Research ArticleBiochemistry

Phosphorylation of the phosphatase PTPROt at Tyr399 is a molecular switch that controls osteoclast activity and bone mass in vivo

See allHide authors and affiliations

Science Signaling  08 Jan 2019:
Vol. 12, Issue 563, eaau0240
DOI: 10.1126/scisignal.aau0240
  • Fig. 1 ROKO and Y399F-HA-PTPROt mice.

    (A) Western blot (WB) showing PTPROt in osteoclasts and osteoblasts (Ob) of homozygous PTPROt knockout (ROKO, labeled RO) and wild-type (WT) mice, as well as in RAW 264.7 (Rw) cells and in HEK293 cells expressing exogenous PTPROt (293). The asterisk marks likely PTPROt alternative splicing products. The antibody against PTPROt cross-reacts with GLEPP-1 (32), and the arrow marks the expected location of the GLEPP-1 band. Molecular size markers are indicated in kilobases. Blotting for tubulin is a loading control. Blot is representative of n = 2 independent experiments using cells from different mice. (B) Schematic diagrams of WT-HA-PTPROt (WT-HA) and Y399F-HA-PTPROt (YF-HA) proteins. PTP, PTP catalytic domain; HA, C-terminal HA epitope tag. The arrow marks the position of the Y399F mutation. The sequence of the C terminus of PTPROt is also shown below the cartoon, with Tyr399 and Phe399 in bold. (C) Western blot showing endogenous PTPROt in OCLs of WT, ROKO (RO), WT-HA, and YF-HA mice. The HA tag slightly reduces the mobility of tagged PTPROt proteins. Blot is representative of n = 4 independent experiments using cells from different mice.

  • Fig. 2 Impaired activity of Y399F-HA-PTPROt OCLs.

    (A) Virtual μCT longitudinal sections of tibias from WT, ROKO, WT-HA-PTPROt (WT-HA), and Y399F-HA-PTPROt (YF-HA) 7-week-old female mice. All images are presented at the same magnification. Images are representative of n = 3 mice from each genotype. (B) Bar graph quantifying the surface area of bone in contact with OCLs (Oc.S/BS) from WT, ROKO, WT-HA, and YF-HA mice. n = 5 to 7 2-month-old mice per bar. Data shown are means ± SE; *P < 0.05 by two-way analysis of variance (ANOVA) with Bonferroni’s post hoc test. a.u., arbitrary units. (C) Bar graph quantifying the concentration of collagen telopeptides in circulation in WT, ROKO, WT-HA, and YF-HA mice. n = 10 to 11 2-month-old mice per bar. Data shown are means ± SE and were analyzed as in (B). (D) OCLs from WT-HA and YF-HA mice stained to show tartarate-resistant acid phosphatase (TRAP) activity, alongside quantification of OCL numbers (means ± SE) per well of a 24-well plate of each genotype, n = 5 mice per bar. (E) Pits indicating resorption (dark areas) in bovine bone that were generated by similar numbers of OCLs from WT-HA and YF-HA mice, alongside quantification of pit resorption by OCLs of both genotypes. Resorption was measured as percentage of bone area covered by pits of the total bone fragment surface area, mean ± SE. n = 7 (YF-HA) or 8 (WT-HA) mice per bar. **P = 0.011 by two-tailed Student’s t test. All scale bars, 500 μm.

  • Fig. 3 PTPROt dephosphorylates Src at Tyr416 and Tyr527.

    (A) FLAG-tagged WT PTPROt, the inactive, substrate-trapping mutant C325S PTPROt (CS), and Src were expressed in SYF cells as indicated. Cell lysates were subjected to FLAG immunoprecipitation (IP) and blotted (WB) for tyrosine phosphorylation (pTyr), PTPROt. Cell lysates were also blotted for Src phosphorylated at Tyr416, Src phosphorylated at Tyr527, and total Src. Blot is representative of n = 3 independent experiments. (B) Quantification of phosphorylation of Src at Tyr527 (pTyr527) and Tyr416 (pTyr416) in SYF cells expressing WT or CS PTPROt relative to cells not expressing PTPROt. Data are means ± SD from n = 5 (Tyr527) or n = 3 (Tyr416) independent experiments. *P ≤ 0.014 by one-way ANOVA with Tukey’s multiple comparisons test. (C) Phosphorylation of Src at Tyr527 and Tyr416 in OCLs from WT and ROKO mice as determined by protein blotting with phospho-specific antibodies alongside a representative blot showing Src phosphorylation. (D) Phosphorylation of Src at Tyr527 and Tyr416 in OCLs from WT-HA-PTPROt (WT-HA) and Y399F-HA-PTPROt (YF-HA) mice, alongside a representative blot showing Src phosphorylation. (E) Src kinase activity measured in OCLs from WT, ROKO, WT-HA, and YF-HA mice as indicated. Data in (C) to (E) are means ± SD obtained from OCLs from n = 7 to 9 male mice, aged 6 to 8 weeks, per bar. *P < 0.05, **P < 0.01 by Student’s t test.

