p75 neurotrophin receptor regulates tissue fibrosis through inhibition of plasminogen activation via a PDE4/cAMP/PKA pathway

Abstract

Clearance of fibrin through proteolytic degradation is a critical step of matrix remodeling that contributes to tissue repair in a variety of pathological conditions, such as stroke, atherosclerosis, and pulmonary disease. However, the molecular mechanisms that regulate fibrin deposition are not known. Here, we report that the p75 neurotrophin receptor (p75(NTR)), a TNF receptor superfamily member up-regulated after tissue injury, blocks fibrinolysis by down-regulating the serine protease, tissue plasminogen activator (tPA), and up-regulating plasminogen activator inhibitor-1 (PAI-1). We have discovered a new mechanism in which phosphodiesterase PDE4A4/5 interacts with p75(NTR) to enhance cAMP degradation. The p75(NTR)-dependent down-regulation of cAMP results in a decrease in extracellular proteolytic activity. This mechanism is supported in vivo in p75(NTR)-deficient mice, which show increased proteolysis after sciatic nerve injury and lung fibrosis. Our results reveal a novel pathogenic mechanism by which p75(NTR) regulates degradation of cAMP and perpetuates scar formation after injury.

Figure 1.

Fibrin deposition is reduced in the sciatic nerve of p75^NTR−/− mice. Immunohistochemistry for fibrin on uninjured wt (a) and 4 d after sciatic nerve crush injury wt (c) and p75 ^NTR−/− mice (e). Immunohistochemistry for p75^NTR on uninjured wt (b) and 4 d after sciatic nerve crush injury wt (d) and p75 ^NTR−/− mice (f). Representative images are shown from n = 20 wt and n = 20 p75 ^NTR−/− mice. (g) Western blot for p75^NTR and fibrin on sciatic nerve extracts from uninjured wt, and wt and p75 ^NTR−/− mice 3 and 8 d after injury. Myosin serves as loading control. Western blots were performed three times. A representative blot is shown. (h) Quantification of fibrin deposition shows significant decrease for fibrin in p75 ^NTR−/− mice (n = 5), when compared with wt mice (n = 4). Bar graph represents means ± SEM (P < 0.003; by t test). Bar, 25 μm.

Figure 2.

p75^NTR regulates expression of tPA in the sciatic nerve after crush injury. In situ zymography in the presence of plasminogen on wt (a) and p75 ^NTR−/− (b) mice and in the absence of plasminogen (c) or in the presence of plasminogen and tPASTOP (d) in p75 ^NTR−/− mice. Arrows indicate the lytic zone. Double immunofluorescence for tPA (green) or p75^NTR (red) on wt (e and h), p75 ^NTR−/− (f) and p75 ^NTR−/− tPA ^−/− mice (g). Uninjured wt sciatic nerve exhibits minimal proteolytic activity (i) and minimal tPA and p75^NTR immunoreactivity (j). Zymographies have been performed on n = 10 wt and n = 10 p75 ^NTR−/− mice. Representative images are shown. tPA (k) and p75^NTR (l) expression in SCs was verified by double immunofluorescence with an S100 (SC marker) antibody. Arrows indicate double-positive cells (k and l, yellow). The experiment was repeated at two different time points (4 and 8 d after crush injury) in n = 4 mice per genotype per time point and representative images are shown. Bar: 400 μm (a–d, i), 150 μm (e–g, j), 20 μm (h, k, and l).

Figure 3.

Loss of tPA rescues the effects of p75^NTR deficiency in plasminogen activation and fibrin deposition in the sciatic nerve. Increased fibrin deposition in the crushed sciatic nerve of p75 ^NTR−/− tPA ^−/− mice (c), when compared with crushed p75 ^NTR−/− sciatic nerve (b). Wt (a) and tPA ^−/− (d) nerves are used for control. In situ zymography shows lack of proteolytic activity in the crushed p75 ^NTR−/− tPA ^−/− sciatic nerves (n = 5) (g), when compared with crushed p75 ^NTR−/− sciatic nerves (n = 20) (f). Crushed wt (e) and tPA ^−/− (h) nerves are used for control. Fibrin immunostainings and zymographies were performed on n = 5 p75 ^NTR−/− tPA ^−/−, n = 20 p75 ^NTR−/−, n = 20 wt, n = 5 tPA ^−/− mice. Representative images are shown. (i) Quantification of proteolytic activity 4 d after crush injury shows statistically significant increase for proteolytic activity in p75 ^NTR−/− mice. Quantification results are based on n = 5 p75 ^NTR−/−, n = 5 p75 ^NTR−/− tPA ^−/−, n = 5 tPA ^−/− and n = 4 wt mice. Bar graph represents means ± SEM (*, P < 0.05; by ANOVA). Bar: 50 μm (a–d), 300 μm (e–h).

