- Email: [email protected]
EXPRESSION OF CATION EXCHANGER NHE AND ANION EXCHANGER AE ISOFORMS IN PRIMARY HUMAN BONE-DERIVED OSTEOBLASTS
Cell Biology International, 1998, Vol. 22, No. 7/8, 551–562 Article No. cb980299, available online at http://www.idealibrary.com on
EXPRESSION OF CATION EXCHANGER NHE AND ANION EXCHANGER AE ISOFORMS IN PRIMARY HUMAN BONE-DERIVED OSTEOBLASTS A. MOBASHERI1,2,3, S. GOLDING2, S. N. PAGAKIS2, K. CORKEY2, A. E. POCOCK3, B. FERMOR3, M. J. O’BRIEN2, R. J. WILKINS2, J. C. ELLORY2 and M. J. O. FRANCIS3* 1
Department of Biomedical Sciences, School of Biosciences, University of Westminster, 115 New Cavendish Street, London W1M 8JS; 2University Laboratory of Physiology, University of Oxford, Parks Road, Oxford OX1 3PT; and 3Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Headington, Oxford OX3 7LD, U.K. Received 11 July 1997, accepted 26 August 1998
The authors used isoform-specific antibodies against cation (NHE) and anion (AE) exchange isoforms in order to establish their specific expression and localization in dispersed human bone-derived cells. Immunocytochemical preparations of permeabilized osteoblasts probed with polyclonal antibodies were optically analysed by conventional immunofluorescence and confocal laser scanning microscopy. These techniques demonstrated the abundant presence of epitopes of the cation exchangers NHE1 and NHE3 and the anion exchanger AE2 in these cells. The NHE1 and NHE3 isoform proteins were predominantly located in subplasmalemmal and nucleoplasmic vesicles. The AE2 isoform was densely localized to a subcellular location characteristic of the Golgi complex. The molecular identity of the AE and NHE isoforms was investigated by RT-PCR that confirmed the presence of NHE1 and NHE3 transcripts in addition to NHE4. RT-PCR and diagnostic restriction analysis of amplified AE cDNA established preferential AE2 expression. Since AE2 has been shown to act as a sulfate transporter at low pH, it is possible that it performs this function in the osteoblast Golgi complex where sulfation reactions occur post-translationally on numerous extracellular matrix macromolecules prior to secretion and mineralization. The Na + /H + exchanger proteins are regulated by mitogenic and non-mitogenic stimuli in the osseus environment and are involved in the large fluxes of ions and protons that necessarily occur during bone formation and resorption and thus play an important role in intracellular ion homeostasis in osteoblasts. 1998 Academic Press
K: osteoblast; Na + /H + exchanger; NHE; anion exchanger; AE; isoform; immunofluorescence; confocal microscopy; RT-PCR; alkaline phosphatase; bone mineralization, pH regulation
INTRODUCTION Bone, the skeletal support system of the vertebrate body, is a dynamic tissue that is continually formed, resorbed and remodelled. Osteoblasts are highly specialized mononuclear cuboidal cells responsible for the formation of organic bone matrix and its subsequent mineralization (Triffitt, *To whom correspondence should be addressed: Dr Martin J. O. Francis, University Lecturer in Biochemistry, Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Headington, Oxford OX3 7LD; E-mail: [email protected] 1065–6995/98/070551+12 $30.00/0
1980; Parfitt, 1984). They originate from multipotential mesenchymal cells, which also give rise to closely related fibroblasts and chondroblasts and somewhat less related myocytes and adipocytes (Grigoriadis et al., 1988, 1990). Osteoblasts actively synthesize and secrete an extracellular matrix consisting of type I collagen, proteoglycan ground substance and non-collagenous proteins. The mineralization process is initiated by osteoblasts through the polarized secretion of matrix vesicles loaded with alkaline phosphatase, an enzyme responsible for calcification of the matrix. 1998 Academic Press
552
We are interested in the molecular identity and isoforms of active and passive transport proteins involved in pH regulation and ion homeostasis in osteoblasts. In this study, we have concentrated our efforts on candidate plasma membrane transport systems involved in pH regulation in primary human bone-derived osteoblasts. Most mammalian cells possess multiple mechanisms of pH homeostasis responsible for maintaining a relatively stable intracellular pH (pHi)7.3. Under physiological conditions these may be divided into the acid and base extruder groups. For instance, in the archetypal cell, sodium independent Cl /HCO3 exchangers also known as band 3-related anion exchangers participate in base extrusion (Alper, 1994) whereas sodium dependent Na + /H + or cation exchangers are involved in acid efflux (Tse et al., 1994). Active transport mechanisms such as the H + ,K + -ATPase pump (Chow and Forte, 1995) complement the combined activity of the cation and anion exchangers in some cell types. In addition to intracellular pH and ion homeostasis, cation and anion exchangers play a significant role in cell volume regulation (Grinstein et al., 1989; Alper, 1994) and in epithelial cells cation exchangers are responsible for acid secretion or Na + reabsorption (Wakabayashi et al., 1997). Four cation exchanger proteins (NHE1–4) and three anion exchanger proteins (AE1–3) have been identified and characterized to date (Alper, 1994; Tse et al., 1994). Immunocytochemical studies suggest that NHE1 is the ubiquitous housekeeping isoform expressed by most epithelial and nonepithelial cells and functional studies have demonstrated its involvement in intracellular pH regulation. However, NHE2, NHE3 and NHE4 are more restricted in their distribution and are predominantly expressed in the epithelial cells of the stomach, kidney and the intestines (Wakabayashi et al., 1997). AE1 commonly referred to as band-3, is the prototype member of the anion exchange family found in the erythropoietic system and the kidney (Kopito and Lodish, 1985; Kudrycki and Shull, 1989). Transcripts for AE2 have been detected in almost all tissues examined whereas mRNA for the third member of the anion exchanger family AE3 appears to be expressed most abundantly in heart and brain (Kudrycki et al., 1990). It has recently become apparent that a number of hormones and growth factors modulate cytosolic pH (pHi) in osteoblasts, and there is some evidence that this in turn influences DNA synthesis and cellular proliferation (Reid et al., 1988). Parathyroid hormone (PTH) and parathyroid hormone-
Cell Biology International, Vol. 22, No. 7/8, 1998
related peptide (PTHRP) are important regulators of normal and pathological bone remodelling. A consequence of PTH/PTHRP action in osteoblasts is intracellular acidification followed by activation of the Na + /H + exchanger system that has been shown to play a central role in modulating pHi recovery (Redhead and Baker, 1988). PTH has been demonstrated to decrease pHi and growth in UMR-106 osteoblast-like cells (Reid et al., 1988). Depression of pHi has important consequences for bone metabolism. In osteoblasts the regulation of pHi is a critical component of hormone stimulated bone remodelling where local fluctuations occur in the osseus extracellular microenvironment (Redhead and Baker, 1988). At the cellular level, Na + /H + exchanger activity is regulated by a variety of stimuli that activate diverse signal transduction systems including protein kinase A, protein kinase C, tyrosine kinase and Ca2+ /calmodulin dependent protein kinase II (Wakabayashi et al., 1997). This extensive functional and regulatory diversity appears to be generated by tissue-specific expression of Na + /H + exchanger isoforms in a variety of tissues. A recent report by Azarani and co-workers (1995) suggests that the established rat osteoblastic cell line UMR-106 expresses only one isoform of the Na + /H + exchanger, namely NHE1, that is activated by PTH/PTHRP via a cyclic AMP dependent pathway. Other investigators working on similar and unrelated cell lines have suggested that more diverse signalling pathways are in operation in bone derived cells (Reid et al., 1988; Sugimoto et al., 1992; Green and Kleeman, 1992; Gupta et al., 1994). The existence of various signal transduction pathways may be reflected by the bone cell-specific expression of multiple NHE isoform proteins differentially responsive to distinct signalling pathways. The conflicting hypotheses regarding the expression and regulation of Na + /H + exchange in bone cells prompted us to investigate the expression of Na + /H + exchanger (and anion exchanger) systems in human bone-derived osteoblasts at the protein and mRNA levels to gain an insight into the process of human osteoblast pH regulation. To this end, we used isoform-specific antibodies to determine which cation and anion exchange systems may be important to these cells in the context of human bone mineralization and investigated their expression and cellular localization by immunofluorescence confocal laser scanning microscopy and the reverse transcriptase polymerase chain reaction (RT-PCR). There may be important species specific differences in the expression of these systems in bone and thus the work described here aims to unravel the
