Lactoferrin Stimulates Osteoblast Differentiation Through PKA and p38 Pathways Independent of Lactoferrin's Receptor LRP1

Authors

  • Wei Zhang,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    2. Beijing Laboratory for Food Quality and Safety, Beijing, China
    Search for more papers by this author
    • WZ and HG contributed equally to this work.
  • Huiyuan Guo,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    2. Beijing Laboratory for Food Quality and Safety, Beijing, China
    Search for more papers by this author
    • WZ and HG contributed equally to this work.
  • Hao Jing,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Yixuan Li,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    2. Beijing Higher Institution Engineering Research Center of Animal Product, Beijing, China
    Search for more papers by this author
  • Xiaoyu Wang,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    2. Beijing Higher Institution Engineering Research Center of Animal Product, Beijing, China
    Search for more papers by this author
  • Hao Zhang,

    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    Search for more papers by this author
  • Lu Jiang,

    1. Beijing Laboratory for Food Quality and Safety, Beijing, China
    Search for more papers by this author
  • Fazheng Ren

    Corresponding author
    1. Key Laboratory of Functional Dairy, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, China
    2. Beijing Laboratory for Food Quality and Safety, Beijing, China
    • Address correspondence to: Fazheng Ren, PhD, College of Food Science and Nutritional Engineering, China Agricultural University, No. 17 Qinghua East Road, Haidian District, Beijing 100083, China. E-mail: renfazheng@263.net

    Search for more papers by this author

ABSTRACT

Lactoferrin (LF) has been established as a potent anabolic factor for bone health both in vivo and in vitro. However, the molecular mechanisms underlying LF's action are still largely unknown. Here, we explore the signaling pathways that mediate LF's beneficial effect on osteoblast differentiation. In primary osteoblast and preosteoblast MC3T3-E1, LF promoted alkaline phosphatase (ALP) activity, osteocalcin (OCN) secretion, and mineralization. Along with this enhanced osteogenic differentiation, activation of p38 mitogen-activated protein kinase (MAPK) was detected in LF-treated MC3T3-E1 cells. Downregulating p38 with selective inhibitor SB203580 or p38α small interfering RNA (siRNA) attenuated the effect of LF on osteogenesis. Furthermore, knockdown of p38α significantly decreased LF-induced Runt-related transcription factor 2 (Runx2) phosphorylation. According to previous studies and our results, we speculated that LF-induced osteoblast proliferation and differentiation were two relatively separate processes controlled by extracellular signal-regulated kinase 1/2 (ERK1/2) and p38 pathways, respectively. Besides p38 MAPK activation, protein kinase A (PKA) was also activated in MC3T3-E1 cells. PKA inhibitor H89 significantly inhibited LF-induced p38 activation, ALP activity, and OCN secretion, indicating that PKA possibly acted as an upstream kinase of p38. In order to further identify the role of LF's receptor low-density lipoprotein receptor-related protein 1 (LRP1), we constructed LRP1 stable-knockdown MC3T3-E1 cells. Neither LRP1 antagonist receptor associated protein (RAP), nor LRP1 knockdown approach could attenuate the LF-induced osteogenesis, implying that LF stimulated osteoblast differentiation via an LRP1-independent pathway. Taken together, the present work indicated that LF stimulated MC3T3-E1 preosteoblast differentiation mainly through LRP1-independent PKA and p38 signaling pathways. These results provided the first evidence of the signaling mechanisms of LF's effect on osteoblast differentiation. © 2014 American Society for Bone and Mineral Research.

Introduction

Osteoporosis is a skeletal disorder characterized by a systemic impairment of bone mass, strength, and microarchitecture, which increases the propensity of fragility fractures. The imbalance between the activities of osteoclast and osteoblast and the uncoupling of bone resorption and formation account for the bone loss.[1-3] Current drugs used to treat osteoporosis are major antiresorptive agents. These drugs target on osteoclast and inhibit the bone resorption process, whereas they have little effect on osteoblast-controlled bone formation.[4] Besides, the antiresorptive drugs have some limitations such as side effects, concurrent comorbidities, and inadequate long-term compliance.[3] Thus, it is desirable to develop natural and nontoxic agents that not only have the satisfactory anabolic effect but also are adequate for long-term therapy.

Lactoferrin (LF), an 80-kDa iron-binding glycoprotein, has been reported to be a potent bone growth factor both in vivo and in vitro.[5] Our previous work reported that oral administration of LF effectively prevented estrogen deficiency-induced bone loss and improved bone microarchitecture in ovariectomized rat model.[6] In vitro, LF was demonstrated to promote osteoblast proliferation and survival, and inhibit osteoclastogenesis.[7, 8] In particular, LF exerted beneficial effect on osteoblast differentiation and extracellular matrix calcification.[9-11] Osteoblast differentiation is a prerequisite of bone formation in vivo and is the primary function of osteoblast.[12] However, the molecular mechanisms underlying LF's effect on osteoblast differentiation remain poorly understood.