  • Fig. 4 Phosphorylation and dephosphorylation of PTPROt at Tyr399.

    (A) Schematic representations of the indicated WT and mutant-tagged PTPROt molecules. Arrows indicate the Y399F mutation in the C terminus and the R331M and C325S mutations, both of which are located in the PTP catalytic domain. C-terminal FLAG (F) or HA (HA) tags are indicated. (B) Western blot showing tyrosine phosphorylation of PTPROt and its association with Src in SYF cells expressing the indicated combinations of Src and tagged WT, R331M, or C325S PTPROt proteins. The WT PTPROt blot contains a sample from cells treated with 0.5 mM sodium pervanadate as a technical positive control for phosphorylation. Y, Tyr399 is intact; YF, Tyr399 is mutated to phenylalanine. Cell lysates were subjected to FLAG immunoprecipitation (IP) and blotted for phosphorylated tyrosine (pTyr), PTPROt, and Src. Input blots were probed for FLAG and Src. Blots are representative of three independent experiments per PTPROt construct. (C) Western blot showing tyrosine phosphorylation of C325S PTPROt (CS) or (C325S, Y399F) PTPROt (CSYF) in SYF cells in response to a 10-min treatment with 0.5 mM sodium pervanadate (PV), as indicated. Blot is representative of three independent experiments. (D) Western blot showing trans-dephosphorylation of inactive, HA-tagged WT PTPROt at pTyr399 at the indicated time points after 30 and 60 min of exposure to active, FLAG-tagged PTPROt (WT). Negative controls include replacement of WT PTPROt with the inactive C325S mutant (CS), or the addition of 0.5 mM sodium pervanadate (WT + V). Blot is representative of three independent experiments. (E) Quantification of PTPROt trans-dephosphorylation at Tyr399. Data are means ± SD from n = 3 independent experiments per bar. *P < 0.05, **P < 0.01 versus WT at time 0 by one-way ANOVA with Dunnett’s multiple comparison test.

  • Fig. 5 Grb2 binds PTPROt at pTyr399.

    (A) Hybridization of a PTPROt peptide containing pTyr399 with a panel of 89 SH2 domains, each fused to GST. The sequence of the peptide is shown with pTyr399 highlighted in bold and bracketed. The ellipses labeled A to J indicate pairs of dots, with each pair representing binding to a specific SH2 domain from the proteins indicated at the right of the panel. Arrowheads and asterisks mark the positions of the SH2 domains of Src and Fyn, respectively, which did not bind the peptide. The complete peptide array analysis is shown in fig. S5. (B) Western blot (WB) showing PTPROt pulled down from lysates of pervanadate-treated OCLs from WT-HA-PTPROt (WT-HA) or Y399F-HA-PTPROt (YF-HA) mice using GST or GST fused to the SH2 domains of Grb2 or Src. Blot is representative of three independent experiments. (C) Western blots showing PTPROt pulled down from OCLs from WT-HA and YF-HA mice using GST or GST fused to full-length Grb2 (Grb2) or to the SH2, N-terminal SH3 (N-SH3), or C-terminal SH3 (C-SH3) domains of Grb2 as indicated. Blot is representative of two independent experiments.

  • Fig. 6 Influence of pTyr399 on association of PTPROt with Grb2 and Src.

