Functional Roles of the Nuclear Localization Signal of Parathyroid Hormone-Related Protein (PTHrP) in Osteoblastic Cells

Abstract

PTHrP is an important regulator of bone remodelling, apparently by acting through several sequence domains. We here aimed to further delineate the functional roles of the nuclear localization signal (NLS) comprising the 88–107 amino acid sequence of PTHrP in osteoblasts. PTHrP mutants from a human PTHrP (−36/+139) cDNA (wild type) cloned into pcDNA3.1 plasmid with deletion (Δ) of the signal peptide (SP), NLS, T¹⁰⁷, or T107A replacing T¹⁰⁷ by A¹⁰⁷ were generated and stably transfected into osteoblastic MC3T3-E1 cells. In these cells, intracellular trafficking, cell proliferation and viability, as well as cell differentiation were evaluated. In these transfected cells, PTHrP was detected in the cytoplasm and also in the nucleus, except in the NLS mutant. Meanwhile, the PTH type 1 receptor (PTH1R) accumulates in the cytoplasm except for the ΔSP mutant in which the receptor remains at the cell membrane. PTHrP-wild type cells showed enhanced growth and viability, as well as an increased matrix mineralization, alkaline phosphatase activity, and osteocalcin gene expression; and these features were inhibited or abolished in ΔNLS or ΔT¹⁰⁷ mutants. Of note, these effects of PTHrP overexpression on cell growth and function were similarly decreased in the ΔSP mutant after PTH1R small interfering RNA transfection or by a PTH1R antagonist. The present in vitro findings suggest a mixed model for PTHrP actions on osteoblastic growth and function whereby this protein needs to be secreted and internalized via the PTH1R (autocrine/paracrine pathway) before NLS-dependent shuttling to the nucleus (intracrine pathway).

PTHrP has emerged as an important cytokine with a variety of cell functions including growth and survival, migration, and differentiation (1). By alternative splicing, the human PTHrP gene generates 3 different mRNA species that are translated into 3 main protein isoforms of 139, 141, and 173 amino acids with a common N terminus (2). Each of these PTHrP isoforms contains various biologically significant epitopes: a signal peptide (SP) in its N terminus, which may drive the protein to the endoplasmic reticulum to enter the secretory pathway; the 1–36 region that has partial homology to PTH, which accounts for its interaction with the PTH type 1 receptor (PTH1R); a midregion, which contains the nuclear localization signal (NLS), an ill-defined bipartite epitope comprising the 88–107 region, which is responsible for trafficking to the nucleus/nucleolus (3–6); and a C-terminal region, containing the highly conserved epitope 107–111, named osteostatin by Fenton et al (7, 8) after describing its antiresorptive activity. The existence of alternative translation initiation sites within the SP allows PTHrP to be driven mainly to the endoplasmic reticulum (secretory pathway) or to the nucleus through the NLS (5, 6). In addition, there appears to be competition between different directions for the PTHrP protein fate, either into the secretory pathway or the nucleus through the SP and NLS, respectively (9).

PTHrP has been reported to exert intracrine actions affecting cell proliferation and/or cell survival in chondrocytes, vascular smooth muscle cells (VSMCs), and colon carcinoma cells (10–13). The underlying mechanisms of these actions, however, remain poorly understood. Previous studies suggest that both the expression of PTHrP and its nuclear/nucleolar localization as a response to mitogenic factors are cell cycle dependent (14–17). PTHrP can interact with RNA in the nucleolus and thus may regulate ribosomal function (18). In addition, the C-terminal 108–139 domain has been shown to be required for the intracrine action of PTHrP on VSMC proliferation (19, 20).