LF receptors are thought to have pivotal roles in mediating multiple functions of LF. To date, various LF receptors have been identified in different tissues and cell types. Intelectin, also called small intestine LF receptor, is the most documented LF receptor that specifically recognizes LF in human small intestine.[13] However, to date, there is no evidence available for the existence of intelectin in osteoblasts. Low-density lipoprotein receptor-related protein 1 (LRP1) is the only known LF receptor that mediates LF's action in osteoblast.[14] LF was shown to promote osteoblast proliferation through LRP1-mediated p42/44 mitogen-activated protein kinase (MAPK) activation,[15] whereas LF-induced osteoblast survival is LRP1-independent.[16] These results suggested that LRP1's roles may vary when mediating different physiological processes of osteoblasts in response to LF. Until now, the role of LRP1 in LF's effect on osteoblast differentiation has not been identified.

The MAPKs, a family of serine/threonine kinases, are well-known mediators connecting various extracellular stimuli and cellular responses.[17] A member of the MAPK family, extracellular signal-regulated kinase 1/2 (ERK1/2, also called p42/44 MAPK), was reported to be responsible for LF's mitogenic effect in osteoblasts,[15] suggesting that the MAPK family may be important for LF's action in osteoblasts. Previous work indicated that p38 MAPK was required for both natural and bone morphogenic protein 2 (BMP-2)-induced differentiation in osteoblasts.[18] Moreover, an increasing amount of research has shown that p38 MAPK is indispensible for osteoblast differentiation induced by various osteotropic factors, such as epinephrine[19] and parathyroid hormone (PTH).[20] ERK1/2 is also reported to be implicated in BMP-2[21] and 1,25(OH)2D3–induced[22] osteoblast differentiation. Until now, the roles of ERKs and p38 MAPK in the differentiation of osteoblast have been disputable. In our preliminary study, we found that LF activated all three major members of the MAPK family, but only p38 was associated with the LF-induced increase of alkaline phosphatase (ALP) activity (Supplementary Fig. S1 and S2). Therefore, it led us to investigate how p38 MAPK mediated the LF's stimulating signal in osteoblasts.

In this work, various approaches such as Western blotting, short hairpin RNA (shRNA) knockdown, and confocal microscopy were used to examine the signaling pathways that transduced the LF's stimulating signal during osteoblast differentiation. Furthermore, we also evaluated the roles of LF's receptor, LRP1, in LF-treated osteoblasts.

Materials and Methods

Reagents

Bovine LF with 95% purity (SDS-PAGE)[23] was provided by the Australian Yosica Holding (Melbourne, Australia). Fetal bovine serum, α modified essential medium (α-MEM), Opti-MEM reduced serum medium, and Lipofectamine 2000 reagent were obtained from Life Technologies (Rockville, MD, USA). DMSO, trypsin, puromycin, fluorescein isothiocyanate (FITC), Hoechst 33258, DiI, and Alizarin Red S were purchased from Sigma-Aldrich (St. Louis, MO, USA). Rabbit antibody against p38, phospho-p38 (p-p38, Thr180/Tyr182), cyclic adenosine monophosphate response element-binding protein (CREB), phospho-CREB (p-CREB, Ser133), anti-mouse and rabbit immunoglobulin G (IgG) peroxidase conjugate antibodies, and p38 activity assay kit were obtained from Cell Signaling Technology (Beverly, MA, USA). Rabbit anti–Runt-related transcription factor 2 (anti-Runx2) antibody, p38α small interfering RNA (siRNA) (M), control siRNA-A, and Protein A/G PLUS-Agarose were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Mouse anti-phosphotyrosine (p-Tyr, 4G10 Platinum) antibody, SB203580, and H89 were purchased from Calbiochem (Cambridge, MA, USA). Rabbit anti-LRP1 antibody was purchased from Epitomics (Burlingame, CA, USA). Rabbit anti-histone H4 and β-actin antibodies were obtained from Beyotime (Nantong, China). Recombinant receptor associated protein (RAP) was obtained from Sinobiological Inc. (Beijing, China).

Cell culture and treatment

The mouse preosteoblastic cell line MC3T3-E1 subclone 4 was obtained from American Type Culture Collection (ATCC, Rockville, MD, USA). Primary osteoblasts were prepared from the calvaria of newborn Sprague-Dawley rats as described.[23] Cells were maintained in α-MEM supplemented with 10% fetal bovine serum (defined as minimal medium) in a humidified 5% CO2 balanced air incubator at 37°C. Cells were subcultured using 0.25% trypsin every 2 to 3 days. For differentiation induction, cells were cultured in minimal medium supplied with 50 µg/mL ascorbic acid and 10 mM β-glycerophosphate (defined as differentiation medium), and the medium was changed every 2 days.

To examine the downstream signaling pathways in response to LF treatment, cells were seeded into 60-mm dishes at density of 2 × 104 cells/cm2 and cultured for 24 hours. Then the cells were starved in α-MEM without serum for 24 hours and cells were pretreated with inhibitors or activator for 1 hour before addition of LF (100 µg/mL).

ALP activity assay

Cells were seeded into 24-well plates at a density of 2 × 104 cells/cm2 and cultured for 7 days (2 days to reach confluence, and another 5 days for differentiation). At the last day, cells were stimulated with various concentrations of LF for 24 hours. For inhibitor assay, corresponding inhibitor was added 1 hour before LF treatment. Cells were then washed three times with physiological saline, and the cell layers were scraped and sonicated in 0.1 M Tris buffer (pH 7.4) containing 1% Triton X-100. ALP activity was quantitated in cell lysate using an ALP activity diagnostic kit (Roche Diagnostics GmbH, Basle, Swiss) with automated clinical chemistry analyzer (Hitachi, Tokyo, Japan). Aliquots of cell lysate were subjected to protein assay using a bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL, USA) and the ALP activities were normalized to the total protein.

