To promote their survival and progression in the skeleton, osteotropic malignancies of breast, lung, and prostate produce parathyroid hormone–related protein (PTHrP), which induces hypercalcemia. PTHrP serum elevations have also been described in multiple myeloma (MM), although their role is not well defined. When we investigated MM cells from patients and cell lines, we found that PTHrP and its receptor (PTH-R1) are highly expressed, and that PTHrP is secreted both as a full-length molecule and as small subunits. Among these subunits, the mid-region, including the nuclear localization sequence (NLS), exerted a proliferative effect because it was accumulated in nuclei of MM cells surviving in starvation conditions. This was confirmed by increased transcription of several genes enrolled in proliferation and apoptosis control. PTHrP was also found to stimulate PTH-R1 in MM cells. PTH-R1's selective activation by the full-length PTHrP molecule or the NH2-terminal fragment resulted in a significant increase of intracellular Ca2+ influx, cyclic adenosine monophosphate (cAMP) content, and expression of receptor activator of NF-κB ligand (RANKL) and monocyte chemoattractant protein-1 (MCP-1). Our data definitely clarify the role of PTHrP in MM. The PTHrP peptide is functionally secreted by malignant plasma cells and contributes to MM tumor biology and progression, both by intracrine maintenance of cell proliferation in stress conditions and by autocrine or paracrine stimulation of PTH-R1, which in turn reinforces the production of osteoclastogenic factors. © 2014 American Society for Bone and Mineral Research.
Parathyroid hormone-related protein (PTHrP) is functionally analogous to the parathyroid hormone (PTH), a major regulator of calcium homeostasis and an osteoclast (OC) activator. These peptides show high homology within their NH2-domain in the 1–34 amino acid (aa) sequence, by which they bind the same G-protein receptor (PTH-R1).[2, 3] During fetal development PTHrP is produced by perichondral cells to adjust placental calcium levels and support the differentiation of endochondral cells to chondrocytes,[4, 5] whereas after birth it concurs to development of mammary glands and lactation. The predominant activity of PTHrP in bone metabolism has been definitely proved in homozygous PTHrP-deficient mice that show abnormal bone morphogenesis, leading to osteopetrosis and short-term survival after birth.[7, 8]
PTHrP has been primarily reported in patients with osteotropic malignancies, such as breast, lung, and renal cancers, who show skeletal disease and hypercalcemia.[9-12] Subsequent studies have described its functional effects through either paracrine or autocrine cell stimulation by tumors with no skeleton involvement and even by several normal tissues.[13-15] An intracrine effect of PTHrP has been also postulated in breast cancer to improve the nucleocytoplasmic trafficking that accelerates cell cycle progression, RNA transcription, and transport.[16-18] Structurally, PTHrP contains three functional domains, namely NH2-terminal, mid-region, and COOH-terminal fragments. Similar to PTH, the NH2-terminal subunit (aa 1–36) avidly links PTH-R1, by which it regulates calcium channels promoting the influx of extracellular Ca2+, and activates both adenylate-cyclase protein A kinase (PKA) and phospholipase Cβ protein kinase (PKC) pathway signals. The mid-region (aa 38–106) has been reported to contain the nuclear localization sequence (aa 87–94) (NLS) that binds DNA in nuclei with potential effects on cell proliferation, because NLS−/− mice show accelerated apoptosis in most tissues, which is associated with delayed growth, senescence, and early death.[19-21] Last, the COOH-terminal subunit (aa 107–139), also known as osteostatin, has been described as a negative regulator of signals induced by the NH2-terminal region and an inhibitor of the osteoclast activity in rats. The functional role of PTHrP fragments in human pathophysiology has been intensively investigated and their generation by posttranslational proteolysis of members of the subtilisin family, has been definitely assessed.[23-25]
Tumors producing skeletal metastases are usually associated with high PTHrP activity that apparently correlates with poor prognosis in both breast and prostate cancers, although a survival advantage has been unexpectedly described in female patients with lung adenocarcinoma and serum elevation of PTHrP. Nevertheless, the role of PTHrP on cancer growth has been definitely confirmed by demonstrating that its suppression, by either neutralizing antibodies in MDA-MB-231 breast cancer cells or by antisense oligonucleotides in mouse models, restrains tumor progression, resulting in marked reduction of tumor size.[27, 28]
Skeleton devastation due to hyperactive osteoclastogenesis is a hallmark of multiple myeloma (MM), and PTHrP has been postulated to concur to myeloma bone disease because it induces the formation of osteolytic lesions.[29, 30] Also, high PTHrP serum levels have been reported in MM, in particular in patients with overt skeleton involvement, although its cellular origin and its specific role has not yet been defined.[31, 32] Based on the interference of malignant plasma cells with other components of the MM marrow microenvironment, leading to enhanced production of OC-activating factors as interleukin 1β (IL-1β), macrophage inflammatory protein 1α (MIP-1α), and receptor activator of NF-κB ligand (RANKL), several authors have hypothesized that MM cells may also directly produce PTHrP for their own activity,[31, 32, 34] although this hypothesis is not supported by either in vitro or in vivo studies. In this context, we have previously proved that, in MM, PTHrP acts as a suppressor of bone remodeling by inducing the expression of E4BP4, a transcriptional repressor gene for osteoblasts (OBs).