In bone, PTHrP and the PTH1R are abundant in both chondrocytes and cells of the osteoblastic lineage. In these cells, PTHrP, through interaction with the PTH1R, can exert diverse autocrine and paracrine actions to modulate bone formation and remodeling (21–23). Mice null for either PTHrP or the PTH1R show skeletal alterations causing death soon after birth or, more dramatically, embryonic lethality, respectively, whereas PTHrP happloinsufficiency causes trabecular osteopenia postnatally (24, 25). Recently, it has been shown that knock-in mice expressing a truncated form of PTHrP missing both the mid and C-terminal regions display premature osteoporosis (26, 27), suggesting the significant role of these PTHrP domains in bone development and metabolism. Intermittent administration of PTHrP(107–139), comprising the entire C-terminal region of the major PTHrP isoform in humans, has also recently been shown to exert osteogenic features in various osteoporosis mouse models (28–30). The osteostatin epitope might be responsible for these actions of this C-terminal PTHrP peptide (31). Both PTHrP(107–139) and osteostatin have been reported to trigger intracellular calcium transients through interaction with membrane calcium channels in an osteoblastic osteosarcoma cell line (32). Of note, these C-terminal PTHrP peptides similarly promoted vascular endothelial growth factor (VEGF) 2 transactivation, apparently by a src-mediated mechanism in osteoblastic cells (33, 34). This signalsome targeted by osteostatin has an impact on the modulation of osteoblastic function (32–34). However, characterization of putative receptors for C-terminal PTHrP remains elusive.

We here sought out to delineate the roles of the NLS domain of PTHrP in regulating osteoblastic function. The putative role of the T¹⁰⁷ residue in this domain of PTHrP is ill defined because this residue overlaps the NLS and the osteostatin epitope. Thus, we also examined the consequences of deletion/mutation of this poorly explored residue on PTHrP actions in osteoblastic cells. To these aims, several PTHrP deletion mutants were transfected into MC3T3-E1 cells, an osteoblastic model extensively used in PTHrP studies (31, 34–39). We demonstrate that the PTHrP epitope comprising both 88–95 and 102–106 amino acid clusters is partially required for PTHrP driving MC3T3-E1 cell proliferation and survival and also cell differentiation. Moreover, our data herein indicate that threonine at position 107 (T¹⁰⁷) contributes to the osteogenic actions of PTHrP in osteoblasts. The present in vitro findings suggest a mixed mechanism whereby PTHrP needs to be secreted and internalized via the PTH1R (autocrine/paracrine pathway) before NLS-dependent shuttling to the nucleus (intracrine pathway) for modulating osteoblastic growth and function.

Materials and Methods

PTHrP constructs and cell transfections

The different constructs used were based on pcDNA3.1 plasmid containing a human PTHrP (−36/+139) cDNA wild type (WT); PTHrP (−36/+139; Δ88–95+102–106) cDNA (ΔNLS-PTHrP); or PTHrP (1–139) cDNA (ΔSP-PTHrP) with an epitope tag at the respective C terminus, corresponding to human influenza hemagglutinin (HA) (for immunodetection) (11). The ΔT¹⁰⁷-PTHrP and T107A-PTHrP constructs were generated from WT or ΔSP-PTHrP plasmids, respectively, by in vitro site-directed mutagenesis, using a standard protocol (Stratagene). PTHrP sequences were confirmed by DNA sequencing (Secugen).

Mouse osteoblastic MC3T3-E1 cells were grown in α-MEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin in a 5% CO₂ humidified incubator at 37°C (31). Cells were plated at 1.5 × 10⁴/cm² in 6-well plates 24 hours before transfection with the different PTHrP constructs. Transfections were carried out in serum-free medium with X-tremeGENE 9 DNA transfection reagent (Roche Diagnostics) for 24 hours at 37°C, as recently described (37). Stably transfected clones were selected by treatment with 500 μg/mL geneticin (G418; Life Technologies) for 3 weeks. Confirmation of PTHrP construct transfection was accomplished by real-time PCR, Western blotting, and immunofluorescence. As a positive control for PTHrP expression, human kidney 2 cells, grown in RPMI 1640 with 10% FBS and antibiotics in 5% CO2 at 37°C, were used. For PTH1R silencing, MC3T3-E1 cells stably transfected with the different PTHrP constructs, at 40%–60% confluence in 6-well dishes, were incubated in OptiMEM (Life Technologies) without FBS for 24 hours. Then, the cells were transfected with 2 independent small interfering RNA (siRNA) sequences against the PTH1R (100 nM) (Life Technologies). Transfection was performed for 48 hours using lipofectamine LTX transfection reagent (Invitrogen), following manufacturer's instructions.