Figure 4.

p75^NTR-mediated regulation of tPA and fibrinolysis in SCs. Primary SC cultures from wt (a) or p75 ^NTR−/− mice (b) on a 3D fibrin gel. Arrowheads indicate the border of fibrin degradation. Quantification of fibrin degradation (c) and tPA activity (d) from wt and p75 ^NTR−/− SCs. Experiments were performed three times in duplicates. Representative images are shown. Bar graph represents means ± SEM (P < 0.01; by t test). Bar, 130 μm.

Figure 5.

Expression of p75^NTR regulates tPA, PAI-1, and fibrinolysis in fibroblasts. 3D fibrin gel degraded by NIH3T3 (a), but not by NIH3T3p75^NTR cells (b). (c) Quantification of fibrin degradation. Experiments were performed seven times in duplicates. Phase-contrast microscopy shows lytic zones in NIH3T3 (d), but not in NIH3T3p75^NTR cultures (e). Zymography shows degradation of casein by NIH3T3 cells (f), whereas NIH3T3p75^NTR cells do not degrade casein (g). Quantification of tPA (h) and uPA (i) activity in supernatants from NIH3T3 and NIH3T3p75^NTR cultures. Experiments were performed five times in duplicates. (j) RT-PCR analysis for tPA, PAI-1, uPA, and GAPDH on cDNA derived from NIH3T3 and NIH3T3p75^NTR cells. (k) RT-PCR analysis for tPA and GAPDH on cDNA derived from uninjured wt, and wt or p75 ^NTR−/− mice three days after nerve injury. Bar graphs represent means ± SEM (statistics by t test). Bar: 1.2 cm (a and b), 130 μm (d–g).

Figure 6.

p75^NTR regulates tPA and PAI-1 via a PDE4/cAMP/PKA pathway. (a) db-cAMP induces fibrinolysis in NIH3T3p75^NTR cells. (b) IBMX increases tPA activity of NIH3T3p75^NTR cells to the levels of NIH3T3 cells. Inhibition of PKA by KT5720 shows decrease of tPA activity in both NIH3T3 and NIH3T3p75^NTR cells (P < 0.0001). (c) PKA activity assay shows decrease of PKA in NIH3T3p75^NTR cells. (d) Forskolin increases tPA mRNA in NIH3T3 and NIH3T3p75^NTR cells. Inhibition of PKA by KT5720 decreases tPA transcript. (e) Quantification of PAI-1 mRNA changes by real time PCR shows a fourfold increase of PAI-1 mRNA in NIH3T3p75^NTR cells compared with NIH3T3 cells. (f) Forskolin increases tPA activity in both wt (P<0.001) and p75 ^NTR−/− (P < 0.00001) SCs. NGF and BDNF do not affect activity of tPA (P > 0.8 and P > 0.3, respectively). (g) Transient overexpression of FL p75^NTR or p75 ICD leads to decreased levels of tPA in NIH3T3 cells. Experiments were performed at least 5 times in duplicates. *, P < 0.0001; **, P < 0.05; ***, P < 0.01. NS: non-significant. Bar graphs represent means ± SEM (statistics by ANOVA).

Figure 7.

p75^NTR decreases intracellular cAMP via PDE4. (a) cAMP levels in NIH3T3 and NIH3T3p75^NTR cells show a reduction of cAMP in NIH3T3p75^NTR cells, as compared with NIH3T3 cells. Treatment with PTX, IBMX, specific inhibitors for PDE1, PDE2, PDE3, and PDE4 (rolipram) shows that only IBMX (IC50 for PDE4 2–50 μM) and rolipram (IC50 for PDE4 0.8 μM) (P < 0.0001) increase levels of cAMP in NIH3T3p75^NTR cells to the levels of NIH3T3 cells. (b) Transient overexpression of FL p75^NTR or p75 ICD leads to decreased levels of cAMP in NIH3T3 cells. (c) siRNA mediated knockdown of p75^NTR levels in NIH3T3p75^NTR cells leads to increased levels of cAMP. (d) siRNA mediated knockdown of p75^NTR in primary rat Schwann cells leads to increased levels of cAMP. p75^NTR levels after siRNA knock down in duplicate samples of NIH3T3p75^NTR cells (e) and SCs (f). Immunostaining to detect cAMP in injured sciatic nerve reveals increased cAMP immunoreactivity in the sciatic nerve of p75 ^NTR−/− mice (h) when compared with wt controls (g). Experiments were performed four times in duplicate. Bar graphs represent means ± SEM (statistics by ANOVA or t test).