Here, we analyze the multifaceted effect of PTHrP in MM bone disease by providing evidence that this protein is produced as a full-length molecule as well as in small subunits by malignant plasma cells. In particular, we demonstrate that the NLS fragment is detectable in nuclei of proliferating MM cells and contributes to their survival, whereas the NH2-terminal fragment enhances the pro-OC tumor activity.
Patients and Methods
MM plasma cells and cell lines
Fourteen patients, 10 with stage IIIB newly diagnosed MM and four with monoclonal gammopathy of unknown significance (MGUS), were enrolled in the study, which was approved by the Ethical Committee of the University of Bari. Patients with MM provided their biological materials before any anti-MM treatments, and serum calcium levels were routinely assessed in parallel with PTHrP measurement by commercial ELISA (Phoenix Pharmaceuticals, Burlingame, CA, USA).
Primary MM plasma cells were isolated by magnetic sorting from iliac crest marrow aspirates, using anti-CD138 monoclonal antibody (mAb)-conjugated microbeads (Stem Cell Technologies, Vancouver, BC, Canada). In addition, five MM cell lines (ALB, MUS, PAS, MCC-2, and GRA) established in our laboratory, along with both U266 and RPMI 8226 MM cell lines, were investigated in parallel with a renal cell carcinoma cell line, RCC-1, for control experiments. Further cell controls included normal OBs obtained from bone biopsies of two healthy individuals undergoing posttraumatic orthopedic surgery.
PTHrP and PTH-R1 protein
Both PTHrP and PTH-R1 expression were evaluated by flow cytometry using phycoerythrin (PE) anti-CD138 (Beckman Coulter, Fullerton, CA, USA), fluorescein isothiocyanate (FITC) anti-PTHrP (full-length and NH2-terminal; Santa Cruz Biotechnology, Santa Cruz, CA, USA; cod. sc-9680), and FITC anti-PTH-R1 (Santa Cruz Biotechnology) mAbs. Cytoplasmic full-length and/or NH2-terminal PTHrP was also investigated by double fluorescence using UV microscopy visualization. Briefly, MM cells were fixed in 4% paraformaldehyde (Sigma-Aldrich, Milan, Italy), permeabilized with 0.1% Triton X-100 (Sigma-Aldrich), and treated with goat anti-PTHrP (Santa Cruz Biotechnology; cod. sc-9680) and FITC anti-goat rabbit antiserum (Sigma-Aldrich). Nuclei were stained with 0.1 μg/mL 4′,6-diamino-2-phenylindole dihydrochloride (DAPI). To visualize PTHrP nuclear translocation by proliferating MM cells, we completed additional experiments using RPMI 8226, both unstimulated and stimulated with 1 nM PTHrP mid-region including NLS (aa 38–94), in lowering serum concentration cultures, namely 10% to 0% of fetal calf serum (FCS), as described. Subsequent analysis was completed by confocal microscopy (DM IRE2; Leica Microsystems, Wetzlar, Germany) after incubation with rabbit anti-PTHrP mid-region (aa 34–53) (Calbiochem, Merck Chemicals Limited, Nottingham, UK; cod. PC09), secondary antibody and nuclei staining with 1 mM TO-PRO-3 Iodide (Invitrogen, Milan, Italy).
PTHrP and PTH-R1 cell expression was also assessed by Western blot (WB) using 10 µg of cell lysate from freshly derived MM cells, and from several MM cell lines and RCC-1. Because PTHrP is cleaved at the posttranslational level in multiple functional subunits, we used two different antibodies in the WB experiments. In particular, the goat anti-PTHrP (Santa Cruz Biotechnology; cod. sc-9680) recognizes full-length and NH2-terminal PTHrP (aa 1–34), whereas the rabbit anti-PTHrP (Calbiochem, Merck Chemicals Limited; cod. PC09) is specific for mid-region PTHrP (aa 34–53). Briefly, blotted polyvinyl difluoride (PVDF) membranes (Bio-Rad, Milan, Italy) were incubated with goat or rabbit anti-PTHrP and mouse anti-PTH-R1 mAbs (Santa Cruz Biotechnology), then with secondary antibodies and finally revealed by enhanced chemiluminescence (ECL Plus; GE Healthcare, Milan, Italy). In addition, WB with rabbit anti-PTHrP was used to evaluate the intracellular content of NLS+-mid region PTHrP in both RPMI 8226 and MUS cells cultured with 10% FCS or in the serum-starved condition, and the relative quantification of NLS+-mid region PTHrP was assessed by ImageJ Software (NIH, Bethesda, MD, USA).