Western analysis

Total cell protein extracts were obtained by cell lysis buffer in 50 mM Tris-HCl, pH 7.4, with 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM phenylmethylsulfonylfluoride, and 0.8 mM aprotinin. The Subcellular Protein Fractionation Kit (Pierce Chemical Co) was used for the obtainment of cytoplasm and nuclear protein extracts. Protein content was determined by the bicinchoninic acid-based protein assay (Thermo Fisher Scientific), using BSA as standard. Protein extracts (20–40 μg) were then separated on 9%–14% polyacrylamide-sodium dodecyl sulfate gels under reducing conditions and transferred onto nitrocellulose membranes (Bio-Rad Laboratories). Blots were blocked with 5% defatted milk in 50 mM Tris-HCl, pH 7.5, with 0.1% Tween 20 for 1 hour at room temperature, and then incubated overnight at 4°C with mouse monoclonal anti-HA antibody (6E2) (Cell Signaling Technology) or rabbit polyclonal PTH1R antibody (IV peptide; Covance). Mouse monoclonal anti-α-tubulin (Sigma-Aldrich) or rabbit polyclonal antilamin-1β (Cell Signaling Technology) antibodies were used as loading controls, respectively. After incubation with the corresponding secondary horseradish peroxidase-coupled IgG (Santa Cruz Biotechnology), blots were developed with the ECL system (GE Healthcare).

Immunofluorescence

Cells were fixed with 4% p-formaldehyde and permeabilized using 0.1% Triton in PBS, pH 7.4. Nonspecific binding was blocked with 5% BSA, followed by overnight incubation with a mouse monoclonal anti-HA antibody, kindly provided by M. Monsalve (Instituto de Investigaciones Biomédicas, CSIC, Madrid), or the rabbit polyclonal anti-PTH1R antibody (Covance) at 4°C in a humidified chamber. Cells were rinsed in 0.1% Triton X-100-PBS before incubation for 1 hour with Alexa fluor 488-conjugated antimouse or antirabbit IgG (Invitrogen). The nucleus was stained with propidium iodide (Sigma-Aldrich). Samples were mounted in FluorSafe Reagent (Calbiochem) and examined using a Leica DM-IRB confocal microscope.

Real-time PCR

Total RNA was extracted from the different MC3T3-E1 cell clones by using Trizol (Invitrogen). cDNA synthesis was performed using avian myeloblastosis virus reverse transcriptase (Promega Corp) with random hexamer primers, and real time PCR was carried out in an ABI PRISM 7500 system (Applied Biosystems). Unlabeled specific primers for human PTHrP, mouse PTH1R and mouse osteocalcin (OC) and TaqMan MGB probes were obtained by Assay-by-Design (Applied Biosystems). To analyze hPTHrP in ΔSP-PTHrP-overexpressing cells, SYBR Premix Ex Taq (Takara) and the following specific primers for hPTHrP and the internal control 18S rRNA were used:

hPTHrP forward, 5′-AGATTTACGGCGACGATTCTTC-3′; reverse, 5′-CTCCGAGGTAGCTCTGATTTCAG-3′; and 18S rRNA forward, 5′-ATGCTCTTAGCTGAGGTGCCCG-3′; reverse, 5′-ATTCCTAGCTGCGGTATCCAGG-3′. Results are expressed in mRNA copy numbers, calculated for each sample using the cycle threshold (Ct) value and normalized against 18S rRNA as described elsewhere (39).

Evaluation of cell growth and cell viability

Growing MC3T3-E1 cells stably transfected with the different PTHrP constructs or vector alone were trypsinized, washed with PBS, pelleted, and resuspended in PBS containing 50 μg/mL propidium iodide and 100 U/mL RNAse A. Cell cycle was analyzed by flow cytometry using a FACS Calibur flow cytometer (BD Biosciences). Cell proliferation was also determined by addition of Alamar Blue solution (AbD Serotec) at 10% (vol/vol) to each well of 24-well plates (10⁴cells/cm²) at day 2 of growth in FBS-containing medium. Four hours thereafter, absorbance at 540 nm was measured in the cell-conditioned medium.