Cell Biology International, Vol. 22, No. 7/8, 1998
physiological complexities of pH regulation in human bone-derived osteoblastic cells at the molecular level. MATERIALS AND METHODS Cell preparation and culture Human bone derived osteoblasts were grown out from trabecular bone chips and cultured essentially as previously described (Gundle et al., 1995a,b). The approval of the Hospital Medical Staff Ethics Committee was obtained for all the experiments described. Briefly, samples of trabecular bone were obtained from the greater trochanter of patients undergoing surgery for hip replacement. Primary cultures of human bone derived osteoblasts were obtained by outgrowth from bone chip explants in DMEM supplemented with 10 m NaHCO3, 25 m HEPES, 2 m -glutamine, 10% heat treated foetal calf serum, 30 units/ml benzylpenicillin and 15 µg/ml streptomycin sulfate, 100 n -ascorbic acid-2-phosphate (Asc-2-P) and 10 n dexamethasone pH 7.3 in 80-cm2 culture flasks in a humidified incubator at 37C and 5% CO2, 95% air until they reached confluence. They were then detached using sequential collagenase and trypsin digestion and plated into 24 well plates containing microscope glass coverslips. For comparing alkaline phosphatase activities, primary cultures of human anterior cruciate ligament (ACL) cells were obtained by outgrowth from explants of normal ACL obtained at knee replacement for osteoarthritis in Dulbecco’s minimum essential medium (DMEM) supplemented with 10 m NaHCO3, 25 m HEPES, 2 m -glutamine, 10% heat treated foetal calf serum, 30 units/ml benzylpenicillin and 15 µg/ml streptomycin sulfate 100 n -ascorbic acid-2-phosphate (ASC-2-P), pH 7.3 in 80-cm2 culture flasks in a humidified incubator at 37C and 5% CO2/95% air until they reached confluence. They were then detached using sequential collagenase and trypsin digestion and plated into 24-well plates. Isoform-specific antibodies A monoclonal antibody raised against the human erythrocyte anion exchange AE1 isoform was purchased from Sigma (B9277). Dr Seth Alper, Harvard Medical School, Boston donated affinitypurified rabbit polyclonal antibodies against the C-terminal peptide of AE1 and AE2. This antibody was originally raised against the C-terminal peptide
553
of AE2 but cross-reacts with AE1 (Stuart-Tilley et al., 1994). Polyclonal antiserum against NHE1 was obtained from Dr Jacques Pouyssegur, Centre National de la Recherche Scientifique, Nice. Polyclonal antisera against NHE2 and NHE3 were obtained from Dr Ming Tse and Dr Mark Donowitz, Johns Hopkins University Medical School, Baltimore. A monoclonal antibody against the human erythrocyte Mg2+ -dependent Ca2+ ATPase (Clone 5F10) was purchased from Sigma (A7952) and the abundant expression of this protein in human bone derived cells was used as a marker for the osteoblastic phenotype (Borke et al., 1988) in addition to quantitative alkaline phosphatase expression. Confocal image acquisition and immunofluorescence microscopy Osteoblasts grown on circular glass coverslips were fixed in freshly prepared 3.7% paraformaldehyde in PBS for 10 min before permeabilizing in 1% SDS in PBS (Brown et al., 1996) for a further 5 min. Non-specific binding sites in the permeabilized cells were blocked overnight with 10–20% normal goat serum (NGS; Sigma) in PBS at 4C. The cells were then probed with primary isoform-specific antibodies and secondary anti-mouse and anti-rabbit IgG (Sigma) conjugated to crystalline tetramethylrhodamine isothiocyanate (TRITC) appropriately diluted in PBS containing 1% NGS before mounting in non-fluorescent mounting media (Hydromount , National Diagnostics). The cells were examined using a Leica confocal laser scanning microscope fitted with an oil immersion objective (100). For excitation of TRITC, an argon–krypton laser was used. With the pinhole filter combination, laser power setting and TRITC filter, optical images were recorded with intervals of approximately 0.5 ìm in the z direction. The images were median filtered (33) to remove noise and saved in either a 512512 pixel TIFF (tagged image file format) format or in a 256256 pixel TIFF format using the public domain NIH Image Analysis software for Macintosh. Filtered images were printed on a Codonics dye sublimation photographic quality printer with a minimum resolution of 600 dpi. Total RNA isolation, cDNA synthesis and RT-PCR RT-PCR was necessary to distinguish between AE1 and AE2 expression as the polyclonal antibody available cross-reacts with both protein isoforms.