Also, soluble full-length and/or NH2-terminal PTHrP released by 72-hour cultured cells was measured by enzyme immunoassay (Phoenix Pharmaceuticals Inc., Burlingame, CA, USA) in 100-fold concentrated supernatants from MM cell lines and RCC-1. Serum levels of PTHrP were also measured in patients enrolled in the study and in normal donors.
Total RNA was extracted from MM cell samples and cell controls using the RNAeasy Mini kit (Qiagen, Chatsworth, CA, USA). To obtain first strand cDNA, 1 μg of total RNA was treated with High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA, USA) by a Mastercycler Personal (Eppendorf SRL, Milan, Italy). PTHrP and PTH-R1 levels were assessed in primary MM cells from patients 2, 4, and 7 as well as in MM cell lines and control RCC-1 cells. Expression levels of RANKL, MIP-1α, and monocyte chemoattractant protein-1 (MCP-1) were assessed on MUS, RPMI 8226, U266 MM cell lines, and on control OBs after incubation with PTHrP NH2-region (aa 1–40) at 100 nM for 48 hours. RNA levels of AKT1, Bcl-2, Bcl-xl, Cyclin D1, and Ki67 were also measured in RPMI 8226 after 48 hours of incubation with 100 nM of PTHrP NLS+ (aa 38–94) in 10% FCS-supplemented cultures, as well as in serum-free cultures to evaluate the effect of NLS in starved cells. β-actin was the control gene for real-time PCR by Sybr Green technology (SYBR Green/ROX qPCR Master Mix; Fermentas, Burlington, Ontario, Canada) using an ABI Prism 7300 Sequence Detector (Applied Biosystems). After initial incubation at 50°C for 2 minutes, samples were denatured at 95°C for 10 minutes, followed by 40 sequential cycles at 95°C for 15 seconds, and 1 minute at 60°C. All samples were analyzed in triplicate and the delta-delta comparative cycle threshold algorithm (2−ΔΔCt) was used to evaluate relative gene expression levels; the primers are listed in Table 1.
|Gene symbol||GenBank accession number||Forward primer 5′>3′||Reverse primer 5′>3′||Amplicon size (bp)|
Effect of PTHrP on proliferation
Full-length PTHrP (aa 1–139), NH2 (aa 1–40), mid-region (aa 67–86) (Bachem Peptides, Rome, Italy), aa 38–94 (kindly provided by Dr. Doug Burton, Boston, MA, USA), and COOH-terminal (aa 107–139) fragments were tested for their proliferative activity. The mid-region (aa 67–86) was used as the NLS− fragment, whereas the aa 38–94 subunit included NLS (NLS+). To this end, MM patient 4, MUS, RPMI 8226, and RCC-1 were incubated with each PTHrP subunit at 100 nM for 24 and 48 hours. Cell viability and proliferation were measured by 3-(4,5-dimethylthiazol-2yl)-5-(3-carboxymetho-xyohenil)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega, Madison, WI, USA). The formazan product, measured by absorbance at 490 nm in a 96-well plate reader (Benchmark; Bio-Rad, Hercules, CA, USA) reflected the extent of living cells by triplets of experiments. The proliferation extent under starvation conditions was measured by sequential serum deprivation (10%, 8%, 4%, 2%, and 0% FCS) in parallel cultures.
Activation of PTH-R1
PTHR-1 activity was investigated on marrow MM cells from patient 4 and from MUS, U266, and RPMI 8226 cells; values were levels of intracellular Ca2+ flux and cyclic adenosine monophosphate (cAMP) accumulation in response to 100 nM of PTHrP full-size molecules (aa 1–139) and the following subunits: NH2-terminal (aa 1–40); mid-region of either NLS− (aa 67–86) or NLS+ (aa 39–94); and COOH-terminal (aa 107-139) fragments. Briefly, intracellular Ca2+ flux was measured each minute up to 10 minutes by flow cytometry with FACSDiva software (Becton-Dickinson, Milan, Italy) after stimulation, staining with Fluo3-AM dye (Sigma-Aldrich) and detection of mean fluorescence intensity (MFI). Control experiments of Ca2+ flux activation were also completed in parallel cell cultures pretreated with 10 µM of nifedipine as Ca2+ inhibitor and results were expressed as mean values from triplets of experiments. Time-dependent variation of cAMP was calculated by competitive immunoassay (Assay Designs Inc., Ann Arbor, MI, USA); the colorimetric reaction at 405 nm (Bio-Rad) was measured. Each assay was completed in triplicate and absorbance was converted into pM/mL using a standard curve.
