we found that TPM1 knockdown further elevated GFAP immunoreactivity and astrocyte activation in LPS-treated TREM2−/− mice compared to LPS-treated TREM2−/− control mice (Fig. 3J, Additional file 4: Fig. S4G). Functionally, we observed decreases in the amplitudes of a- and b-waves of ERG in LPS-treated TREM2−/− mice under both scotopic and photopic conditions (Fig. 3K, L). However, TPM1 knockdown further reduced a- and b-wave amplitudes of ERG in LPS-treated TREM2−/− mice (Fig. 3K, L). Also, we found that TPM1 knockdown increased latency for a-waves under scotopic condition at 1 and 3 cd s/m2 light intensities in LPS-treated TREM2−/− mice compared to PBS-treated TREM2−/− mice (Additional file 4: Fig. S4H, I). Taken together, these results indicate that TPM1 knockdown aggravates LPS-induced inflammation and function decline in the TREM2−/− mouse retina.

Fig. 3figure 3

TPM1 knockdown exacerbates inflammation in the TREM2−/− mouse retina. AE qPCR analysis of TNF-α, IL-1β, IL-6, COX-2 and iNOS in TREM2−/− mice after treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001), n = 7 mice in each group. FH, Retina whole-mounts stained with antibodies against Iba-1 and CD68 (F) and quantification of numbers of Iba-1+ (G) and of Iba-1+CD68+ microglia (H) in the IPL and OPL of TREM2−/− retinas after treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001). n = 3 in PBS group, n = 4 in LPS + siCTR/siTPM1-1 group. Arrowheads show microglia, scale bar, 20 µm. I Retina sections stained with Iba-1 and CD68 antibodies in TREM2−/− retinas following treatments with PBS, or with LPS and siTPM1-1 or siCTR. Arrowheads show the migration of activated microglia to the ONL. The boxed regions are highly magnified at the bottom showing migrated microglia in the ONL. Scale bars, 20 µm. J Retina sections stained with an antibody against GFAP in TREM2−/− retinas after treatments with PBS, or with LPS and siTPM1-1 or siCTR. Arrowheads show activated astrocytes and Müller cells. Scale bars, 20 µm. K, L Scotopic and photopic electroretinogram (ERG) recordings on TREM2−/− mice. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS, **p < 0.01, ***p < 0.001, ****p < 0.001; LPS + siCTR vs. LPS + siTPM1-1, #p < 0.05, ##p < 0.01). n = 13, 10, 11 mice in PBS, LPS + siCTR, and LPS + siTPM1-1 groups, respectively

TPM1 knockdown changes glial cell-specific transcriptomes in the TREM2−/− mouse retina

We showed that TPM1 knockdown suppressed LPS-induced inflammation in the WT mouse retina but aggravated LPS-induced inflammation in the TREM2−/− mouse retina. To decipher the differential roles of TPM1, we performed RNA sequences and subsequent transcriptome-wide analysis of genes and pathways in the TREM2−/− or WT mouse retina with TPM1 knockdown. We found that 440 genes were upregulated and 91 genes were downregulated in siTPM1- and LPS-treated WT retinas compared to siCTR- and LPS-treated WT retinas (Fig. 4A). Meanwhile, 953 upregulated genes and 92 downregulated genes were identified in siTPM1- and LPS-treated TREM2−/− mouse retinas compared to siCTR- and LPS-treated TREM2−/− mouse retinas (Fig. 4B), suggesting that TPM1 knockdown induces robust gene expression changes in LPS-treated TREM2−/− retinas. Indeed, we found that TPM1 knockdown promoted upregulation of more genes in LPS-treated TREM2−/− retinas than in LPS-treated WT retinas (Fig. 4C). By analyzing differentially expressed genes (DEGs), we found that the genes associated with microglia and astrocytes were remarkably upregulated in LPS-treated WT retinas (Additional file 5: Fig. S5A, B, and E). However, we found that TPM1 knockdown decreased expression of the genes that were associated with M1 microglia (Cxcl10) and A1 astrocytes (Gbp2) (Additional file 5: Fig. S5A, B, and E), but increased expression of the genes related to M2 microglia (Arg1 and Temem119) and A2 astrocytes (Cd109 and Cd14) in LPS-treated WT retinas compared to siCTR- and LPS-treated WT control retinas (Additional file 5: Fig. S5A, C, D, F, and G), suggesting that TPM1 knockdown suppressed LPS-induced glial reactivity in WT retinas, which is consistent with our morphological observation (Fig. 2F–H). Similarly, we found that LPS treatment elevated the expression of DEGs which were associated with both microglia (M1 and M2) and astrocytes (A1 and A2) in TREM2−/− mice (Fig. 4D–F, J, and L–N). Interestingly, we found that TPM1 knockdown further increased expression of the genes that were related to M1—(Ccl7, Ccl2 and Cxcl10) and M2—(Arg1 and Tmem119) microglia, and A1—(C3, Serping1, Srgn, Gbp2, Gfap) and A2—(Tgm1, Cd109, Ptgs2, Emp1 and B3gnt5) astrocytes in LPS-treated TREM2−/− retinas compared to either siCTR- and LPS-treated TREM2−/− control retinas or PBS-treated TREM2−/− control retinas (Fig. 4D–S), suggesting that TPM1 knockdown further enhances glial cell reactivity in LPS-treated TREM2−/− retinas, which is consistent with morphological alterations in microglia and astrocytes (Fig. 3F–J). Together, these results indicate that TPM1 knockdown increasingly induces transcriptomic alterations in microglia and astrocytes in LPS-treated TREM2−/− retinas.