Osteocalcin secretion assay

MC3T3-E1 cells were seeded into 24-well plates at a density of 2 × 104 cells/cm2 and cultured for 14 days (2 days to reach confluent, and another 12 days for differentiation). At the last day, cells were stimulated with various concentrations of LF for 24 hours. For inhibitor assay, corresponding inhibitor was added 1 hour before LF treatment. Then the amount of osteocalcin (OCN) in the culture supernatant was measured using a mouse osteocalcin enzyme immunoassay (EIA) kit (Biomedical Technologies Inc., Stoughton, MA, USA) according to the manufacturer's instruction.

Mineralization assay

Confluent cells were incubated in the differentiation media (containing indicated concentrations of LF) in six-well plates until day 35 (MC3T3-E1) or day 21 (primary osteoblast) from seeding. The differentiation media with vehicle or LF were changed every 2 days. For detecting bone nodules, the cultures in the wells were rinsed using an ice-cooled PBS and fixed with 70% ethyl alcohol. They were stained for 10 minutes with 40 mM Alizarin red S (pH 4.2). After washing with water, the samples were observed under light microscope and the representative pictures were photographed.

Western blotting and immunoprecipitation

Cell pellets (60-mm dishes) were lysed in 100 µL of lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 5 mM EDTA, 1 mM Na3VO4, 1 mM NaF, 10 µg/mL aprotinin, 10 µg/mL leupeptin, 10 µg/mL pepstatin, 10 mM iodoacetamide, and 1 mM phenylmethylsulfonyl fluoride [PMSF]) for 30 minutes on ice. Nuclear extracts were prepared with NE-PER nuclear and cytoplasmic extraction kit (Pierce). Protein concentration was determined using a BCA protein assay kit. For Western blotting, equal amounts of cleared cell lysates were separated by SDS-PAGE, followed by transfer of proteins to polyvinylidene difluoride membrane (Millipore, Temecula, CA, USA). The membrane was incubated with primary antibodies, followed by incubation with horseradish peroxidase-conjugated secondary antibody. The washed blot was developed using enhanced chemiluminescence reagent (Millipore) and exposed to X-ray films. For analysis of Runx2 phosphorylation, the cell extracts were incubated with anti-Runx2 antibody, after which the immune complexes were precipitated with protein A/G Plus agarose beads. Proteins were eluted from the beads and subjected to Western blotting. The phosphorylated Runx2 was detected by a specific mouse anti-phosphotyrosine antibody.[24]

p38 kinase activity assay

p38 activity was assessed by p38 kinase activity assay kit according to the manufacturer's instruction. Briefly, immobilized phospho-p38 MAPK (Thr180/Tyr182) monoclonal antibody was used to immunoprecipitate phospho-p38 MAPK, and then an in vitro kinase assay was performed using activating transcription factor 2 (ATF-2) as a substrate. The activity of p38 MAPK was determined by its ability to phosphorylate ATF-2, as detected by Western blotting using anti-phospho-ATF-2 (p-ATF-2, Thr71) antibody. The total amount of p38 in cell extracts was also measured as the loading control.

siRNA knockdown of p38 expression

MC3T3-E1 cells were seeded into six-well plates and grown 24 hours to reach 50% confluence before transfection. Then the medium was changed with opti-MEM reduced serum medium and MC3T3-E1 cells were transfected with p38α siRNA (M) and control scramble siRNA-A using Lipofectamine 2000. Forty-eight hours after siRNA transfection, cells were treated with LF for indicated periods and prepared for Western blotting analysis, ALP assay, and OCN mRNA analysis.

LRP1 stable knockdown with lentiviral LRP1 shRNA vectors

Lentiviral vectors (pGLV3/H1/GFP/Puro; GenePharma, Shanghai, China) carrying either four shRNA oligonucleotide sequences (LRP1-shRNA1-4, L1–L4) targeted to mouse LRP1 mRNA or a scramble shRNA (negative control, NC-shRNA, NC) were used. The shRNA sequences were provided in supplementary Table S1. MC3T3-E1 cells were seeded at 1 × 104 cells/cm2 in six-well plates the day before transfection. Lentivirus at 100 multiplicity of infection (MOI) and 8 µg/mL polybrene were added to α-MEM with 10% FBS in 1 mL total volume. After 24 hours, the medium was replaced for expansion of transduced cells. Forty-eight hours later the medium was removed and cells were selected by puromycin treatment (2 µg/mL) for 7 days. Knockdown of LRP1 was confirmed by Western blotting. The LRP1 shRNA that had the highest knockdown efficiency was used for all subsequent studies.

RT-PCR analysis

MC3T3-E1 cells were transfected with p38α siRNA as described in siRNA knockdown of p38 expression, and after transfection, the medium was replaced for expansion of transduced cells for 48 hours. Then the media were removed and cells were exposed to 100 µg/mL LF for 24 hours. The mRNA level of osteocalcin (OCN) was analyzed by semiquantitative RT-PCR reaction with respect to GAPDH level as controls. The primer sequences were as follows: for OCN, forward 5′-TGCTTGTGACGAGCTATCAG-3′ and reverse 5′-GAGGACAGGGAG GATCAAGT-3′; for GAPDH, forward 5′-TGGCAAAGTGGAGATTGTTGC-3′ and reverse 5′-AAGATGGTGATGGGCTTCCCG-3′. Total RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instruction. First-strand cDNA was synthesized from 2 µg of total RNA using first-strand cDNA synthesis kit (Takara Biotechnology, Dalian, China).