MC3T3-E1 cells were maintained in serum-depleted medium for up to 6 days. Cell death at 2 and 6 days after serum depletion was determined by Trypan blue exclusion as reported (38, 39), using a Countess Automated Cell Counter (Life Technologies). Apoptosis was estimated at 24 hours in serum-depleted medium as the percentage of hypodiploid cells by using flow cytometry as described above. When present, the PTH1R antagonist (Asn¹⁰, Leu¹¹, D-Trp¹²) PTHrP(7–34) amide PTHrP[7–34]) was added to the culture medium at a final concentration of 1 μM (30).

Matrix mineralization and alkaline phosphatase activity (ALP) assays

MC3T3-E1 cells transfected with the different constructs were grown in the medium mentioned above supplemented with 50 μg/mL ascorbic acid and 10 mM β-glycerophosphate (osteogenic medium) for 15 days. When present, PTHrP(7–34) (1 μM) was added every other day to the culture medium. Matrix mineralization was evaluated by alizarin S red staining. The stain was dissolved with 10% cetylpyridinum chloride in 10 mM sodium phosphate, pH 7.4, and absorbance at 620 nm was measured (31). ALP activity was measured in cell extracts obtained with 0.1% Triton X-100, using p-nitrophenylphosphate as substrate (31). ALP activity was normalized to cell protein content determined as described above.

Statistics

Results are mean ± SD. Statistical analysis was performed by Mann-Whitney test or Kruskal-Wallis nonparametric ANOVA followed by Dunn's post hoc test, when appropriate. P < .05 was considered significant.

Results

Characterization of MC3T3-E1 cell clones constitutively overexpressing different PTHrP mutants

In order to characterize the functional roles of the NLS in the PTHrP molecule in osteoblasts, we established MC3T3-E1 cells overexpressing the different constructs depicted in Figure 1. Empty pcDNA3.1 vector transfectants served as controls. We selected clones showing PTHrP mRNA overexpression within a similar range for further studies (specific clones used are shown with an arrow in Figure 2A). Because the TaqMan probe in the Assay-by-Design (Applied Biosystems) system for PCR amplification of the human PTHrP constructs anneals in the SP region, we used Western blotting and SYBR green-based real time PCR to determine PTHrP protein and mRNA expression, respectively, in the ΔSP construct (Figure 2A, inset). Protein expression of the PTH1R was similarly and significantly down-regulated in the PTHrP-overexpressing cells but in the ΔSP-PTHrP mutant (Figure 2B).

Figure 1.

The PTHrP constructs used for stably transfection into MC3T3-E1 cells. HA-tag represents influenza HA epitope tag for immunodetection in these transfected cells. The numbers above vertical lines indicate the site of amino acid residues in PTHrP sequence. SP, signal peptide.

Figure 2.

PTHrP and PTH1R expression analysis of MC3T3-E1 clones stably transfected with the different constructs or pCDNA3.1 vector alone (V). Human (h)PTHrP gene expression assessed by real-time PCR of 2 clones for each construct is shown. For comparison, hPTHrP expression in native human kidney (HK)2 cells is also depicted (A). Clones expressing comparable levels of PTHrP mRNA were selected for each construct for further studies as shown here. Arrows indicate the selected clones used in this study. PTHrP expression in the ΔSP-PTHrP mutant was compared to V and WT-PTHrP using SYBR green-based real time PCR (A, right inset). PTHrP (A, left inset), and PTH1R (B) protein expression in these clones was also evaluated by Western analysis. A representative autoradiogram is shown. Relative cytoplasmic and nuclear localization of PTHrP protein in the different cell clones studied were examined by Western blot (C and D). The arrowheads on the left of each autoradiogram denote the position of protein markers (identified by their molecular weight). HA, hemagglutinin.