554
Cell Biology International, Vol. 22, No. 7/8, 1998
Table 1. DNA sequence of oligonucleotide primers for RT-PCR. Sequences are from 5 to 3 direction and the estimated size of the RT-PCR product is indicated. In the AE row, isoform-independent oligonucleotide primers complementary to mRNA sequences within membrane spanning domains 1 and 6 of AE are shown. The predicted size of all expected AE isoforms is given in parentheses. The neutral base inosine (I) was used in the AE forward primer due to degeneracy in codon usage between AE isoforms Isoform NHE1 NHE2 NHE3 NHE4 AE â-actin
Sequence
PCR Product predicted size (bp)
TCTGCCGTCTCAACTGTCTCTA (forward) CCCTTCAACTCCTCATTCACCA (reverse) GCAGATGGTAATAGCAGCGA (forward) CCTTGGTGGGGGCTTGGGTG (reverse) ACGTCCAGGACCCCTACATC (forward) TGGACCTCCTCAAACACGGCCAG (reverse) GGCTGGGATTGAAGATGTATGT (forward) GCTGGCTGAGGATTGCTGTAA (reverse) TTC ATC TAC TTT GCI GCC CTG (forward) CGA CAG GAG GGC CGT GTT GGG (reverse) TTC AAC TCC ATC ATG AAG AAG TGT GAC GTG (forward) CTA AGT CAT AGT CCG CCT AGA AGC ATT (reverse)
Total RNA was prepared from the osteoblasts using a Qiagen RNeasy kit (Qiagen). Before PCR amplification an oligo-dT primed cDNA library was prepared from total RNA using SuperScript II reverse transcriptase (Life Technologies). PCR of AE was performed as previously described (Golding et al., 1997) using isoform-independent oligonucleotide primers complementary to mRNA sequences within membrane spanning domains 1 and 6 of AE. The primer sequences used were as follows: AEM1-F(orward) primer 5 TTC ATC TAC TTT GCI GCC CTG-3 and AEM6-R(everse) primer 5 -CGA CAG GAG GGC CGT GTT GGG-3 . Inosine (I), a neutral base, was used in the primers due to degeneracy in codon usage between AE isoforms. The following â-actin primer sequences were used as a positive PCR control: 5 -TTC AAC TCC ATC ATG AAG AAG TGT GAC GTG-3 (Forward) and 5 -CTA AGT CAT AGT CCG CCT AGA AGC ATT-3 (Reverse). The PCR cycling protocol consisted of 1 cycle of 95C3 min, then 40 cycles of 95C1 min, 55C1 min, 72C2 min, followed by 1 final cycle of 72C5 min. Isoformspecific primers and cycle conditions for NHE PCR were as described in Borensztein et al. (1995) (Table 1), but with the NHE3 primer sequences replaced by 5 -ACG TCC AGG ACC CCT ACA TC-3 (forward) and 5 -TGG ACC TCC TCA AAC ACG GCC AG-3 (reverse) as the primers described were found to be rat-specific. PCR was also performed using cDNA prepared in the absence of reverse transcriptase as a negative control. Following amplification, AE PCR products
422 310 460 501 339 (AE1), 573 (AE2), 457 (AE3) 310
were purified and concentrated using the Wizard PCR Purification System (Promega, U.K.), and resuspended in 18 ìl of nuclease-free water to which was added 2 ìl 10 restriction enzyme buffer (supplied) and 5 U AvaII restriction enzyme (Boehringer Mannheim, U.K.). Samples were then incubated at 37C for 3 h before separation on a 4% NuSieve GTG agarose (FMC, Flowgen, U.K.) gel in TAE buffer (40 m Tris/acetate, 1 m EDTA) for 4 h at 40 V. DNA assay. Solubilized DNA was measured fluorometrically. DNA isolated from cells was solubilized with 10 m EDTA, adjusted to pH 12.3 with NaOH and incubated at 37C for 30 min. The pH was adjusted to 7.0 with 1 KH2PO4. Hoechst 33258 (300 ng) was added and fluorescence measured using a Perkin Elmer LS 30 fluorimeter with excitation and emission wavelengths set at 355 nm and 450 nm and compared to that of known standard DNA samples. Alkaline phosphatase activity. Alkaline phosphatase activity was determined by measuring the release of p-nitrophenol from p-nitrophenyl phosphate disodium as described in Sigma technical bulletin no. 104. For quantitative studies, the amount of p-nitrophenol released was determined using a Gilford System 2600 spectrophotometer at 410 nm. Sites of alkaline phosphatase activity were stained red for qualitative studies and microscopic evaluation.