Fig. 4figure 4

TPM1 knockdown changes transcriptome associated with glial cells in the TREM2−/− mouse retina. AC Volcano plots showing differentially expressed genes (DEGs) in WT or TREM2−/− mice after treatments with LPS and siTPM1-1 or siCTR. Red dots indicate up-regulated genes and green dots show down-regulated genes (n = 3 mice in each group). D A heatmap showing DEGs associated with microglia and astrocytes in TREM2−/− mice following treatments with PBS, or with LPS and siTPM1-1 or siCTR. ES The expression of DEGs associated with M1 (EG) or M2 microglia (H, I) and with A1 (JN) or A2 astrocytes (OS) in TREM2−/− mice after treatments with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001). n = 3 mice in each group. FPKM fragments per kilobase of transcript per million fragments mapped

TPM1-induced inflammation is microglia-dependent

To specifically investigate the role of microglia in TPM1-mediated inflammation, we genetically eliminated microglia from the retina. To this end, we crossed CX3CR1CreER transgenic mice expressing tamoxifen-inducible Cre recombinase under the control of the CX3C chemokine receptor 1 promoter with Rosa26iDTR mice expressing Cre-inducible diphtheria toxin receptor (iDTR), and generated a new CX3CR1CreER:Rosa26iDTR mouse line. We conditionally depleted microglia from CX3CR1CreER:Rosa26iDTR mice by administration of tamoxifen (TAM) and subsequent administration of DT 28–30 days later (Fig. 5A). We found that TPM1 was significantly elevated in the microglia-depleted retina of CX3CR1CreER:Rosa26iDTR mice compared to CX3CR1CreER:Rosa26iDTR control mice (Fig. 5B, C, and F), suggesting that microglia depletion triggers TPM1 upregulation. Moreover, we found that microglia depletion increased GFAP immunoreactivity (Fig. 5E), which was confirmed by WB analysis (Fig. 5B, D). Some processes of activated astrocytes migrated into the INL, even onto the ONL in microglia-depleted CX3CR1CreER:Rosa26iDTR retinas (Fig. 5E). Meanwhile, we found that microglia deletion induced the upregulation of TNF-α, IL-1β, IL-6, COX-2 and iNOS (Fig. 5G–K). Consistently, previous studies reported that microglia depletion elevated inflammatory signatures and astrocytes activation [24,25,26]. Together these results indicate that microglia deletion induces TPM1 upregulation and inflammation. Surprisingly, we observed that TPM1 knockdown did not reduce the expression levels of these pro-inflammatory cytokines elicited by microglia deletion in CX3CR1CreER:Rosa26iDTR retinas (Fig. 5F–K), suggesting that TPM1 regulates inflammation in a microglia-dependent manner.

Fig. 5figure 5

TPM1-induced inflammation is microglia-dependent. A Retinal whole-mounts from CX3CR1CreER:Rosa26iDTR mice following tamoxifen (TAM) and subsequent diphtheria toxin (DT) treatments were stained with an antibody against Iba-1. Scale bar, 500 µm. BD Western blot analysis (B) and quantification of TPM1 and GFAP (C, D) in microglia-depleted CX3CR1CreER:Rosa26iDTR mouse retinas following TAM and DT treatment. Data are presented as mean ± SEM and analyzed by unpaired two-tailed Student’s t test (control vs. DT treatment, **p < 0.01). n = 4 mice in each group. E Retina sections stained with an antibody against GFAP in microglia-depleted CX3CR1CreER:Rosa26iDTR mice following TAM and DT treatment. Arrowheads show activated astrocytes and Müller cells. Scale bars, 20 µm. FK qPCR analysis of TPM1, TNF-α, IL-1β, IL-6, COX-2 and iNOS in microglia-depleted CX3CR1CreER:Rosa26iDTR mice after siCTR or siTPM1-1 treatment. Data are presented as mean ± SEM and analyzed one-way ANOVA with Tukey’s multiple comparison test (compared to Control or DT-siCTR, *p < 0.05, **p < 0.01). n = 4, 5,5 mice in Control, DT-siCTR and DT-siTPM1-1 groups, respectively