Confocal microscopy

LF was labeled with fluorescent dye FITC with a 1:5 (mol/mol) molar excess according to the manufacturer's instruction. FITC-labeled LF (FITC-LF) was then purified through a fluorescent dye removal column (Pierce). FITC-LF was added to the cultured cells at a final concentration of 100 µg/mL under the experimental conditions indicated, after which the cells were fixed in 4% paraformaldehyde. MC3T3-E1 cells membranes were stained with 1 µg/mL DiI, and nuclei were labeled with 4 µg/mL Hoechst 33258. Fluorescent images were captured using a Zeiss LSM 710 confocal laser scanning microscope (Carl Zeiss, Jena, Germany) with a Plan-Apochromat ×63/1.40 M27 glycerol immersion objective. The same gain, pinhole, and laser power settings were employed for imaging each treatment to achieve comparative results. The laser lines at 405 nm, 488 nm, and 514 nm were used for excitation of Hoechst 33258, FITC, and DiI, respectively. The emission wavelength range for each dye was selected according to the ZEN 2009 SP1 software (Carl Zeiss) and corresponding single optical sections were collected using an image format of 1024 × 1024 pixels.

Statistical analysis

All experiments were repeated at least three times. Data are presented as means ± SD. Statistical significance was evaluated by using Student's t test when only two groups were compared. If more than two groups were compared, evaluation of significance was performed using one-way analysis of variance (ANOVA) followed by Duncan's post hoc test. In all tests, statistical significance was set at p < 0.05.

Results

LF promotes differentiation and mineralization in MC3T3-E1 cells and primary osteoblasts

To study the mechanisms underlying LF's osteoinductive function, we first examined the effect of LF on the differentiation and mineralization in MC3T3-E1 cells and primary osteoblasts. As shown in Fig. 1A, C, LF treatment for 24 hours dramatically increased ALP activity in the range of 10 to 1000 µg/mL in both MC3T3-E1 cells and primary osteoblasts. The maximal stimulation was observed at 100 µg/mL (MC3T3-E1) and 1000 µg/mL (primary osteoblast). Likewise, LF also stimulated OCN secretion in two cells, although the maximal stimulation was observed at 50 µg/mL (MC3T3-E1) and 100 µg/mL (primary osteoblast) (Fig. 1B, D). Importantly, LF significantly promoted the mineralization of both cells (Fig. 1E). LF could stimulate MC3T3-E1 cells mineralization at range of 1 to 1000 µg/mL, whereas the lowest effective concentration of LF was higher (10 µg/mL) for primary osteoblasts. This is probably due to the different sensibility to LF and different capacity of differentiation of these two types of osteoblasts. During the LF-stimulated mineralization process, primary osteoblasts had higher differentiation capacity. So the primary osteoblast may not be as sensitive as MC3T3-E1 to LF stimuli, and a low concentration of LF (1 µg/mL) is insufficient to stimulate the mineralization of primary osteoblast. These results showed that LF is a potent factor for osteoblast differentiation and mineralization.

Figure 1.

LF increases ALP activity, OCN secretion and mineralization in MC3T3-E1 cells and primary osteoblasts. Differentiated MC3T3-E1 cells and primary osteoblasts were treated with the indicated concentration of LF for 24 hours as described in Materials and Methods. Then the MC3T3-E1 cells and primary osteoblasts were subjected to ALP activity assay (A, C) and OCN assay (B, D). Values of ALP and OCN are normalized to the total amount of protein. (E) MC3T3-E1 cells and primary osteoblasts reached confluence and were cultured in differentiation medium until day 35 and day 21 from seeding, respectively. During the mineralization, culture medium was changed every 2 days with or without LF. The extent of mineralized matrix in the plates was determined by Alizarin Red S staining. Each value is the mean ± SD of three independent experiments. Means without a common letter differ, p < 0.05. ALP = alkaline phosphatase; OCN = osteocalcin; LF = lactoferrin; OB = osteoblast.

LRP1 is not involved in mediating the LF's osteoinductive effect in MC3T3-E1 cells

LRP1 has been reported to mediate LF's mitogenic effect in osteoblasts,[15] so it we were very interested to explore the role of LRP1 in LF-induced osteoblast differentiation. To address this issue, we established LRP1 stable knockdown cells by introducing LRP1-shRNA lentiviral vectors into MC3T3-E1 cells using lentiviral transfection. After the transfection, positive cells were selected with puromycin and the knockdown of LRP1 was confirmed by Western blotting. The results showed that LRP1 shRNA L4 had the highest knockdown efficiency (Fig. 2A) and it was used for all subsequent studies. Next, we evaluated the osteoinductive effect of LF in LRP1 stable knockdown MC3T3-E1 cells. The results of ALP and OCN assays showed that knockdown of LRP1 did not influence LF's effect on ALP activity and OCN production in MC3T3-E1 cells (Fig. 2C, D). RAP is a natural antagonist for the LRP receptor family and blocks the binding of all known ligands to the receptor.[25] First, RAP activity was confirmed by detecting the inhibition effect of RAP on ERK1/2 activation as shown in Fig. 2B. Then, RAP was added before LF treatment to block the interaction between LF and LRP1. The results showed that LF-induced enhancements of ALP activity and OCN secretion were not inhibited by RAP (Fig. 2E, F). These results are consistent with LRP1 knockdown experiments, and clearly showed that LF stimulates osteoblast MC3T3-E1 cells via an LRP1-independent mechanism.