Cell Biology International, Vol. 22, No. 7/8, 1998
RESULTS Immunofluorescence and confocal analysis of NHE and AE isoform expression in human osteoblasts Immunofluorescence and confocal laser scanning microscopy was utilized to investigate the specific expression and subcellular localization of epitopes of cation and anion exchange proteins abundantly expressed in osteoblasts. Isoform specific polyclonal antibodies and secondary TRITCconjugated anti-rabbit IgG were used to stain cation and anion exchange proteins in permeabilized osteoblasts in which antigenic sites were retrieved by SDS pre-treatment (Brown et al., 1996). We used a confocal microscope in addition to a conventional fluorescence microscope since it allows 3-dimensional visualization of the protein of interest and reveals more information about its morphological and subcellular localization. The immunofluorescence data revealed abundant expression of the Mg2+ -dependent Ca2+ -ATPase (PMCA1 isoform) in the plasma membrane and subplasmalemmal vesicles of paraformaldehyde fixed, SDS permeabilized osteoblasts (Fig. 2A) The expression of the alkaline phosphatase (Fig. 1) was used as a bone specific marker for the osteoblastic phenotype. Comparison of alkaline phosphatase expression levels in human bone-derived osteolasts and anterior cruciate ligament cells provided convincing evidence for a true osteoblastic phenotype in the bone-derived osteoblasts (see Table 2). The cation exchangers NHE1 and NHE3 were both present as indicated by the intense fluorescent staining of both proteins in osteoblasts (Fig. 2B, E, respectively). The fluorescent immunostaining for the NHE1 isoform was predominantly nucleoplasmic with nucleolar exclusion; the most abundant staining was present in a particulate intracellular pattern as observed by optical sectioning (Fig. 2B). This finding is concordant with the concept of ubiquitous NHE1 isoform expression but perhaps not consistent in a straightforward manner with a housekeeping acid efflux role in dispersed osteoblasts. In contrast to NHE1 the fluorescent immunostaining of the NHE3 isoform was plasmalemmal and nucleoplasmic. The monoclonal anti-human erythrocyte AE1 did not recognize any epitopes in osteoblasts (results not shown). Although no evidence for AE1 expression was found the anion exchanger AE2 isoform was abundantly expressed in osteoblasts (Fig. 2C). In contrast to the NHE proteins, most of the staining for AE2 was predominantly intracellular and polar suggesting that the bulk of the AE2 protein is in an intracellular compartment characteristic of the
555
Golgi apparatus. It is important to note that the intracellular AE2 protein was not detectable without prior permeabilization and antigen retrieval with SDS and non-permeabilized cells exhibited almost no immunoreactivity when probed with the AE2 specific antibodies. This finding is consistent with the established notion that newly biosynthesized AE2 protein is translated in the endoplasmic reticulum and further processed in the cis, medial and trans-Golgi compartments before trafficking to the plasma membrane (Alper, 1994). However, the high intracellular abundance of the AE2 isoform in osteoblasts suggests that most of the AE2 protein may not be targeted to the plasma membrane but intended for prolonged transit in the organelles. In all experiments performed, cells probed with secondary antibody alone exhibited negligible background staining even when the laser power was set to maximum (results not shown). RT-PCR analysis of AE and NHE expression in human osteoblasts We have previously used RT-PCR to demonstrate the presence of transcripts encoding AE2 in isolated articular chondrocytes (Golding et al., 1997). Here we used an identical approach to detect transcripts encoding AE2 in human osteoblasts. This was carried out to confirm the results obtained by immunofluorescence and confocal microscopy. Although the anti-erythrocyte AE1 monoclonal antibody did not recognize epitopes in osteoblasts, the polyclonal AE antibody can recognize shared epitopes on the AE1 and AE2 isoforms. Therefore, experiments were designed and performed to demonstrate the molecular identity of the AE system present in these cells. PCR amplification of cDNA between membrane spanning regions 1 and 6 of AE produced one cDNA product of a size (570 bp) consistent with the amplification of AE2 cDNA rather than AE1 which, based on other cloned sequences, would have been amplified to give a cDNA product 100 bp smaller in size (Fig. 3) (Golding et al., 1997). As further proof, the osteoblast PCR product was digested with the restriction enyme AvaII. Using the Webcutter v2.0 program (http://www.firstmarket.com/cutter/) and the published mRNA sequences for rat kidney AE1 (Kudrycki and Shull, 1989), human kidney AE2 (Medina et al., 1997) and mouse brain AE3 (Kopito et al., 1989) between MEM1 and MEM6 it was deduced that digestion by AvaII would produce cDNA fragments of 16 bp, 140 bp and 183 bp (AE1); 66 bp, 240 bp and 267 bp (AE2); and 68 bp, 277 bp and 112 bp (AE3). As can be seen in Fig. 4,
556
Cell Biology International, Vol. 22, No. 7/8, 1998
Fig. 1. A: High-power phase-contrast light micrograph of cultured human bone-derived osteoblasts. B: Expression of alkaline phosphatase by the cells demonstrated by histochemical staining. Original magnification=400, reproduced at 95%.
Fig. 2. Confocal laser scanning microscope (CLSM) optical sections (0.5 ìm, in the z direction) through osteoblasts probed with polyclonal antibodies against the Mg2+ -dependent Ca2+ -ATPase, the cation exchanger NHE1 and anion exchanger AE2 isoforms. A: The polar expression of the Mg2+ dependent Ca2+ -ATPase (PMCA1 isoform) in human primary bone derived osteoblasts is one marker for the osteoblastic phenotype (N=nucleus; bar=10 ìm). B: Cation exchange (NHE1 isoform) distribution in human primary bone derived osteoblasts. The staining is predominantly nucleoplasmic with nucleolar exclusion. C: Distribution of the band-3 like anion exchanger (AE2 isoform) in human primary bone derived osteoblasts. The pattern of immunofluorescent staining is typically organellar and the bulk of the AE2 protein is most likely in the Golgi complex (arrows indicate intracellular compartment). D: Control cells where primary antibody was omitted. E: Conventional immunofluorescence micrograph of osteoblasts probed with a polyclonal NHE3 specific antibody and secondary TRITC-conjugated IgG.