Measurement of cytokine expression
MUS, RPMI 8226, and U266 cells were incubated for 48 hours with PTHrP NH2-terminal fragment at 100 nM and subsequently measured by real-time PCR for their expression of RANKL, MIP-1α, and MCP-1. Each cytokine was also quantified as secreted protein in relative supernatants by commercial ELISAs (Immunological Sciences, Rome, Italy) and absorbance values at 450 nm were converted in pg/mL using the Microplate Manager 4.0 software (Bio-Rad). Moreover, cytoplasmic content of RANKL and MCP-1 were revealed by flow cytometry after cell fixation and permeabilization with Fix and Perm (An Der Grub Bio Research GMBH, Kaumberg, Austria). Briefly, 2 × 105 cells were incubated with primary antibody for 30 minutes at 4°C, then washed and treated with secondary antibody (Sigma-Aldrich).
PTH-R1 gene silencing
The role of PTH-R1 on RANKL, MCP-1, and MIP-1α induction and modulation was investigated by silencing PTH-R1. To this, 5 × 105 cells were incubated in a six-well tissue culture plate with a pool of three target-specific 20-nt to 25-nt small interfering RNAs (siRNAs) designed to knock down PTH-R1 (Santa Cruz Biotechnology) using a scrambled sequence as control. Efficiency of silencing was monitored by real-time PCR after 24 hours, and after the PTHrP NH2-fragment (aa 1–40) stimulation, RANKL, MIP-1α, and MCP-1 levels were also revealed by real-time PCR.
Pro-OC activity of PTHrP-treated myeloma cells
To explore the osteoclastogenic potential of PTHrP-treated myeloma cells, we purified peripheral CD14+ monocytes by immunomagnetic technique from healthy donors and cocultured these cells with PTHrP NH2-fragment-treated RPMI 8226 cells. Also, we incubated parallel preparations of CD14+ monocytes with RANKL (Provitro GmbH, Berlin, Germany; 100 ng/mL) and macrophage colony-stimulating factor (M-CSF; ORF Genetics, Kopavogur, Iceland; 30 ng/mL) as positive control, or with untreated RPMI 8226 as negative control. Osteoclast formation was evaluated in relation to the evidence in culture of large multinucleated cells by light microscopy after May-Grünwald-Giemsa staining that were positive to specific tartrate-resistant acid phosphatase (TRAcP) enzymatic reaction (Sigma Aldrich). These cells were numbered with respect to the surface of culture as cells/cm2 and each coculture was completed using three different samples of monocytes from normal donors.
Differences between means of data were calculated by both Student's t and Mann-Whitney tests and p values <0.05 were considered significant.
Calcium and PTHrP serum levels
Hypercalcemia occurred only in patient 4, in parallel with the highest serum PTHrP level detected (28.4 pM/L). PTHrP levels in the other MM patients were variable, except in patients 2, 5, and 8. The patients with MGUS showed the lowest levels (1.0–2.9 pM/L) at amounts comparable to normal donors (Table 2).
|Patients||PTHrP concentration (pM/L)|
Malignant plasma cells express PTHrP and PTH-R1
Figure 1A shows both PTHrP and PTH-R1 flow cytometry mean levels by cell preparations. Those from MM patients were significantly higher for both PTHrP-positive and PTH-R1-positive cells than those from MGUS patients (p < 0.05 in both instances), which were our negative controls, for the absence of the bone disease. A major difference was observed in the levels of the PTHrP-positive population. They were as high as around 40% in primary cells from MM patients; ie, up to fivefold higher than in those with MGUS (p < 0.03). Notably, primary plasma cells from both MGUS and MM patients expressed lower percentages of positive cells as compared to MM cell lines which uniformly showed higher levels of both PTHrP and PTH-R1. In particular, MUS, U266, and RPMI 8226 showed percentages of PTHrP-positive cells similar to those of RCC-1 control cells (>90%), with similar elevations of PTH-R1-positive populations in both U266 and RPMI 8226 cell lines.
Intracellular content of PTHrP was also variably detected by double immunofluorescence using goat anti-PTHrP mAb and DAPI, and Fig. 1B (left) depicts a representative pattern of cytoplasmic detection of PTHrP in RPMI 8226 cells, whereas flow cytometry expression of PTH-R1 is representatively shown (right) for primary MM cells (patient 4), MM (MUS), and control (RCC-1) cell lines.
The WB detection of both PTHrP and PTH-R1 in these cell preparations is represented in Fig. 1C; both proteins were revealed in each cell sample. In particular, PTHrP was detected in different isoforms as, in addition to the 18-kDa band related to the full-length molecule (aa 1–139), other bands of lower size (∼10 to 3 kDa) were revealed in all cell lysates. Furthermore, by using rabbit anti-mid-region PTHrP, we specifically revealed a single band of approximately 7 kDa related to the mid-region, as described for other tumors.[9-11, 23-25]
We also measured RNA levels of both PTHrP and PTH-R1 (Fig. 1D) in selected plasma cell samples, based on the highest cytometric expression. Transcription of both genes was variable and in apparent agreement with flow cytometry data. Notably, primary MM plasma cells from patient 4 showed the highest level of PTHrP RNA with respect to the other patients, whereas the MM cell lines expressed comparable levels of both PTHrP and PTH-R1.