Figure 2.

LF promotes MC3T3-E1 cell differentiation in an LRP1-independent manner. (A) Following transfection of MC3T3-E1 cells with four shRNAs (L1–L4) targeted toward LRP1, the knockdown effect was validated through Western blotting using a rabbit monoclonal antibody against LRP1 light chain (85-kDa fragment). shRNA L4 had the best knockdown effect and was used in subsequent studies. (B) The activity of LRP1 antagonist RAP was validated by its inhibitory effect on ERK activation. MC3T3-E1 cells were pretreated with 50 nM RAP for 1 hour and treated with LF for another 24 hours, and then the cells' proteins were subjected to Western blotting. (C, D) Stable LRP1 knockdown MC3T3-E1 cells or negative control cells were differentiated and treated with 100 µg/mL LF or vehicle for 24 hours as described in ALP activity assay and Osteocalcin Secretion Assay. Then ALP activity assay and OCN assay were performed. (E, F) MC3T3-E1 cells were differentiated as described in ALP activity assay and Osteocalcin Secretion Assay and 50 nM RAP was added 1 hour before 100 µg/mL LF treatment. Then ALP activity assay and OCN assay were performed. # = significant difference between the LF-treated group and the LF-inhibitor co-treated group, as analyzed by Student's t test (p < 0.05); ns = no significant difference between the two groups (p > 0.05). LRP1 = low-density lipoprotein receptor-related protein 1; shRNA = short-hairpin RNA; MC = MC3T3-E1 cells without transfection; NC = negative control shRNA-transfected cells; ALP = alkaline phosphatase; OCN = osteocalcin; LF = lactoferrin; RAP = receptor associated protein; ERK = extracellular signal-regulated kinase; p-ERK = phospho-ERK.

LF activates p38 MAPK pathway in MC3T3-E1 cells

p38 MAPK has been reported to be an important pathway in regulating differentiation of osteoblasts.[18, 26] Here, our results showed that 100 µg/mL LF dramatically increased p38 phosphorylation within the initial 3 hours. After this rapid activation, the phosphorylation of p38 gradually declined to the basal level (Fig. 3A). Besides, LF was shown to stimulate p38 phosphorylation in a dose-dependent manner in MC3T3-E1 cells (Fig. 3B). Interestingly, once p38 was activated by LF, the phospho-p38 was translocated into the nuclei (Fig. 3C). Furthermore, LF dramatically increased p38 kinase activity at 10 to 100 µg/mL (Fig. 3D). These results together showed that LF could dramatically activate the p38 MAPK pathway in MC3T3-E1 cells.

Figure 3.

LF stimulates the activation of the p38 MAPK pathway in MC3T3-E1 cells. (A) Serum-starved MC3T3-E1 cells were incubated with 100 µg/mL LF or vehicle for the indicated time and then phospho-p38 was detected by Western blotting. (B) Concentration dependency of p38 phosphorylation induced by LF. Serum-starved cells were incubated with various concentrations of LF for 2 hours. Then the phospho-p38 was detected. (C) Serum-starved MC3T3-E1 cells were treated with 100 µg/mL LF for the indicated time and nuclear extracts were prepared for determination of phospho-p38 nuclear translocation. (D) In vitro kinase assay was performed for detection of p38 kinase activity in MC3T3-E1 cells treated with or without LF for 2 hours. Phospho-p38 was immunoprecipitated by an agarose conjugated phospho-p38 (Thr180/Tyr182) antibody and then in vitro kinase assay was performed using ATF-2 as a substrate. Phosphorylation of ATF-2 was detected by Western blotting. MAPK = mitogen-activated protein kinase; p-p38 = phospho-p38; ATF-2 = activating transcription factor 2; p-ATF-2 = phospho-ATF-2; LF = lactoferrin; IP = immunoprecipitation; IB = immunoblotting.

LF stimulates osteoblast differentiation through the p38 MAPK pathway

Because p38 activation was observed in LF-treated osteoblasts, the role of this pathway in mediating LF signaling was evaluated. As shown in Fig. 4A, blocking p38 MAPK by inhibitor SB203580 significantly attenuated the LF-induced p38 activation and simultaneously suppressed the elevation of ALP activity (Fig. 4B) and OCN secretion (Fig. 4C). In order to confirm the inhibitor results, we used an siRNA knockdown approach to examine the effect of p38α knockdown on LF-induced osteogenesis. As shown in Fig. 4D, MC3T3-E1 cells were transfected with p38α siRNA or control siRNA, and p38α siRNA transfection effectively decreased the basal level of p38 protein expression. Consistent with the inhibitor results, knockdown of p38α also significantly suppressed LF-induced enhancement of ALP activity (Fig. 4F) and OCN mRNA level (Fig. 4G). Because Runt-related transcription factor 2 (Runx2) is the major bone-specific transcription factor responsible for bone marker gene transcription including ALP and OCN during osteoblast differentiation,[27, 28] the activation of Runx2 was assessed. Lactoferrin could significantly increase the phosphorylation of Runx2 (Supplementary Fig. S3). Interestingly, LF-mediated phosphorylation of Runx2 was also significantly attenuated by p38α knockdown (Fig. 4E). These results demonstrated that p38 played an essential role in LF-induced differentiation of MC3T3-E1 cells.