558
Cell Biology International, Vol. 22, No. 7/8, 1998
Table 2. Expression of alkaline phosphatase by human bone derived osteoblasts (HBDO) compared with nonosteoblastic anterior cruciate ligament (ACL) cells compared over a period of 12 days. These quantitative measurements show roughly a seven-fold difference in alkaline phosphatase expression in HBDO cells compared to ACL cells on day 6 and a two-fold increase on day 12. S.D., Standard deviation of the mean; n=9. Units of alkaline phosphatase expression shown are per ìg of culture DNA Days in culture
ACL
..
HBDO
..
3 6 9 12
15.64 18.50 23.66 56.03
4.46 0.62 0.17 6.70
77.0 137.80 114.45 123.80
6.50 11.54 8.74 11.51
AvaII digestion of the osteoblast PCR product produced fragments of a size consistent with the original product being AE2. Using RT-PCR, we have also been able to amplify NHE isoforms 1, 3 and 4 from osteoblasts (Fig. 5) using protocols described elsewhere (Borensztein et al., 1995). However, we were unable to detect transcripts for NHE2 by RT-PCR (not shown). Therefore, taken together these results indicated that human primary derived osteoblasts transcribe the AE2 isoform of AE and NHE isoforms 1, 3 and 4. Overall, the results presented here suggest that human bone derived cells possess at least three distinct Na + /H + exchangers (NHE1, NHE3 and NHE4) and at least one anion exchanger (AE2). DISCUSSION In this study we have demonstrated that human bone derived osteoblasts express the NHE1 and
Fig. 3. RT-PCR amplified a 570 bp cDNA fragment from cDNA library prepared from the human bone derived osteoblast that corresponds to regions between membrane spanning regions 1 and 6 of the AE2 isoform.
NHE3 isoforms of the Na + /H + exchanger system at the protein level and the NHE4 isoform at the mRNA level. Although we have presented no functional evidence for an acid efflux ‘housekeeping’ role for NHE1, our data indicates that this Na + / H + exchanger is present in human osteoblasts. The expression of NHE1 in osteoblasts was expected as Northern blot analysis in UMR-106 osteoblastic cells (Azarani et al., 1995) and immunocytochemistry in almost all other cells and tissues examined have found this protein (Tse et al., 1991). The extensive literature on NHE1 expression clearly indicates that this isoform is ubiquitously expressed and plays a ‘housekeeping’ role in almost all cells (for reviews see Orlowski and Grinstein, 1997 and Wakabayashi et al., 1997). Physiologically, the presence of cation exchange mechanisms has been previously demonstrated by numerous groups in osteoblast-like cells (Redhead and Baker, 1988; Reid et al., 1988; Sujimoto et al., 1988; Dascalu et al., 1992; Graham and Tashjian Jr, 1992). The NHE1 exchanger is activated by changes in cell volume (particularly hyperosmolarity or cell shrinkage) and is entirely inhibited by amiloride and its analogues in osteoblasts and its immunostaining in human osteoblasts suggests that this protein may participate in pH regulation and acid efflux in these cells. Our data also suggests that the NHE1 and NHE3 proteins are differently distributed in osteoblasts. The pattern of NHE1 staining appears to be particulate and predominantly intracellular whereas NHE3 is present in the plasma membrane and endomembrane compartments. Immunolocalization and subcellular fractionation studies in other tissues have shown that NHE1 and other antiporters are not distributed homogeneously or even exclusively in the plasma membrane (Orlowski and Grinstein, 1997). NHE1 has been found along the borders of lamellipodia (Grinstein et al., 1993) and in endomembrane compartments of renal (Van Dyke, 1995) and hepatic (Hensley et al., 1990) endosomes. Indeed, NHE3 has also been found in internalized membranes, subapical membrane vesicles and juxtanuclear clusters of renal epithelial cells (Orlowski and Grinstein, 1997). The presence of antiporters in endosomes and subapical membrane vesicles suggests that intracellular stores could serve as a functional reservoir of spare Na + /H + exchangers that may be called upon under certain physiological circumstances. Thus, in human osteoblasts NHE proteins are compartmentalized in distinct subcellular locations. Whether they perform specific intracellular functions or remain quiescent during storage remains to be determined. It is
Fig. 4. Restriction digestion of the osteoblast AE PCR product. The PCR product shown in Fig. 3 was digested with AvaII, producing 2 detectable fragments of 240 bp and 270 bp which is more consistent with the sizes of the predicted digestion products for AE2 than for other AE isoforms. A third predicted band of 66 bp was undetectable under the gel conditions used.