Finally, measurement of in vitro PTHrP production by MM cell lines is reported in Table 3. Again, variable amounts of the peptide were detected by different cell cultures the highest levels occurred in RPMI 8226, U266, and MUS supernatants, at a magnitude almost equivalent to the RCC-1 control cell line.
|Cell lines||PTHrP concentration (ng/mL)|
These data provided evidence that primary plasma cells from patients with either MGUS or MM, as well as MM cell lines, express variable amounts of PTHrP and PTH-R1, and that PTHrP is produced by cultured MM cells in form of both full-length molecule and smaller subunits.
PTHrP induces proliferation in MM cells by NLS
By using the PTHrP full-length form (aa 1–139) as well as the NH2-terminal (aa 1–40), NLS− mid-region (aa 67–86), and C-terminal (aa 107–139) fragments, we observed in all cell preparations, including primary MM cells from patient 4, MUS and RPMI 8226 cells, and RCC-1, a minimal, almost undetectable effect as compared to basal values of untreated cells (p > 0.5), both in 24-hour and 48-hour cultures (Fig. 2A). By contrast, the NLS+ mid-region fragment (aa 38–94) induced evident proliferation effect in almost all samples both after 24 hours and 48 hours of culture, with significant increase of the proliferative rate after 48 hours in all instances, compared to other PTHrP fragments (p < 0.05).
Because PTHrP exerts a role in cell survival under serum-starvation conditions, we tested the full-length molecule and relative fragments on MUS, RPMI 8226, and RCC-1 cultured with decreasing concentration of FCS. Our MTS assay confirmed that in contrast with other molecular forms of PTHrP, the NLS+ mid-region (aa 38–94) was effective in maintaining the proliferative rate, allowing cell survival at concentrations of FCS lower than 10% and even in serum-free medium (Fig. 2B). Indeed, levels of absorbance remained almost unaltered with minimal decrease in cultures with no FCS, thus confirming the effect of the NLS+ mid-region on MM cells in a fashion similar to other cancers.
To verify this effect, we also inspected the localization of an exogenously added fragment (aa 38–94) in proliferating cells both in 10% FCS and in serum-free cultures. Based on relative intracellular minor concentration, we failed to detect by immunofluorescence the nuclear localization of the endogenous mid-region fragment. Figure 2C (top), shows that NLS+ mid-region increased presence in nuclei of serum-starved RPMI 8226 cells (right) as compared to those normally cultured with complete medium (10% FCS; left). We thus interpreted its major content in nuclei of starved cells as functional for cell proliferation and survival. To validate this hypothesis, we analyzed by WB the relative content of NLS+ mid-region in both RPMI 8226 and MUS cultured either in 10% FCS or in serum-deprived medium. The ImageJ software analysis revealed that the NLS+ mid-region content in serum starved cells was approximately twofold and 10-fold higher for RPMI 8226 and MUS, respectively, if compared to control 10% FCS cultured cells (Fig. 2C, bottom). Furthermore, we evaluated in these cells the transcription levels of genes enrolled in the cell cycle, such as AKT1, Bcl-XL, and Cyclin D1. Real-time PCR data, shown in Fig. 2D, supported the evidence that NLS+ mid-region-treated cells upregulate the transcription of those genes in starvation stress conditions, thus proving the NLS+ mid-region involvement in cell proliferation and apoptosis control (p < 0.02).
PTH-R1 is functional in MM cells
We next addressed the question of whether PTH-R1 was responsive to PTHrP in MM cells. Thus, after stimulation with PTHrP, we measured both intracellular Ca2+ flux and cAMP content as receptor activity.
Figure 3A shows the Ca2+ influx variations in primary MM cells from patient 4, MUS, U266, and RPMI 8226 cell lines. All cell preparations underwent fast increase of intracellular Ca2+ flow within 4 to 5 minutes with maximum influx after 8 minutes, as compared to basal values of the Fluo3-AM dye uptake. However, this occurred only when the cells were stimulated by either PTHrP in its full-length form, or as the NH2-terminal fragment (aa 1–40). Conversely, no effect was recorded by other fragments, because in all instances the Ca2+ influx (left) was similar to values of unstimulated cells, as well as to control cells Ca2+-inhibited by nifedipine (right).
However, the size of Ca2+ flow increase induced by the PTHrP full-length molecule and the NH2-fragment (aa 1–40) was apparently dissimilar between MM cell preparations, as compared to basal levels. Among the cell lines, MUS showed high basic activity (MFI ≅ 1400) that rapidly increased (MFI ≅ 1900), whereas PTH-R1 basic activation in other cell lines was apparently lower (MFI ≅ 1000). This aspect reveals a putative basilar activity of the receptor in all cell lines that appears to be independent of PTHrP exogenous stimulation.