We next evaluated and compared the cellular localization of PTHrP in MC3T3-E1 cells transfected with the aforementioned constructs using Western analysis of cytoplasmic and nuclear fractions and immunofluorescence. As expected, we found that these cells transfected with the WT or any of the mutated constructs showed increased cytoplasmic PTHrP levels, compared with empty vector-transfected cells (Figure 2C). PTHrP protein was also detected in the nuclear fraction of these cells, except in ΔNLS-PTHrP-overexpressing cells. Notably, increased nuclear levels of PTHrP were observed in WT-PTHrP compared with that in the other PTHrP mutants, even though total PTHrP expression was similar in all these cells studied (Figure 2, A and D). Consistent with Western data, using immunofluorescence, intense PTHrP positivity was found thoroughly in the cytosol and the nucleus in WT cells in the presence of an intact SP; PTHrP staining was less prominent but still detected in the cytoplasm and the nucleus of the ΔSP-PTHrP mutant but barely present in the nucleus in cells in which T¹⁰⁷ was deleted or mutated to A¹⁰⁷ (Figure 3). As expected, the ΔNLS-PTHrP mutant displayed no signal in the nucleus (Figure 3).

Figure 3.

Intracellular localization of each PTHrP construct stably transfected in MC3T3-E1 cells. Immunofluorescence to detect HA-tag or propidium iodide (PI) staining or both (merge) was shown and described in detail in the text. Cells transfected with pcDNA3.1 vector alone were used as negative control. Images represent the results of 3 independent observations.

The localization of the PTH1R was also examined in these pcDNA3.1-transfected (control) cells as well as in WT-PTHrP, ΔT¹⁰⁷-PTHrP, ΔNLS-PTHrP, and ΔSP-PTHrP mutants. All of these cells showed membrane-delimited PTH1R localization, but only those with an intact SP, namely WT-PTHrP, ΔT¹⁰⁷-PTHrP, and ΔNLS-PTHrP, also displayed a cytoplasmic pattern of immunostaining. Nuclear PTH1R localization was undetected in any of these mutants studied (Figure 4).

Figure 4.

Cellular localization of PTH1R in MC3T3-E1 cells stably transfected with each PTHrP construct. Immunofluorescence to detect this receptor by using anti-PTH1R or propidium iodide (PI) staining or both (merge) was shown and described in detail in the text. Cells transfected with pcDNA3.1 vector alone were used as negative control. Images represent the results of 3 independent observations.

An intact NLS and the PTH1R are required for the effects of PTHrP on cell growth and viability in osteoblastic MC3T3-E1 cells

The proliferative capacity of MC3T3-E1 cells overexpressing the different PTHrP constructs was examined. As an initial approach, using Alamar Blue assay, we found that the increased proliferation displayed by WT-PTHrP-transfected cells was significantly attenuated in the ΔT¹⁰⁷-PTHrP, ΔNLS-PTHrP, or ΔSP-PTHrP mutants (Figure 5A). We next aimed to confirm and extend these observations by using flow cytometry. As expected, WT-PTHrP-transfected cells displayed a greater percent of cells in S+G₂/M phases of the cell cycle than those transfected with the empty vector (control) in FBS-containing medium (Figure 5B). Interestingly, the proliferative capacity of each PTHrP mutant affecting T¹⁰⁷, NLS, or SP was significantly lower than that of WT-PTHrP cells but still more than the empty vector control (Figure 5B).

Figure 5.

Cell growth and cell viability of MC3T3-E1 cells, overexpressing or not different PTHrP constructs (cells transfected with pcDNA 3.1 vector alone, V). Cell growth at day 2 was estimated by Alamar blue assay (A) or by analysis of the cell cycle using flow cytometry (B). Cell death was evaluated by either Trypan blue exclusion at day 2 (C) or by flow cytometry at 24 hours (D) in FBS-depleted conditions, as described in the text. Some PTHrP-overexpressing cells were transiently transfected with PTH1R siRNA or incubated with the PTH1R antagonist PTHrP(7–34) (B and D). Results are mean ± SEM of at least 3 experiments in triplicate. *, P < .05; **, P < .01 vs V; ^#, P < .05 vs WT.