    (A) Western blot showing coprecipitation of transgenically expressed Src and endogenous Grb2 with FLAG-tagged PTPROt from lysates of HEK293 cells expressing WT or R331M PTPROt. Blots are representative of three independent experiments. (B) Western blots showing coprecipitation of transgenically expressed Src and endogenous Grb2 with R331M PTPROt (RM) and (R331M Y399F) PTPROt (RMYF) in HEK293 cells (1st IP). Precipitated material was eluted from the beads and then precipitated with Src antibodies (2nd IP). Blots are representative of three independent experiments. (C) Bar graph showing the amount of Src that coprecipitates with RM PTPROt and RMYF PTPROt in the 1st IP of (B). Data are means ± SD from n = 3 independent experiments per bar, P value by paired, two-tailed t test. (D) PLA performed on OCLs from WT-HA-PTPROt (WT-HA) and Y399F-HA-PTPROt (YF-HA) mice. Red signals indicate interaction between antibodies against HA and Src. Staining of OCLs from WT mice, which only produce nontagged endogenous PTPROt, is a negative control. Cells are also stained for actin (green), which marks the podosomal array (the sealing zone–like structure) at the cell periphery, and for DNA [4′,6-diamidino-2-phenylindole (DAPI), blue] to mark nuclei. The boxed areas in the top row of images are magnified below. (E) Quantification of PLA signals in OCLs from WT-HA and YF-HA mice. The number of red PLA signals in individual OCLs was divided by the number of nuclei in each OCL, which is proportional to OCL size. Data (means ± SD) are shown normalized to the average values of WT-HA cells. Each bar represents data from n = 8 cells from nonoverlapping fields obtained in two independent experiments. Statistical analysis was performed by two-way ANOVA on original log-transformed data, accounting for batch and treatment effects. (F) Single-antibody PLA (negative control) performed on OCLs from WT-HA mice with primary antibodies against either HA or Src as indicated, and with both secondary antibodies. Actin and DNA are stained as in (D). All scale bars, 20 μm.

  • Fig. 7 Model depicting the dual roles of PTPROt toward Src.

    Active Src, which is phosphorylated at pTyr416, physically associates with PTPROt, which dephosphorylates pTyr416. The mode of interaction between Src with PTPROt in this case is not known, but the end result is a reduction in Src activity. When PTPROt is phosphorylated at Tyr399 by active Src molecules (and possibly by other kinases), the SH2 domain of Grb2 binds PTPROt at pTyr399, and Grb2 recruits Src to the phosphatase. PTPROt then dephosphorylates Src at the inactivating site pTyr527, thus activating the kinase. PTPROt autodephosphorylation (dashed line) in cis or trans helps to terminate its Src-stimulating activity.

  • Table 1 μCT analysis of tibiae from WT and ROKO mice.

    Parameters shown are BV/TV (trabecular bone volume as percentage of total volume), Tb.Th (trabecular thickness), Tb.Sp (trabecular separation), and Tb.N (trabecular number). n = 8 mice of each sex per genotype, aged 2 months. Data are means ± SD and were analyzed by unpaired, two-tailed Student’s t test. NS, not significant.

    GenotypeSexBV/TV
    (%)
    Tb.Th
    (μm)
    Tb.Sp
    (μm)
    Tb.N
    (mm−1)
    WTM12.5 ± 2.522.6 ± 1.1165.3 ± 39.35.39 ± 0.88
    ROKOM11.5 ± 2.522.1 ± 1.9175.3 ± 21.35.14 ± 0.10
    P valueNSNSNSNS
    WTF13.4 ± 2.022.6 ± 1.5147.1 ± 19.65.93 ± 0.67
    ROKOF11.8 ± 2.920.9 ± 1.3164.2 ± 43.45.63 ± 1.21
    P valueNS0.02NSNS
  • Table 2 μCT analysis of tibiae from WT-HA-PTPROt (WT-HA) and Y399F-HA-PTPROt (Y399F-HA) mice.

    Parameters shown are BV/TV (trabecular bone volume as percentage of total volume), Tb.Th (trabecular thickness), Tb.Sp (trabecular separation), Tb.N (trabecular number), and tibial length. n = 6 mice of each sex per genotype, aged 2 months. Data are means ± SD and were analyzed by unpaired, two-tailed Student’s t test. Y399F-HA-PTPROt data are from mouse line 8; WT-HA-PTPROt data are combined from mouse lines 2 and 3. See table S1 for a similar comparison to Y399F-HA-PTPROt mouse line 16.

    GenotypeSexBV/TV
    (%)
    Tb.Th
    (μm)
    Tb.Sp
    (μm)
    Tb.N
    (mm−1)
    Length
    (mm)
    WT-HAM15.0 ± 2.022.8 ± 2.1127.3 ± 22.76.77 ± 1.0317.43 ± 0.15
    Y399F-HAM22.0 ± 2.426.1 ± 2.498.4 ± 20.18.17 ± 1.1617.47 ± 0.19
    P value0.0240.0160.0420.051NS
    WT-HAF12.3 ± 2.524.2 ± 0.8175.0 ± 38.65.18 ± 1.0916.45 ± 0.29
    Y399F-HAF18.6 ± 4.425.5 ± 2.7114.4 ± 21.27.30 ± 1.1916.98 ± 0.13
    P value0.013NS0.0070.0090.002
  • Table 3 Transplanting bone marrow from Y399F-HA PTPROt mice into WT-HA PTPROt mice increases bone mass.