Figure 4.

LF stimulates osteoblast differentiation through the p38 MAPK pathway. (A) MC3T3-E1 cells were incubated for 1 hour in the presence or absence of 20 µM SB203580, and then 100 µg/mL LF was added and incubated for 2 hours. The inhibitor activity of SB203580 was evaluated by detecting its inhibitory effect on p38 kinase activity. (B, C) MC3T3-E1 cells were incubated with 20 µM SB203580 for 1 hour prior to addition of 100 µg/mL LF. ALP activity in B and OCN secretion in C were measured as described in ALP activity assay and Osteocalcin Secretion Assay. (D) Knockdown effect of p38α siRNA on p38 expression. Cells were transfected with 50 nM control siRNA or p38α siRNA by Lipofectamine 2000 agent for 24 hours, followed by cultured for another 48 hours, and then the p38 expression was detected by Western blotting. (E) MC3T3-E1 cells were seeded into six-well plates and cultured for 24 hours to reach 50% confluence. Cells were transfected with 50 nM control siRNA or p38α siRNA by Lipofectamine 2000 agent for 24 hours, and the transfected cells were cultured for another 48 hours. Then the phosphorylation level of Runx2 was determined as described in Materials and Methods. (F, G) MC3T3-E1 cells were transfected with control siRNA and p38α siRNA as described in Materials and Methods. Vehicle or 100 µg/mL lactoferrin was added for 24 hours, and the suppression effect of p38α siRNA on LF-induced ALP activity (E) and OCN mRNA level (F) was measured. Each value is the mean ± SD of three independent experiments. # = significant difference between LF-treated group and LF-inhibitor (or siRNA) co-treated group, as analyzed by Student's t test (p < 0.05). LF = lactoferrin; MAPK = mitogen-activated protein kinase; SB = SB203580; ATF-2 = activating transcription factor 2; p-ATF-2 = phospho-ATF-2; ALP = alkaline phosphatase; OCN = osteocalcin; siRNA = small interfering RNA; Runx2 = Runt-related transcription factor 2; p-Tyr = phospho-tyrosine; IP = immunoprecipitation; IB = immunoblotting.

PKA pathway is required for LF-induced p38 activation and differentiation in MC3T3-E1 cells

It has been shown that PKA was implicated in the activation of p38 signaling,[20, 29] so we evaluated the role of this pathway in mediating LF stimuli. As shown in Fig. 5A, LF dramatically increased the phosphorylation of CREB, which is the direct substrate of activated PKA, whereas this effect could be suppressed by selective PKA inhibitor H89. Importantly, blocking PKA with H89 significantly attenuated the LF-induced p38 phosphorylation and p38 kinase activity (Fig. 5B, C), implying that LF-induced p38 activation is PKA-dependent in MC3T3-E1 cell. This result was further confirmed by the PKA activator forskolin, which could activate p38 MAPK in MC3T3-E1 cell (Supplementary Fig. S5). The kinetics of LF-induced p38 and CREB phosphorylation were also examined (Fig. 5D). Although p38 phosphorylation was promoted within 15 minutes, the maximal stimulation was observed at 75 minutes after LF addition. However, for PKA activation, phospho-CREB quickly reached to highest level at 15 minutes and sustained from 30 to 120 minutes in LF-treated cells. These results indicated that the PKA pathway was more quickly activated than p38 in response to LF treatment. Finally, H89 was applied to analyze the role of PKA in LF's action. H89 pretreatment dramatically attenuated the LF-induced ALP activity (Fig. 5E) and OCN production (Fig. 5F), which was similar to p38 inhibitor's effect. Taken together, these results showed that the PKA pathway was involved in LF's effect on osteoblast differentiation and probably acted as an upstream kinase of p38 MAPK.

Figure 5.

The PKA pathway is involved in LF-induced p38 activation and differentiation in MC3T3-E1 cells. (A) Inhibition effect of PKA inhibitor H89 on CREB activation. MC3T3-E1 cells were incubated for 1 hour in the presence or absence of 20 µM H89, and 100 µg/mL LF was added and incubated for 2 hours. The phosphorylation of CREB was detected by Western blotting. (B, C) Inhibition effect of H89 on LF-induced p38 phosphorylation and p38 kinase activity. MC3T3-E1 cells were incubated for 1 hour in the presence or absence of 20 µM H89, and 100 µg/mL LF was added and incubated for 2 hours. Cell lysates were prepared and subjected to Western bloting to analyze the expression of phospho-p38 (B), or to an assay kit for p38 kinase activity (C). (D) Serum-starved cells were treated with vehicle or 100 µg/mL LF for the indicated times and the phosphorylation level of p38 and CREB were analyzed by Western blotting. MC3T3-E1 cells were incubated with 20 µM H89 or vehicle for 1 hour prior to treatment with 100 µg/mL LF. ALP activity (E) and OCN production (F) were measured as described in Materials and Methods. # = significant difference between the LF-treated group and the LF-inhibitor (or siRNA) co-treated group, as analyzed by Student's t test (p < 0.05). PKA = protein kinase A; LF = lactoferrin; CREB = cyclic adenosine monophosphate response element-binding protein; ATF-2 = activating transcription factor 2; ALP = alkaline phosphatase; OCN = osteocalcin; IP = immunoprecipitation; IB = immunoblotting.