Figure 8.

p75^NTR interacts with PDE4A4/5. (a) Endogenous PDE4A5 co-IPs with p75^NTR in NIH3T3p75^NTR cells. Lysates were immunoprecipitated with anti-p75^NTR and probed with anti-PDE4A or anti-p75^NTR. Due to the low endogenous levels of PDE4A, higher exposure was necessary to detect PDE4A5 in the lysates (see Fig. S3 c). (b) FRET emission ratio change of NIH3T3 and NIH3T3p75^NTR cells for the pm-AKAR2.2 in response to forskolin. FRET change represents membrane activation of PKA (c) Mapping of the p75^NTR sites required for interaction with PDE4A5. Schematic diagram of HA-tagged p75^NTR intracellular deletions. TM, transmembrane domain; DD, death domain. Lysates were immunoprecipitated with an anti-HA antibody and probed with anti-PDE4A or anti-p75^NTR. (d) Mapping of the PDE4A4 sites required for interaction with p75^NTR. Schematic diagram of the C-terminal deletion of PDE4A4. Arrow indicates the deletion site. Lysates were immunoprecipitated with anti-p75^NTR and probed with anti-PDE4A or anti-p75^NTR. (e) Co-IP of purified, recombinant proteins reveals that both PDE4A4 and PDE4A5 interact with the ICD of p75^NTR, but PDE4D3 does not. (f) PDE4A4 peptide library screened with recombinant GST-p75^NTR ICD revealed three distinct domains of PDE4A4 (asterisks in d) that interact with the ICD of p75^NTR: the LR1 domain (peptides 40 and 41), the catalytic domain (peptides 135 and 136) and the unique C terminus (peptides 172 and 173). (g) Alanine scanning mutagenesis shows that substitution of C862 abolishes the interaction of p75^NTR with the 173 peptide that is unique to PDE4A.

Figure 9.

p75^NTR regulates fibrin clearance in the lung. LPS induces fibrin deposition (red) in the wt lung (b), when compared with the saline-injected lung (a). Lungs derived from p75 ^NTR−/− mice show less fibrin deposition (c). In situ zymography after 3 h of incubation shows clearance of casein in the lung of saline-injected wt (d), when compared with LPS injected wt lung (e). Lung from LPS-treated p75 ^NTR−/− mouse shows enhanced proteolytic activity (f), when compared with the wt mouse (e). Immunoreactivity for PAI-1 is increased in wt lung derived from LPS-treated mouse (h), when compared with saline-treated control (g). Lung from LPS-treated p75 ^NTR−/− mouse shows decreased PAI-1 (i), when compared with the wt LPS-treated mouse (h). (j) Western blot of fibrin precipitation from the lung shows an up-regulation of fibrin in the LPS-treated wt lung, when compared with the p75 ^NTR−/− lung. (k) Western blot for PAI-1 in the lung shows a decrease of PAI-1 in the p75 ^NTR−/− lung, when compared with the wt lung. Images are representative of n = 10 wt and n = 9 p75 ^NTR−/− mice. Western blots have been performed for n = 4 wt and n = 4 p75 ^NTR−/− mice. Bar: 150 μm (a–c), 75 μm (a–c, inset), 200 μm (d–f), 150 μm (g–i).

Figure 10.

Proposed model for the role of p75^NTR in the cAMP-mediated plasminogen activation. p75^NTR interacts with PDE4A4/5 resulting in degradation of cAMP and thus a reduction of PKA activity. Decrease in cAMP reduces expression of tPA and increases PAI-1, resulting in reduction of plasmin and plasmin-dependent extracellular proteolysis. Reduction of plasmin results in reduced fibrin degradation and ECM remodeling. Because plasmin can proteolytically modify nonfibrin substrates, such as growth factors and cytokines, this mechanism may be upstream of various cellular functions.