Cell Biology International, Vol. 22, No. 7/8, 1998 559
560
Fig. 5. Evidence for the presence of mRNA for isoforms 1, 3 and 4 of the cation exchanger (NHE). PCR using isoform specific NHE primers and â-Actin primers demonstrated the presence of NHE1, NHE3, NHE4 and â-Actin (positive control) mRNA. PCR products of approximately 422 bp (NHE1), 460 bp (NHE3), 501 bp (NHE4) and 310 (â-Actin) were amplified from a first strand cDNA library of human bone-derived osteoblasts.
important to discuss the possible biological roles of the various NHE isoforms in bone. In mineralizing bone Ca2+ transport and resulting changes in local Ca2+ concentrations are essential for matrix calcification. The elevated presence of hormones and growth factors and increases in intracellular Ca2+ concentrations may activate protein kinases such as protein kinase C and the Ca2+ /calmodulin dependent protein kinase II, eventually leading to phosphorylation and activation of NHEs (particularly NHE1; Sardet et al., 1990). The calcium binding protein calmodulin has been shown to bind strongly to the NHE1 cytoplasmic domain (Bertrand et al., 1994). Of the five known Na + /H + exchangers, four of them (NHE1–4) have been proposed to contain conserved calmodulin binding domains; these Na + /H + exchangers bind calmodulin in a Ca2+ dependent manner (Wakabayashi et al., 1997). Therefore, Ca2+ /calmodulin-mediated activiation of NHE1 and other Na + /H + exchanger isoforms in human bone cells (NHE3 and NHE4) may play an important role in alkalinization of bone fluid compartment during matrix mineralization. We have previously demonstrated that human bone derived osteoblasts abundantly express four isoforms of the Na + ,K + -ATPase (á1, á3, â1 and â2; Mobasheri et al., 1997). Therefore, under physiological conditions, the osteoblast plasma membrane Na + ,K + -ATPase which consists of 4 different heterodimeric protomers (á1/â1, á1/â2, á3/â1 and á3/â2) generates an inwardly directed Na + gradient to provide a constant driving force for H + extrusion by the Na + /H + exchange system.
Cell Biology International, Vol. 22, No. 7/8, 1998
The presence of three isoforms of NHE (NHE1, NHE3 and NHE4) in osteoblasts suggests that these cells exploit the steep inward Na + gradient across the plasma membrane to regulate pHi more effectively. The sensitivity of each of the NHE isoforms present to mitogenic (hormones: PTH/PTHRP; growth factors; TGF-â) and nonmitogenic stimuli (impact loading of bone, osmotic changes, shear stress, cell spreading and migration) suggests that the regulation of pHi in human bone derived osteoblasts is subject to finely tuned regulatory mechanisms and multiple second messenger pathway systems. Furthermore, the fact that human bone derived osteoblast express several distinct forms of the Na + /H + exchange system compared to established osteoblast like cell lines that express only one isoform of the Na + /H + exchange system (NHE1) suggest that cell lines do not represent ideal physiological models for studies of pH regulation and ion transport in bone. The specific expression of the anion exchanger AE2 in osteoblasts was suspected before this study was initiated. We have previously detected abundant expression of the AE2 protein in related chondrocytes where the AE1 and AE3 isoforms were undetectable at the mRNA and protein levels (Golding et al., 1997). The absence of AE1 and AE3 was not a surprising discovery as the AE1 isoform appears to be unique to the erythropoietic system and the basolateral membrane of collecting cells in the kidney and AE3 is mainly expressed by cardiomyocytes and neuronal cells (Alper, 1994). In this study the AE2 isoform was abundantly detected in cultured human osteoblasts. However, the subcellular ‘organellar’ localization of AE2 was markedly different from that of the NHE proteins. The concentration of negatively charged anions in the extracellular matrix of bone and cartilage is relatively low and thus it is unlikely that any AE protein would function as a plasma membrane Cl /HCO3 exchanger (Wilkins and Hall, 1992) or participate in base extrusion. Studies performed by Redhead and Baker (1988) strongly indicate that sodium-independent AE systems exist in cultured rat calvarial osteoblasts but they do not participate directly in pHi regulation; recovery from intracellular acid loading generated by exposure to NH4Cl in these cells was Na + dependent and blocked by the Na + /H + exchanger inhibitor amiloride but unaffected by the anion exchange inhibitors 4-acetamido-4 -isothiocyanatostilbene-2,2 -disulfonic acid (SITS) and 4,4 -diisothiocyanato-stilbene-2,2 -disulfonic acid (DIDS). Due to its low abundance in the plasma membrane of human osteoblasts, it is more likely