Data related to cAMP produced in response to PTHrP are included in Fig. 3B. As can be seen, only full-length PTHrP (aa 1–139) and its NH2-fragment (aa 1–40) provoked a significant increase of cAMP in samples, with levels of both primary MM cells from patient 4 and MM lines becoming up to about three times as high as their basal cAMP content (equivalent to 3 pM/mL, p < 0.02 in all instances). By contrast, as observed in Ca2+ influx analysis, no effect was revealed by stimulating the same cell samples with the other PTHrP fragments, thus suggesting that PTH-R1 expressed by MM plasma cells is sensitive only when the PTHrP molecule including the NH2-terminal fragment (aa 1–40) acts as external stimulator of cells, as occurs in physiology.
PTHrP NH2 fragment (aa 1–40) enhances RANKL and MCP-1 in MM cells
Because MM cells were apparently sensitive to the PTHrP NH2-terminal fragment (aa 1–40), we explored its effect in inducing osteoclastogenic molecules such as RANKL, MIP-1α, and MCP-1 that are physiologically produced by OBs in response to PTH-R1 activation. Figure 4A shows the variation of relative RNAs after stimulation with PTHrP NH2-terminal subunit. A dramatic increase of both RANKL and MCP-1 was revealed, contrasting with only a minor effect in MIP-1α transcription. The highest increment was detected in RANKL transcription on RPMI 8226 (2−ΔΔCt: 4.281 ± 0.4), whereas lower though still significant values were revealed in other myelomas (p < 0.05). Similarly, high MCP-1 RNA was detected in all MM cell lines, whereas MIP-1α transcription was moderately recorded in MUS cells (2−ΔΔCt: 1.856 ± 0.5), contrasting with almost unchanged values in other cell lines. These results were validated by increased transcription levels of these genes in control OB preparations after PTHrP NH2-fragment (aa 1–40) stimulation (p < 0.05). In parallel, ELISA assays for soluble RANK-L, MCP-1, and MIP-1α released by the PTHrP-NH2–stimulated MM cells showed corresponding data. In fact, we observed an increased secretion of RANK-L (RPMI 8226: 65.4 ± 7.9 pg/mL versus 12 ± 5.8 pg/mL; U266: 42.7 ± 2.5 pg/mL vs 15.7 ± 4.2 pg/mL; MUS: 53.9 ± 3.4 pg/mL vs 17.8 ± 5.7 pg/mL) and MCP-1 (RPMI 8226: 32.6 ± 6.2 pg/mL vs 12 ± 5.8 pg/mL; U266: 47.7 ± 1.8 pg/mL vs 16.2 ± 2.5 pg/mL; MUS: 60.9 ± 4.5 pg/mL vs 15.2 ± 3.7 pg/mL), compared to untreated cells, in contrast with the MIP-1α secretion that was almost similar in all instances. The increased RANKL and MCP-1 expression was confirmed by flow cytometry (Fig. 4B), which revealed in all instances its upregulation after the stimulation with the PTHrP NH2-terminal subunit (aa 1–40) as compared to relative basal values. Increase of RANKL was particularly evident in RPMI 8226, whose values were more than three-fold higher after stimulation (95% of positive cells), whereas both U266 and MUS showed a minor enhancement of RANKL expression (p < 0.05). Similarly, MCP-1 was variably upregulated in treated cells and reached the highest value in RPMI 8226 (75.4% ± 5.7%). These results suggest that stimulation of PTH-R1 by PTHrP NH2-subunit (aa 1–40) reinforces the production of RANKL and MCP-1 by malignant plasma cells in a similar fashion to what occurs with OBs and stromal cells.
This PTH-R1 response to PTHrP NH2-fragment (aa 1–40) was also verified on RPMI 8226 cells silenced in their receptor expression by the small interfering RNA (siRNA) technique (30% versus 95.4% in untreated cells; and 90.2% in scrambled-treated cells). Thus, once stimulated with the same PTHrP subunit, these cells showed reduced RNA expression of both RANKL and MCP-1, as compared to control cells (Fig. 4C), and this reduction was also evident for both RANKL and MCP-1 secretion detected by ELISA in PTHrP (aa 1–40)-treated MM cells (data not shown). These findings proved the effectiveness of PTHrP, because both the full-length molecule (aa 1–139) and the NH2-subunit (aa 1–40), in activating the PTH-R1 pathway that in cells of the mesenchymal lineage promotes the secretion of osteoclastogenic factors.