Endocytosis of LF is not required for activation of p38 MAPK and PKA

LF has been reported to be endocytosed into cells and interact with intracellular molecules.[30] Thus, we evaluated whether the endocytosis process contributed to LF-induced activation of p38 MAPK and PKA in osteoblasts. To monitor the endocytosis, LF was labeled with FITC, and the cell membrane and nuclei were labeled with DiI and Hoechst 33258, respectively. Endocytosis of LF was abrogated by placing cells at 4°C or in hypertonic medium (400 mM sucrose).[15] Confocal micrographs showed single optical sections of cells labeled with fluorescent dye (Fig. 6A). After 30 minutes of treatment, FITC-labeled LF (100 µg/mL, color green) was endocytosed by MC3T3-E1 cells (Fig. 6B). There was no labeled LF observed within cells placed at 4°C, a temperature at which receptor binding is intact but endocytosis is inhibited, even 30 minutes after addition (Fig. 6C). The internalization of LF was also inhibited in the presence of 400 mM sucrose (Fig. 6D), a condition under which endocytosis is also abrogated. These results confirmed that endocytosis of LF was significantly blocked by incubating the cells at 4°C or culturing in hypertonic medium. Under these conditions, LF still activated p38 MAPK (Fig. 6E, F) and CREB (Fig. 6G, H), which was similar to the results observed under normal conditions. These results suggested that endocytosis of LF was not required for the activation of p38 MAPK and PKA, and other potential LF receptors probably existed. Moreover, we also found that LRP1 is not the only endocytic receptor that mediates the internalization of lactoferrin, because the endocytosis of lactoferrin was barely affected in LRP1-knockdown MC3T3-E1 cell (Supplementary Fig. S4).

Figure 6.

Endocytosis of LF is not required for activation of p38 MAPK and PKA. (AD) Confocal micrographs show single optical sections of MC3T3-E1 cells labeled with nuclear stain Hoechst 333258 (blue) and membrane stain DiI (red) in the presence or absence of FITC-LF. (A) Control cells without FITC-LF. (B) MC3T3-E1 cells treated with 100 µg/mL FITC-LF for 30 minutes. (C, D) MC3T3-E1 cells were pretreated for 10 minutes at 4°C (C) or with 400 mM sucrose (D) before 100 µg/mL FITC-LF treatment for 30 minutes. The arrowheads demonstrate the endocytosed-LF (green) in MC3T3-E1 cell. The main panels are merged pictures that contain images of membrane (red), nuclei (blue) and FITC-LF (green), while the inserts are only FITC-LF images. MC3T3-E1 cells were pretreated for 10 minutes with 400 mM sucrose (hypertonic medium) before treatment with 0 to 100 µg/mL LF for 30 minutes. Whole-cell lysates were sequentially subjected to Western blotting analysis for phospho-p38 MAPK (E) and phospho-CREB (G). MC3T3-E1 cells were incubated at 4°C for 15 minutes and then treated with 100 µg/mL LF for the indicated times. Whole-cell lysates were sequentially subjected to Western blotting analysis for phospho-p38 MAPKs (F) and phospho-CREB (H). LF = lactoferrin; FITC = fluorescein isothiocyanate; FITC-LF = FITC-labeled LF; MAPK = mitogen-activated protein kinase; p-p38 = phospho-p38; PKA = protein kinase A; CREB = cyclic adenosine monophosphate response element-binding protein; p-CREB = phospho-CREB; MC = MC3T3-E1 cells; MC + LF = MC3T3-E1 cells treated with FITC-labeled LF.

Discussion

LF has been recognized as a new anabolic factor that potently promotes bone formation and inhibits bone resorption.[5] In this study, we showed that LF stimulated osteoblast differentiation mainly through LRP1-independent PKA and p38 MAPK pathways. This work, to the best of our knowledge, is the first evidence of the signaling mechanism underlying LF's beneficial effect on osteoblast differentiation.

Osteoblast differentiation is a tightly regulated process characterized by the coordinated expression of various osteoblast phenotypic markers, such as ALP and OCN.[31, 32] In our study, LF significantly promoted osteoblast ALP activity, OCN production, and mineralization in both MC3T3-E1 and primary osteoblasts (Fig. 1), indicating that LF is a potent stimulator for osteoblast differentiation and mineralization.

LF can exert multiple functions either by binding to the membrane receptors to trigger signal transduction, or by entering the target cells via endocytosis. Our results showed that endocytosis is not required for LF-induced p38 and PKA activation in MC3T3-E1 cells (Fig. 6EH). Therefore, LF seems to stimulate osteoblast differentiation via receptor-mediated mechanisms in osteoblasts. To date, many LF receptors have been identified in different tissues and cells.[13] Among these receptors, LRP1 is the only known receptor that transduces LF's signal in osteoblasts. Previous research indicated that LF activated p42/44 MAPK through LRP1 and consequently accelerated osteoblast proliferation.[15] It is thus intriguing to know whether LRP1 is the key mediator for transduction of LF's osteoinductive signaling during osteoblast differentiation. In this work, we established LRP1 stable knockdown MC3T3-E1 cells by lentiviral shRNA transfection (Fig. 2A). Inhibition of LRP1 with shRNA knockdown or antagonist RAP cannot abolish the LF's stimulating action on the ALP activity and OCN production in osteoblasts (Fig. 2CF). These results clearly showed the irrelevance of LRP1 in LF-induced osteoblast differentiation, and suggested the existence of other potential receptors exploited by LF to regulate osteoblast differentiation.