TPM1 knockdown elicits inflammation-related transcriptomic alterations and cell apoptosis in the TREM2−/− mouse retina

Ingenuity pathway analysis of DEGs revealed that most of DEGs were strongly associated with inflammation process, including phagosome formation, neuroinflammation signaling pathway in pairwise comparisons between LPS-treated WT with or without TPM1 knockdown (Fig. 6A), and between LPS-treated TREM2−/− mice with or without TPM1 knockdown (Fig. 6B). Among these DEGs, we found that Myl1, Oprk1, Myh6, Myo10, Ptgdr2, which are related to phagocytosis, were downregulated in siTPM1- and LPS-treated WT retina but upregulated in siCTR- and LPS-treated WT control retinas (Additional file 6: Fig. S6A), suggesting that TPM1 knockdown boosts phagocytic activity suppressed by LPS in WT mice. Interestingly, we found that LPS treatment triggered overexpression of phagosome formation related genes in TREM2−/− mice (Fig. 6C), including C3ar1, Ccr7, Itgb2, Ccr1, P2ry6, Itgax, Fgr, Rac2, Hck, Clec7a, and those genes were further elevated in siTPM1- and LPS-treated TREM2−/− mice, suggesting that TPM1 knockdown results in the enhancement of LPS-induced autophagy in TREM2−/− mice. Among induced DEGs that were related to the neuroinflammation signaling pathway, we found that LPS significantly elevated expressions of pro-inflammatory cytokines including Cxcl10, CD40 and CD86 in WT mice compared to WT control mice (Additional file 6: Fig. S6B). However, we found that TPM1 knockdown remarkably inhibited Cxcl10 (Additional file 6: Fig. S6B) and improved anti-inflammatory cytokines including Csf1r, Irak4 and Trem2 in LPS-treated WT compared to siCTR- and LPS-treated WT control mice (Additional file 6: Fig. S6B), suggesting that TPM1 knockdown inhibits LPS-induced neuroinflammation in WT mice, which is consistent with our in vitro (Additional file 2: Fig. S2A–G) and in vivo (Fig. 2A–H) results. Furthermore, we found that LPS significantly increased inflammatory mediators including Cxcl10, Rac2, Cybb, Casp1, Tlr13, Ncf1 and Tlr7 in TREM2−/− mice (Fig. 6D), and those genes were further elevated in LPS-treated TREM2−/− mice following siTPM1 treatment, indicating that TPM1 knockdown leads to the enhancement of LPS-induced neuroinflammation in TREM2−/− mice, which is consistent with our in vivo results (Fig. 3A–J).

Fig. 6figure 6

TPM1 knockdown elicits inflammation-related transcriptomic alterations and cell death in the TREM2−/− retina. A, B Ingenuity Pathway Analysis (IPA) of top 20 canonical pathways in WT (A) or TREM2−/− mice (B) after treatments with LPS and siTPM1-1 or siCTR. p < 0.001. C, D DEGs in the phagosome formation pathway (C) and neuroinflammation signaling pathway (D) in TREM2−/− mice after treatments with LPS and siTPM1-1 or siCTR. E, F, TUNEL staining (E) and quantification of TUNEL-positive cells (F) in TREM2−/− mice following treatments with PBS, or with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, **p < 0.01, ***p < 0.001). n = 4 mice in each group. G, H qPCR analysis of Bax and Caspase-3 in TREM2−/− mouse retinas after treatments with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, ***p < 0.001, ****p < 0.001). n = 5 mice in each group

To examine whether the increased inflammation induced by TPM1 knockdown resulted in more cell apoptosis in LPS-treated TREM2−/− mice, we performed TUNEL staining. We observed that LPS significantly increased the numbers of TUNEL-positive cells in siCTR-treated TREM2−/− retinas (3.73-fold) compared to PBS-treated TREM2−/− control retinas (Fig. 6E, F). Interestingly, we found that TPM1 knockdown caused more TUNEL-positive cells (1.98-fold) in LPS-treated TREM2−/− retinas than in siCTR- and LPS-treated TREM2−/− control retinas (Fig. 6E, F), suggesting that TPM1 knockdown aggravates cell apoptosis in LPS-treated TREM2−/− mice. Mechanistically, we found that LPS treatment significantly increased mRNA levels of Bax (65.2%) and Caspase-3 (70.66%), which are associated with cell apoptosis, in siCTR-treated TREM2−/− retinas compared to PBS-treated TREM2−/− control retinas (Fig. 6G, H). TPM1 knockdown further increased expression of Bax (59.2%) and Caspase-3 (115.9%) in LPS-treated TREM2−/− retinas compared to siCTR- and LPS-treated TREM2−/− mouse retinas (Fig. 6G, H).