To verify the functional role of osteoclastogenic factors released by myeloma cells in response to PTHrP stimulation, we cocultured OC progenitors as CD14+ peripheral monocytes with either PTHrP-treated or untreated RPMI 8226 cells. Results are reported in Fig. 5A that shows the mean values of OC maturation in response to PTHrP-treated myeloma cells as well as to other stimuli for control. As can be seen, these preparations of PTHrP-stimulated cells exerted by themselves a moderate pro-OC differentiation effect if compared to parallel controls using PTHrP-untreated RPMI 8226 cells. We interpreted this result as dependent on the increased expression of both RANKL and MCP-1 by the peptide. However, the OC differentiation potential was significantly enhanced by adding M-CSF to those cultures (p < 0.02). Cocultures of monocytes with PTHrP-untreated RPMI 8226 cells also showed enhancement of the OC mean number that, however, was approximately one third lower if compared to the correspondent value in cocultures using PTHrP-stimulated cells (p < 0.03). Representative images of these experiments are presented in Fig. 5B in which both May-Grünwald-Giemsa and TRAcP staining show the enriched presence of OC-like multinucleated cells in cocultures of monocytes incubated with PTHrP-treated RPMI cells in the presence of M-CSF.
These results definitely supported the reinforced pro-OC effect induced by the increased expression of osteoclastogenic factors conditioned by PTHrP in myeloma cells.
Despite different functions of PTHrP in solid osteotropic cancers, as an osteoclastogenic factor promoting bone metastasization and hypercalcemia, a survival factor for tumor cells or proliferation activator, very little information is available regarding its role in MM. Previous studies reported serum elevations of the peptide in MM patients showing hypercalcemia and skeleton devastation,[27, 31] whereas other work postulated a potential paracrine effect of PTHrP secreted by malignant plasma cells to reinforce the production of RANKL by marrow stromal cells, thus contributing to the progression of the bone disease. Here, we provide evidence that PTHrP is secreted in multiple soluble forms by malignant plasma cells and we also identified different functions for these bioactive fragments that apparently act as either intracrine or paracrine stimulators of tumor cells, in addition to their putative activity on stromal cells.
By investigating marrow plasma cells from patients and MM cell lines, we found high PTHrP expression and secretion by tumor cells as either full size molecule of 18 kDa (aa 1–139), or small subunits of 10 to 3 kDa. In addition, we revealed elevated PTHrP levels in sera from MM patients included in the study, whereas serum PTHrP in MGUS subjects was comparable to normal donors.
Because it has been reported that the molecule is produced and cleaved in multiple fragments, sequentially conserved in humans, of similar size by breast, lung and prostate cancer cells to exert differential functions,[9-11, 23-25] we assumed that in our analysis these proteolytic peptides of molecular weight lower than 18 kDa corresponded to those active subunits. Thus, we tested on MM cells the effects of full-length PTHrP (aa 1–139), NH2-terminal (aa 1–40), NLS− mid-region (aa 67–86), NLS+ mid-region (aa 38–94), and COOH-terminal (aa 107–139) fragments. We found that, in contrast with other molecular forms, the NLS+ fragment restores both cell survival and proliferation in absence of growth factors, whereas the NH2-fragment enhances, in a paracrine fashion, the production of osteoclastogenic factors as RANKL and MCP-1 by MM cells, thus emphasizing the role of PTHrP in MM tumor progression.
A major result of our work concerns the evidence that MM cells, similarly to other osteotropic cancers, produce PTHrP, detectable as bioactive forms to exert different functional effects. Indeed, the NLS+ mid-region (aa 38–94) is proteolytically degraded in cytoplasm and internalized by nuclei where this fragment primes in an intracrine fashion the proliferation of MM cells by its complementarity to the 5′-GAGTAGAATTCTAATATCTC-3′ DNA sequence. Other authors, however, have suggested that, to induce a similar effect, the NLS subunit may be transported and internalized in nuclei by cytoplasmic importin-β and that this mechanism is protective for tumor cells in preventing their apoptosis or anoikis induced by starvation.[41, 42, 46] By contrast, an opposite proapoptogenic effect has been ascribed to the NLS+ mid-region (aa 38–94) in cell kinetic analysis in breast cancer. We demonstrated in MM cells a definite proliferative effect of this subunit in the serum starvation culture condition and we assumed that starved cells are capable of progressively activating several cytoplasmic transporters, such as importin-β, to localize the NLS fragment in nuclei for their own cell survival. This interpretation was supported by the increased transcription of major antiapoptosis or proliferation genes such as AKT1, Bcl-XL, and Cyclin D1 in NLS-supplemented starved cells. However, comparable upregulation of these genes was not detected in serum-replete cultured myeloma cells after treatment with NLS+ mid-region (aa 38–94) in relation to the constitutive overexpression of surviving genes in cells actively proliferating in the presence of growth factors. In this context, our data are in agreement with other studies demonstrating the dramatic reduction of tumor growth after PTHrP ablation in a PyMT-MMTV breast cancer mouse model bearing a genetic manipulation in the mammary epithelium; a manipulation that was associated with delayed cancerogenesis, inhibition of tumor progression, defective capability of bone metastasization, and reduced expression of AKT1, Ki67, Bcl-2, Cyclin D1, and Factor VIII. Furthermore, PTHrP blocking antibodies administered to mice resulted in reduction of tumor growth, thus allowing the authors to suggest PTHrP as a target for novel therapies.