    Bone marrow cells from female WT-HA-PTPROt mice (WT-HA, controls) or Y399F-HA PTPROt mice (Y399F-HA) aged 5 weeks were implanted in lethally irradiated male WT-HA-PTPROt mice of the same age, and tibiae were analyzed 1 month later by μCT. Parameters shown are BV/TV (trabecular bone volume as percentage of total volume), Tb.Th (trabecular thickness), Tb.Sp (trabecular separation), and Tb.N (trabecular number). n = 7 or 8 acceptor mice in each category. Data are means ± SD and were analyzed by unpaired, two-tailed Student’s t test. The success of the transplantation procedure was evaluated by polymerase chain reaction (PCR) of DNA of macrophages isolated from the mice at sacrifice for the presence of DNA markers specific for the X and Y chromosomes (fig. S3).

    Donor genotypeAcceptor
    genotype
    BV/TV
    (%)
    Tb.Th
    (μm)
    Tb.Sp
    (μm)
    Tb.N
    (mm−1)
    WT-HAWT-HA9.3 ± 1.319.6 ± 4.3178.2 ± 29.15.17 ± 0.90
    Y399F-HAWT-HA12.2 ± 1.622.2 ± 2.9161.5 ± 27.25.58 ± 0.95
    P value0.00110.089NSNS
  • Table 4 Bone formation parameters in femurs from WT versus ROKO and from WT-HA-PTPROt versus Y399F-HA-PTPROt mice.

    Parameters shown are BFR/BS (bone formation rate normalized to bone surface), N.Ob/B.Pm (number of osteoblasts normalized to bone perimeter), Ob.S/BS (osteoblast surface normalized to bone surface), and N.Ot/B.Ar (number of osteocytes normalized to bone area). n = 7 or 8 male mice per genotype, aged 2 months. Data are means ± SD and were analyzed by two-tailed Student’s t test. NS, not significant.

    GenotypeSexBFR/BS
    (μm3/μm2 per day)
    N.Ob/B.Pm
    (/μm)
    Ob.S/BS
    (%)
    N.Ot/B.Ar
    (/μm2)
    WTM0.827 ± 0.4576.39 ± 1.993.74 ± 1.35797.1 ± 109.6
    ROKOM0.965 ± 0.1658.04 ± 2.764.84 ± 1.54852.1 ± 180.5
    P valueNSNSNSNS
    WT-HAM1.323 ± 0.4994.84 ± 2.422.84 ± 1.30797.5 ± 255.2
    Y399F-HAM1.056 ± 0.5447.18 ± 2.103.70 ± 0.66717.6 ± 224.2
    P valueNSNSNSNS

Supplementary Materials

  • www.sciencesignaling.org/cgi/content/full/12/563/eaau0240/DC1

    Fig. S1. Detection of the Y399F mutation engineered into the Ptpro gene of mice.

    Fig. S2. OCLs from WT and ROKO mice.

    Fig. S3. Verification of engraftment after bone marrow transplantation.

    Fig. S4. Activity of WT PTPROt and several PTPROt mutants toward PNPP.

    Fig. S5. PTPROt pTyr399 binds to specific SH2 domains.

    Fig. S6. PTPROt coprecipitates with Grb2 in a phosphorylation-dependent manner.

    Fig. S7. Staining of OCLs with antibodies against Src and HA.

    Table S1. μCT analysis of tibiae from WT-HA-PTPROt versus Y399F-HA-PTPROt (line 16) mice.

  • This PDF file includes:

    • Fig. S1. Detection of the Y399F mutation engineered into the Ptpro gene of mice.
    • Fig. S2. OCLs from WT and ROKO mice.
    • Fig. S3. Verification of engraftment after bone marrow transplantation.
    • Fig. S4. Activity of WT PTPROt and several PTPROt mutants toward PNPP.
    • Fig. S5. PTPROt pTyr399 binds to specific SH2 domains.
    • Fig. S6. PTPROt coprecipitates with Grb2 in a phosphorylation-dependent manner.
    • Fig. S7. Staining of OCLs with antibodies against Src and HA.
    • Table S1. μCT analysis of tibiae from WT-HA-PTPROt versus Y399F-HA-PTPROt (line 16) mice.

    [Download PDF]

Navigate This Article