The putative involvement of the PTH1R for intracellular trafficking and nuclear localization of PTHrP has previously been suggested (6). Thus, we used siRNA technology to suppress PTH1R gene expression, or the PTH1R antagonist PTHrP(7–34), in MC3T3-E1 cells overexpressing intact or mutated PTHrP, and explored its consequence on the cell growth. PTH1R silencing caused almost complete down-regulation of PTH1R mRNA levels, as compared with those in control cells, assessed by real-time PCR (Supplemental Figure 1). PTH1R suppression significantly reduced the mitogenic response in WT-PTHrP-overexpressing cells, compared with that displayed by the other PTHrP mutants studied, which were unaffected by such suppression or by treatment with PTHrP(7–34) (Figure 5B).

We also assessed cell viability in MC3T3-E1 cells overexpressing the different PTHrP constructs. By using Trypan blue staining, WT-PTHrP cells as well as ΔT¹⁰⁷-PTHrP or T107A mutants similarly exhibited lower percent of cell death than control cells cultured for 2 days in serum-depleted conditions (Figure 5C); similar cell protection conferred by all of these PTHrP constructs was observed at day 6 of serum starvation (data not shown). This was confirmed by using flow cytometry to estimate apoptosis after 24 hours of serum depletion (Figure 5D). In contrast, NLS deletion or antagonizing the PTH1R by specific siRNA transfection or addition of PTHrP(7–34) reversed this cell protection conferred by PTHrP overexpression in both experimental settings (Figure 5, C and D). Of interest, SP deletion was similarly effective as the PTH1R siRNA or PTHrP(7–34) inhibiting the antiapoptotic effects observed in WT-PTHrP cells (Figure 5D). In fact, PTH1R silencing or PTHrP(7–34) addition did not further affect ΔSP-PTHrP action on cell apoptosis (Figure 5D).

Relative contribution of NLS- and PTH1R-related pathways in the effects of PTHrP on osteoblastic MC3T3-E1 cell differentiation

Several parameters related to cell differentiation were subsequently determined in MC3T3-E1 cells overexpressing intact or mutated PTHrP. A dramatic increase in matrix mineralization and ALP activity occurred related to PTHrP up-regulation in these osteoblastic cells (Figure 6, A and B). Similar effects were observed in OC mRNA expression, a late osteoblast differentiation marker with a recognized role in bone matrix mineralization (40) (Figure 6C). Up-regulation of these differentiation markers was significantly decreased in the ΔNLS-PTHrP mutant and, in general, attenuated in the ΔT¹⁰⁷-PTHrP mutant (Figure 6, A-C). Of note, matrix mineralization and OC mRNA levels were similarly diminished by either SP deletion or antagonizing the PTH1R, compared with those in WT-PTHrP-overexpressing cells (Figure 6, A and C).

Figure 6.

Cell differentiation in MC3T3-E1 cells after stably transfection with pcDNA3.1 vector alone (V) or the different PTHrP-overexpressing constructs. Matrix mineralization (A) and ALP activity (B) were determined by alizarin S staining and using p-nitrophenyl phosphate as substrate, respectively, at day 15. OC mRNA expression was analyzed by real-time PCR in these cell clones. Some PTHrP-overexpressing cells were transiently transfected with PTH1R siRNA or incubated with PTHrP(7–34) in the course of the experiment (A and C). Results are mean ± SEM of at least 3 experiments in triplicate. *, P < .05;**, P <.01 vs V; ^#, P < .05 vs WT.

Discussion

We here demonstrate that constitutive PTHrP overexpression promotes proliferation and protects from apoptosis in osteoblastic MC3T3-E1 cells. These findings are consistent with previous observations using a similar experimental approach in other normal cell types including chondrocytes, VSMCs, and mesangial cells (10–12, 41), and in prostate and colon cancer cells (13, 42). In the present study, we also show that MC3T3-E1 cells overexpressing PTHrP have an increased osteogenic differentiation capacity. Furthermore, the mitogenic and differentiation responses associated with PTHrP overexpression in these cells were found to be dependent on the presence of an intact NLS in the PTHrP molecule. In this regard, a recent study has shown that knock-in mice lacking the sequence encoding for PTHrP(67–137), comprising the midregion, the NLS, and the C-terminal region, die prematurely with profound skeletal alterations, related to low osteoblast proliferation and decreased expression of major osteogenic genes (including OC) in bone (27). Importantly, exogenous administration of PTHrP(67–139) or the complete PTHrP sequence in vivo or to primary calvaria osteoblasts in vitro failed to rescue these alterations in these mutated mice, suggesting that the missing bony effects of PTHrP in these mice are primarily intracrine (27).