Another remarkable point of our study is related to the function of PTH-R1 in MM cells. Previous studies demonstrated the inverse expression of PTHrP and PTH-R1 in MM cell lines as well as the increased secretion of PTHrP after stimulation with transforming growth factor β (TGF-β), although no functional correlation has been hypothesized between PTHrP and PTH-R1. That PTH-R1 is functional on malignant plasma cells has been shown by anecdotic observation describing the acceleration of tumor progression in MM patients receiving PTH for treatment of osteoporosis. It has also been reported that PTHrP activates PTH-R1 in relation to the homology of its NH2-terminal domain with PTH and that receptor stimulation by the PTHrP subunit (aa 1–40) results in both intracellular PKA and PKC activation. We confirmed the PTH-R1 sensitivity to PTHrP in MM cells and demonstrated that the receptor is specifically susceptible to the NH2-terminal fragment (aa 1–40) because, in all instances, we recorded an increased influx of Ca2+ and cAMP content. This bioactivity was not observed using the mid-region fragments, even those including NLS, nor using osteostatin, whose receptor has not yet been identified.
Therefore, because PTH-R1's binding by either PTH or PTHrP in PTH-R1–expressing cells of mesenchymal origin (such as OBs) leads to the activation of both RANKL and MCP-1 production,[43, 52] we measured the production of these osteoclastogenic factors and their increase in the tumor microenvironment from malignant plasma cells, as reported.[53, 54] We also measured MIP-1α as an additional pro-OC factor associated with hypercalcemia in MM.[55, 56] In particular, we found that both RANKL and MCP-1 are greatly increased by all MM cells in response to stimulation with the NH2-terminal fragment, thus supporting the activation of the PTH-R1 pathway in response to the peptide. The direct involvement of PTH-R1 in reinforcing both RANKL and MCP-1 production was confirmed by their enhanced RNA transcription in cultured RPMI 8226, MUS, and U266 cells, and by the opposite effect after PTH-R1 silencing. Finally, this RANKL upregulation was sufficient in vitro, in the presence of M-CSF, to significantly increase the differentiation of monocytes to OCs. Therefore, at least in our experimental model, stimulation of malignant plasma cells by the PTHrP NH2-terminal fragment definitely upregulates the secretion of pro-osteoclastogenic cytokines from tumor cells and contributes to the formation of osteolytic lesions.
A striking point of our study is related to the double functional effect of PTHrP produced by MM cells. Both the intracrine and the paracrine activities of the PTHrP peptide drive cell functions that are essential for tumor progression. The release of soluble PTHrP is likely promoted by tumor cells to maintain their capacity for survival in stress conditions through nuclear localization. By contrast, the paracrine and/or autocrine stimulation of PTH-R1, together with other osteoclastogenic factors produced by tumor activated stromal cells, is essential to reinforce the pro-osteoclastogenic activity of malignant plasma cells in their metastasization program. However, in our experimental model, PTHrP and related fragments were exogenously administered to cell cultures to resemble the marrow pathophysiology of the MM microenvironment, in which other cells usually release the peptides. This approach is in line with the work by other authors to obtain a detectable effect of PTHrP and its fragments in in vitro studies.[19, 41, 57]
In conclusion, we have identified novel pathogenetic functions of PTHrP in MM related to tumor survival, proliferation, and OC hyperactivation. Because serum elevation of PTHrP in patients with solid tumors produces severe skeleton devastation and leads to a poor prognosis, a systematic analysis of PTHrP levels in MM patients would be advisable, to determine whether this peptide is a novel marker of disease progression.
All authors state that they have no conflicts of interest.
This work was supported by the Italian Association for Cancer Research (Grant IG11647), the Italian Ministry of University and Research (PRIN 2009WZHMWJ), the University of Bari “Aldo Moro,” Bari, Italy (ex 60%, 2010). We thank Dr. Doug Burton for providing the PTHrP NLS+ mid-region (aa 38–94), Dr. Claudio Luparello for discussion of data, Dr. Elena Ranieri for the RCC-1 cell line, and Dr. Maria Pia Scavo for confocal microscopy analysis.
Authors' roles: Study design: PC and FS. Study conduct: AS, SS, AEB, and MV. Data collection and analysis: MDM and MT. Data interpretation: PC and AS. Drafting manuscript: PC. Revising manuscript content: FS. Approving final version of manuscript: PC, AS, SS, MDM, MT, AEB, MV, and FS. PC and FS take the responsibility for the integrity of the data analysis.