Collectively, these results suggest that TPM1 knockdown promotes additional increase in neuroinflammation and more cell apoptosis in TREM2-deficient mouse retinas following LPS treatment.

TPM1 knockdown dysregulates the PKA/CREB signaling pathway in the TREM2−/− mouse retina

Ingenuity pathway analysis of DEGs revealed that CREB signaling in neurons was involved in LPS-treated WT or TREM2−/− mouse retinas following TPM1 knockdown (Fig. 7A, B). Among induced genes associated with the CREB signaling pathway, Cx3cr1, Mc4r, Ptger3, Hrh1, Tacr2, Creb3l2 and Creb3l4, were downregulated in siCTR- and LPS-treated WT retinas but upregulated in siTPM1- and LPS-treated WT retinas (Fig. 7A), suggesting that TPM1 knockdown reverses LPS-induced inhibition of the CREB signaling pathway. Indeed, our in vitro (Additional file 2: Fig. S2H–Q) and in vivo (Fig. 2K–P) results also verified that TPM1 knockdown counteracted the decline of p-CREB induced by LPS treatment. Interestingly, we found that the vast majority of genes that are associated with the CREB signaling pathway, such as C3ar1, Ccr1, Ccr7, P2ry6, C5ar1, Fgf2 and Gpr65, were overexpressed in LPS-treated TREM2−/− mouse retinas relative to PBS-treated TREM2−/− control retinas (Fig. 7B), suggesting that LPS activates the CREB signaling pathway in TREM2−/− mice. Evidently, western blot analysis revealed elevated production of p-CREB in LPS-treated TREM2−/− retinas (Fig. 7C–E). TPM1 knockdown further increased the expression level of the genes associated with the CREB signaling pathway in LPS-treated TREM2−/− retinas compared to siCTR- and LPS-treated TREM2−/− mouse retinas (Fig. 7B), suggesting that TPM1 knockdown further enhances the activation of the CREB signaling pathway induced by LPS in TREM2−/− mice. Furthermore, we observed that LPS administration significantly elevated the expression of p-CREB but not p-PKA and PKA in TREM2−/− retinas (Fig. 7F–J), indicating the dysregulation of the PKA/CREB signaling pathway in LPS-treated TREM2−/− mice. After TPM1 knockdown, p-CREB also showed an increase trend in LPS-treated TREM2−/− retinas relative to siCTR- and LPS-treated TREM2−/− control retinas (Fig. 7F–K), even though p-CREB expression level was not statistically significant between two groups, suggesting that TPM1 knockdown dysregulates the PKA/CREB signaling pathway in LPS-treated TREM2−/− retinas.

Fig. 7figure 7

TPM1 knockdown dysregulates the PKA/CREB pathway in the TREM2−/− retina. A, B DEGs related to the CREB signaling in neurons in WT (A) or TREM2−/− mice (B) following treatments with PBS, or with LPS and siTPM1-1 or siCTR. CE Western bot analysis (C) and quantification of p-CREB and CREB (D, E) in WT or TREM2−/− mice following LPS or PBS treatment. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to LPS + TREM2−/−, *p < 0.05, **p < 0.01; PBS + WT vs. LPS + WT, *p < 0.05). n = 4 mice in each group. FK Western blot analysis (F) and quantification of p-PKA, PKA, p-CREB, CREB and TPM1 (GK) in TREM2−/− mice after treatments with LPS and siTPM1-1 or siCTR. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to PBS or LPS + siCTR, *p < 0.05, ***p < 0.001). n = 5 or 6 mice in each group. LQ qPCR analysis of TPM1, TNF-α, IL-1β, IL-6, COX-2 and iNOS in TREM2−/− mice after treatments with LPS, siTPM1-1 and 666-15, a potent and selective CREB inhibitor. Data are presented as mean ± SEM and analyzed by one-way ANOVA with Tukey’s multiple comparison test (compared to LPS + siCTR or LPS + siTPM1-1, *p < 0.05,

留言 (0)

沒有登入
gif