Netrin-1 functions as a suppressor of bone morphogenetic protein (BMP) signaling

Abstract

Netrin-1 is a secreted protein that is well known for its involvement in axonal guidance during embryonic development and as an enhancer of cancer cell metastasis. Despite extensive efforts, the molecular mechanisms behind many of the physiological functions of netrin-1 have remained elusive. Here, we show that netrin-1 functions as a suppressor of bone morphogenetic protein (BMP) signaling in various cellular systems, including a mutually inhibitory interaction with the BMP-promoting function of leucine-rich repeats and immunoglobulin-like domains (LRIG) proteins. The BMP inhibitory function of netrin-1 in mouse embryonic fibroblasts was dependent on the netrin receptor neogenin, with the expression level regulated by both netrin-1 and LRIG proteins. Our results reveal a previously unrecognized function of netrin-1 that may help to explain several of the developmental, physiological, and cancer-promoting functions of netrins at the signal transduction level.

Figure 1

Netrin-1 inhibits BMP signaling in MEFs. (a–c) Wild-type MEFs were starved in serum-free medium and then treated with various concentrations of BMP4, BMP6, or BMP9 together with various concentrations of netrin-1 for an hour. Then, BMP signaling was assessed via the analysis of nuclear pSmad1/5 levels. The graphs show the average means with error bars showing the standard deviations of three independent experiments performed in duplicate (one-way ANOVA: ^nsp ≥ 0.05; ^#p < 0.05; ^##p < 0.01; ^###p < 0.001). (d) Wild-type MEFs were treated with various concentrations of netrin-1 at different times before, with, or after the addition of BMP4 [5 ng/ml] to the wells. One hour after the addition of BMP4, BMP signaling was assessed via the analysis of nuclear pSmad1/5 levels. The x-axis shows the time at which netrin-1 was added relative to the addition of BMP4: negative values indicate the addition of netrin-1 before the addition of BMP4, ‘0′ indicates the simultaneous addition of netrin-1 and BMP4, and positive values indicate the addition of netrin-1 after the addition of BMP4. The graph shows the average means with the standard deviations of four independent experiments performed in duplicate (one-way ANOVA: ^####p < 0.0001). (e) Id1 mRNA expression levels in response to BMP4 and netrin-1. MEFs were treated, or not, with 20 ng/ml BMP4 and/or 0.25 μg/ml netrin-1 for x hours followed by analysis of RNA expression levels through real-time qRT-PCR. Shown are the Id1 levels normalized to the Rn18s levels. The graph represents the average means with standard deviations of four independent experiments performed in duplicate (Student’s t-test: **p < 0.01; ***p < 0.001). (f) MEFs of different Ntn1 and Lrig genotypes (Ntn1^+/+ or Ntn1^−/− and Lrig-wild-type or Lrig-null [Lrig1^−/−;Lrig2^−/−;Lrig3^−/−], respectively) were treated with various concentrations of BMP4, and BMP signaling was assessed via analysis of the nuclear pSmad1/5 levels. The graph shows the means with the standard deviations of four Ntn1^+/+;Lrig-wild-type or Ntn1^+/+;Lrig-null biological replicates and three Ntn1^−/−;Lrig-wild-type or Ntn1^−/−;Lrig-null biological replicates determined through three independent experimental repeats. The error bars represent the standard deviations (Student’s t-test: **p < 0.01; ****p < 0.0001. Colors of asterisks represent the respective genotype, according to the graph symbols, compared to Ntn1^+/+;Lrig-WT). (g, h) Cross-titration of netrin-1 and LRIG1 (g) or LRIG3 (h). LRIG1- or LRIG3-inducible Lrig-null and Ntn1^−/− MEFs were treated with different concentrations of doxycycline (between 0 and 1 μg/ml) to induce various levels of LRIG protein expression. BMP signaling was induced using 5 ng/ml BMP4 together with different amounts of netrin-1 (between 0 and 1 μg/ml) for an hour. The fluorescence response for the nuclear pSmad1/5 levels induced by BMP4 is plotted on the y-axis. The doxycycline-induced expression levels of FLAG-tagged LRIG1 or LRIG3 were measured and plotted on the x-axis. Each dot represents the quantification of pSmad1/5 and LRIG-FLAG from each measurement. The graphs show data obtained from three independent experiments performed in duplicate for each concentration of netrin-1 and doxycycline. The R² values for the scattered plots are tabulated in Tables 1 and 2.