The present findings indicate that, in addition to the NLS, the PTH1R also contributes to the observed actions of PTHrP on osteoblastic function. Our data show that exogenous administration of the PTH1R antagonist PTHrP(7–34) mimics the inhibitory effects of PTH1R silencing in WT-PTHrP cells, suggesting that PTHrP interaction with its membrane receptor via an autocrine/paracrine pathway is a major contributor to PTHrP effects on osteoblastic cells. Moreover, analysis of PTHrP in cytoplasmic and nuclear fractions showed that impairment of PTHrP secretion by SP deletion did not increase PTHrP presence in the nucleus as expected but even decreased it. These results suggest that secretion and probably internalization of PTHrP via the PTH1R is an important step prior to PTHrP nuclear internalization. In this regard, ΔSP-PTHrP may not reproduce WT-PTHrP effects because PTHrP is not as effectively shuttled into the nucleus in the ΔSP-PTHrP mutant as in WT-PTHrP cells due to the absence of autocrine/paracrine PTHrP-PTH1R interactions. Secreted PTHrP would be impaired in the ΔSP-PTHrP mutant, thus preventing PTH1R signaling that would hamper PTHrP transport to the nucleus and function.

The results of this study suggest that PTHrP induces proliferation and matrix mineralization by at least 2 partially redundant mechanisms: an autocrine/paracrine SP/PTH1R-dependent mechanism, and an intracrine NLS-dependent mechanism. Therefore, PTHrP secretion and subsequent activation of its receptor would be able to induce proliferative and mineralization signals in the ΔNLS-PTHrP mutant. In this regard, previous reports have shown that exogenous administration of PTHrP(1–36), osteostatin, or PTHrP(107–139) can increase osteoblast cell proliferation and differentiation both in vivo and in vitro (23, 28, 31). On the other hand, in SP-deleted osteoblastic cells, PTHrP would be less efficient for cell trafficking but still be able to reach the nucleus via the NLS region and thus affect osteoblastic function.

The present results also indicate that osteoblast protection against apoptosis and stimulation of OC gene expression induced by PTHrP seem to require both the PTH1R and NLS. PTHrP appears to need both PTH1R-dependent autocrine/paracrine signaling/trafficking and also NLS-dependent nuclear shuttling to induce the aforementioned events in osteoblastic cells. Some studies favor the idea that the PTH1R, which also contains an NLS domain itself, can act as a chaperone to facilitate PTHrP targeting to the nucleus (15, 43), although other data do not support this notion (44). Of interest, exposure of osteoblastic cells to exogenous PTHrP has been shown to induce PTH1R shuttling from the nucleus to the cytoplasm (45). In this regard, our results using immunofluorescence provide no evidence of PTH1R localization in the nucleus of any of the PTHrP-overexpressing cell mutants. The present findings suggest a cytoplasmic chaperone role rather than a cotransporting function into the nucleus for the PTH1R. Thus, PTHrP would not shuttle directly into the nucleus, but it probably needs to be secreted (requiring an intact SP), internalized via the PTH1R and transported near the nucleus, and then released from the receptor and shuttled into the nucleus via the NLS to promote cell survival and OC expression.

The important role of the NLS for the prosurvival action of PTHrP in osteoblastic cells as found here is consistent with previous observations in other cell types, namely chondrocytes and muscular smooth muscle cells (10–12). Also of note, antiapoptotic effects have been reported in the same osteoblastic cell type used herein, but stably transfected with a PTH1R plasmid construct (35). In addition, other studies have reported that cell protection by PTHrP may be conferred through the autocrine/paracrine pathway in renal cells and other osteoblastic cell types (39, 41, 46).