LRP5 and LRP6 are multiligand and non-endocytic members of the LRP receptor family that are structurally related to LRP1. LRP5 and LRP6 have been identified as critical regulators of osteoblast function and skeletal mass.[33, 34] But it is unknown if LRP5 and LRP6 possess signaling functions in LF-induced osteoblast differentiation. Besides, several new LF receptors have been identified recently, such as nucleolin[35] and GAPDH.[36] Nucleolin is a major ubiquitous 105-kDa phosphoprotein located mainly in the fibrillar components of nucleoli. Interestingly, nucleolin has been described as a shuttle between the cell surface and the nucleus and it was proposed as a mediator for the extracellular regulation of nuclear events.[37] To investigate the possible involvement of LRP5, LRP6, nucleolin, and GAPDH, the effect of LF on the expressions of these receptors was determined by Western blotting. The results showed that LRP5, LRP6, nucleolin, and GAPDH were all expressed in MC3T3-E1 cells, but, unfortunately, we did not observe any significant enhancement in response to LF treatment in MC3T3-E1 cells (Supplementary Fig. S6). Thus, the specific roles of these potential receptors in LF's action still need further investigation.

MAPKs are well-studied pathways that are implicated in the regulation of osteoblast differentiation, and different regulation patterns have been observed in response to various factors.[18, 38, 39] Grey and colleagues[15] reported that ERK1/2 mediated the LF's mitogenic signal in osteoblasts. However, blocking ERK1/2 with inhibitor U0126 did not affect LF-induced enhancement of ALP activity in osteoblasts (Supplementary Fig. S2). In the present work, LF significantly activated p38 kinase (Fig. 3), and more importantly, the p38 pathway is one of the mechanisms by which LF promoted the activation of Runx2 (Fig. 4E) and the differentiation of osteoblasts (Fig. 4B, C, F, G). Therefore, it is plausible to propose that LF-induced osteoblast proliferation and differentiation were two relatively separate processes controlled by ERK1/2 and p38 pathways, respectively. Similarly, Suzuki and colleagues[19] reported that, in MC3T3-E1 cells exposed to epinephrine, ERKs played an essential role in cell replication, whereas p38 was involved in the regulation of ALP expression. Previous studies and our results suggested that the separate operation of MAPK pathways may probably be a general signal transduction fashion for some extracellular factors in osteoblasts. Previous researches have demonstrated that Runx2 activation can be regulated by either ERK or p38, or both,[24, 39, 40] depending on the type of stimuli and the cell models applied in the particular study. Our results confirmed that LF activated Runx2 through the p38 pathway and consequently regulated the differentiation process.

PKA is a classical signal transducer and is involving in some growth factors' osteotropic effects.[20] In our study, the PKA pathway was quickly activated in the presence of LF (Fig. 5D) and was involved in the regulation of osteoblast differentiation (Fig. 5E, F). Moreover, we also found that PKA was an upstream kinase of p38 MAPK (Fig. 5B, C, D) and the crosstalk between PKA and p38 was crucial for LF's osteoinductive effect. Except for mediating signal transduction in osteoblasts, the PKA pathway was also associated with the regulation of bone-related hormones, such as prostaglandin E2 (PGE2). A recent study showed that fluid flow stimulation activated the PKA pathway and consequently induced the expression of Cox2, which is the rate-limiting enzyme for PGE2 synthesis.[41] Interestingly, increased expression of Cox2 and elevated PGE2 levels were observed in LF-treated MC3T3-E1 cells.[14] This evidence provided the possibility that LF could promote osteoblast activity by hormones' autocrine actions through the PKA pathway.

In conclusion, in this study we report that LF stimulates osteoblast differentiation mainly through LRP1-independent PKA and p38 signaling pathways. LF-induced osteoblast proliferation and differentiation appears to be mutually independent intracellular events that are regulated by ERK1/2 and p38 pathways, respectively. Endocytosis of LF is not required for LF's osteoinductive effects, and the role of this process needs further study (Fig. 7). This work provides some new insight into the molecular mechanisms underlying the anabolic effect of LF on bone and will promote the further use of LF as a natural antiosteoporosis reagent.

Figure 7.

Schematic diagram of summarized signaling mechanisms underlying LF's anabolic effect on osteoblasts. LF = lactoferrin; LRP1 = low-density lipoprotein receptor-related protein 1; PKA = protein kinase A; ERK1/2 = extracellular signal-regulated kinase 1/2; Runx2 = Runt-related transcription factor 2.

Disclosures

All authors state that they have no conflicts of interest.

Acknowledgments

This work was supported by grants from the National Natural Science Foundation of China (31000771) and National Science and Technology Support Program (2012BAD12B08).

Authors' roles: WZ and HG conducted most of the experimental work and contributed to the drafting of the manuscript; YL and XW contributed to the experimental work; HJ, LJ, and HZ conducted the data and results analysis; FR contributed to the design of the research and had primary responsibility for final content. All authors read and approved the final manuscript.

Ancillary