Figure 2

Netrin-1 inhibits BMP signaling in various mammalian cell lines. Western blot analysis showing the inhibitory effect of netrin-1 on the BMP4-induced pSmad1/5 levels in cell lines of different origins. (a) MEFs and human embryonic kidney (HEK293) cells. (b) Human ovarian cancer (Ovcar3 and Ovsaho), malignant melanoma (A375), breast cancer (MDA-MB-415 and T-47D), glioblastoma (TB101 and TB107), colorectal carcinoma (HCT116 and LoVo), and acute lymphoblastic pre-B cell leukemia (697 and RCH-ACV). Representative Western blots of pSMAD1/5 and actin after the cells were treated with BMP4 and netrin-1 are shown together with the quantifications of the pSMAD1/5/actin ratios. Uncropped blots are shown in Fig. S4. Shown are the means and individual values of four independent experiments (Student’s t-test: ^NSp ≥ 0.05; *p < 0.05; **p < 0.01).

Figure 3

Neogenin is required for netrin-1-induced inhibition of BMP signaling, and its expression levels are regulated by netrin-1 and LRIG proteins. (a) Effects of netrin-1 on BMP signaling in Neo1 wild-type (Neo1^+/+) and Neo1-deficient (Neo1^−/−) MEFs. The cells were treated with different concentrations of BMP4 in the presence or absence of 0.25 μg/ml netrin-1. Signaling was analyzed using a pSmad1/5 assay. The graph shows the average means with the standard deviations of four biological replicates. (b) Neogenin expression levels in MEFs of different genotypes (Ntn1^+/+ or Ntn1^−/− and Lrig-wildtype or Lrig-null [Lrig1^−/−; Lrig2^−/−; Lrig3^−/−]). Neogenin and actin levels were analyzed by Western blotting. The graph represents the average means with standard deviations of band intensity ratios (neogenin/actin) from four Ntn1^+/+;Lrig-wild-type biological replicates and three Ntn1^−/−;Lrig-wild-type or Ntn1^−/−;Lrig-null biological replicates that were determined with three independent experiments. (c, d) Concentration-dependent (c) and time-dependent (d) effects of netrin-1 treatment on neogenin expression levels in Ntn1^−/−;Lrig-wild-type and Ntn1^−/−;Lrig-null MEFs. Representative blots and graphs showing the average means with standard deviations of four independent experiments are shown (one-way ANOVA: ^#p < 0.05; ^###p < 0.001; ^####p < 0.0001. Post-hoc test with multiple comparisons [compared to netrin-1, 0 ng/ml] adjusted p-value: ^nsp ≥ 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (e–g) Effects of induced expression of LRIG1 (e), LRIG2 (f), and LRIG3 (g) on neogenin expression levels. LRIG expression was induced through the treatment of LRIG-inducible MEFs with 100 ng/ml of doxycycline, overnight. Neogenin and actin levels were analyzed using Western blotting. Uncropped blots are shown in Fig. S7. Representative blots and graphs representing average means with standard deviations of four independent experiments are shown. (Student’s t-test; *, p < 0.05; **, p < 0.01; ***, p < 0.001. Colors of asterisks in (a) represent the respective genotype, according to the graph symbols, compared to Neo1^+/+).

Figure 4

Netrin-1 inhibits adipogenesis and chondrogenesis in vitro. (a–e) Netrin-1 inhibits MEF adipogenesis. Wild-type MEFs were induced with an adipogenic cocktail together with or without 20 ng/ml BMP4 and/or 0.3 µg/ml netrin-1. After nine days, the cells were fixed, stained with Oil Red O, and imaged, and the staining was quantified. (a–d) Representative micrographs of MEFs treated with the adipogenic cocktail together with (a) no further additions, (b) BMP4, (c) BMP4 and netrin-1, and (d) netrin-1. Scale bars, 0.6 mm. (e) Quantification of the Oil Red O-stained areas as outlined in a-d. Shown are the means and individual values of eight independent experiments. (Student’s t-test: ***p < 0.001). (f) Netrin-1 inhibits chondrogenesis. ATDC5 cells were stimulated with BMP4 alone or in combination with different concentrations of netrin-1 for 72 h. Thereafter, chondrogenesis was assessed via an alkaline phosphatase (ALP) assay. The graph represents the average means with standard deviations of four independent experiments performed in duplicate. (One-way ANOVA: ^####p < 0.0001. Post-hoc test with multiple comparisons [compared to netrin-1, 0 ng/ml] adjusted p-value: **p < 0.01; ***p < 0.001; ****p < 0.0001).