Consistent with previous findings in osteoblasts and other cell types (47, 48), we here observed PTH1R protein down-regulation in PTHrP-overexpressing MC3T3-E1 cells, related to receptor internalization. On the other hand, this effect was not observed in the SP-deleted mutant, probably due to the lack of PTHrP-PTH1R autocrine/paracrine interactions that trigger desensitization mechanisms involving receptor internalization and down-regulation. Even in a scenario of PTH1R down-regulation, we found an increase of cell proliferation, survival, and differentiation in the WT-PTHrP-overexpressing cells over those in control cells. Thus, the decreased PTH1R levels are unlikely to be responsible for the observed alterations in osteoblast function in these mutants. Furthermore, internalization of the PTH1R is not strictly associated to this receptor desensitization because its internalization appears to prolong, rather than to shorten, receptor signaling (49–50).

The role of T¹⁰⁷ in PTHrP is controversial. This is the first amino acid of the osteostatin sequence (7, 8), and it has also been included in the sequence constituting the NLS domain by most investigators (3, 5, 46, 51) but not by others (19, 20). We here show that deletion or mutation of this amino acid in the PTHrP molecule significantly hampers its stimulatory effects on MC3T3-E1 cell proliferation and several osteoblast differentiation markers, namely ALP activity and OC gene expression. Our present immunofluorescence data suggest that T¹⁰⁷ might contribute to the translocation of PTHrP from the cytoplasm to the nucleus (Figure 3). In this respect, phosphorylation of Thr or Ser residues in the NLS of proteins other than PTHrP has shown to favor nuclear entry (52–53). Also of note, a previous study in VSMCs demonstrates that the C-terminal terminus 108–139 of PTHrP does not interact with its nuclear entry but plays a key transactivating role in its intracrine mitogenic action (19). Thus, the possibility that T¹⁰⁷ could be a putative phosphorylation substrate for an uncharacterized kinase, affecting its nuclear trafficking and/or actions in the nucleus, awaits further studies. In any event, contrary to the observed role of the NLS in the prosurvival effect of PTHrP overexpression, T¹⁰⁷ does not appear to be essential in this regard in osteoblastic MC3T3-E1 cells.

In conclusion, the present study reveals the complexity of the mechanisms of action of PTHrP in osteoblastic cells. Our in vitro findings suggest that PTHrP acts on osteoblastic growth and function in a mixed model in which this protein needs to be secreted and internalized via the PTH1R (autocrine/paracrine pathway) before NLS-dependent shuttling to the nucleus (intracrine pathway). A proposed model for different autocrine, paracrine, and intracrine mechanisms whereby PTHrP may affect osteoblast cell functions is depicted in Figure 7.

Figure 7.

Proposed model for autocrine, paracrine, and intracrine mechanisms whereby PTHrP can affect osteoblast cell functions. Three different pathways are shown: 1) Mixed pathway: SP, PTH1R, and NLS-dependent mechanism; 2) autocrine-paracrine; SP and PTH1R-dependent mechanism; 3) intracrine; NLS-dependent mechanism. Cytoplasmic and/or nuclear effectors would act as transducers of PTHrP actions through these different pathways. PTHrP is depicted with its main domains: N- and C-terminal (term) domains and NLS.

Funding Statement

This study was supported by grants from Spanish Instituto de Salud Carlos III (PI050117, RD06/0013/1002, RD12/0043/0008, PI11/00449) and Ministerio de Educación (SAF2005-05254). A.G.-M., J.A.A., D.L., and S.P.-N. are recipients of a postdoctoral research contract from Ministerio de Ciencia e Innovación-Juan de la Cierva program (JCI-2009-04360 and JCI-2011-09548), Comunidad Autónoma de Madrid (S-2009/Mat-1472), and RETICEF, (RD06/0013/1002 and RD12/0043/0008), respectively. M.M. and A.L.-H. were supported by Ministerio de Economía y Competitividad (FI12/00458) and Ministerio de Educación-FPU program (AP2009-1871), respectively, and Fundación Conchita Rábago.