Cell Biology International, Vol. 22, No. 7/8, 1998
that the AE2 isoform in osteoblasts is modified for the uptake of sulfate (SO42 ) and concentrating this anion in the Golgi complex where sulfation reactions routinely occur on numerous extracellular matrix macromolecules. This isoform of AE has been shown to mediate SO42 transport in reconstituted proteoliposomes and its steep pH dependence of SO42 transport; maximal at pH 5.5 and reduced to less than 10% (of the value at pH 5.5) at pH 8.5 suggests that AE2 participates in the pH regulation of SO42 transport (Sekler et al., 1995). This would explain the absence of the AE2 isoform in the plasma membrane of human osteoblasts and the reports of non-participation of anion exchange mechanism in pH regulation in other osteoblast models. In summary, our findings suggest that one AE and three NHE proteins are expressed in adult human bone derived osteoblasts. Although it is unlikely that other isoforms of AE are present in osteoblasts, the expression of novel isoforms of NHE cannot be ruled out at this stage. Exposure to pharmacologically active compounds such as amiloride and its analogues and a number of growth factors may affect the expression, abundance and turnover of the NHE proteins in bone as has been shown in other tissues. The presence of multiple NHE isoforms in human osteoblasts suggests that intracellular pH regulation in these cells may be achieved more effectively by several isozymes of Na + /H + exchange. The abundant intracellular localization of AE2 in osteoblasts is an intriguing observation and biochemical studies are required to determine its precise role in endomembrane compartments. If the abundant AE2 present in internal membranes of the osteoblast is truly functional, it is tempting to speculate that this protein is primarily involved in organellar sulphate transport and may play a role in organellar anion exchange where transport and accumulation of SO43 is important for sulfation of secreted matrix macromolecules. ACKNOWLEDGEMENTS The authors wish to thank Dr Seth L. Alper (Molecular Medicine and Renal Units, Beth Israel Hospital, Harvard Medical School, Boston) and Dr Jacques Pouyssegur (Centre de Biochemie, Centre National de la Recherche Scientifique, Nice) for their generous gifts of polyclonal antibodies against AE2 and NHE1 respectively. The generosity of Dr Mark Donowitz and Dr Chung-Ming Tse (Departments of Medicine and Physiology, The
561
Johns Hopkins University School of Medicine, Baltimore) is acknowledged for providing polyclonal antibodies against NHE2 and NHE3. The Nuffield Department of Orthopaedic Surgery, Nuffield Orthopaedic Centre, Oxford University supported most of the work described in this paper through grants to M.J.O.F. A Wellcome Trust programme grant to J.C.E. and a scholarship to A.M. from the Arthritis Research Campaign U.K. also funded this study. REFERENCES A SL, 1994. The band 3-related AE anion exchanger gene family. Cell Physiol Biochem 4: 265–281. A A, O J, G D, 1995. Parathyroid hormone and parathyroid hormone-related peptide activate the Na + /H + exchanger NHE-1 isoform in osteoblastic cells (UMR-106) via a cAMP-dependent pathway. J Biol Chem 270: 23,166–23,172. B B, W S, I T, P´ J, S M, 1994. The Na + /H + exchanger isoform 1 (NHE 1) is a novel member of the calmodulin-binding proteins. Identification and characterization of calmodulinbinding sites. J Biol Chem 269: 13,703–13,709. B JL, E EF, M J, K P, M KG, P JT, R BL, K R, 1988. Epitopes of the human erythrocyte Ca2+ -Mg2+ -ATPase pump in human osteoblast-like cell plasma membranes. J Clin Endocrinol Metab 67: 1299–1304. B P, F M, L K, B M, P M, 1995. RT-PCR analysis of Na + /H + exchanger mRNAs in rat medullary thick ascending limb. Am J Physiol 268: F1224–F1228. B D, L J, ML M, S-T A, T R, A S, 1996. Antigen retrieval in cryostat tissue sections and cultured cells by treatment with sodium dodecyl sulfate (SDS). Histochem Cell Biol 105: 261–267. C DC, F JG, 1995. Functional significance of the beta subunit for heterodimeric P-type ATPases. J Exp Biol 198: 1–17. D A, N Z, K R, 1992. Hyperosmotic activation of the Na + -H + exchanger in a rat bone cell line: temperature dependence and activation pathways. J Physiol 456: 503–518. G S, M A, W RJ, E JC, 1997. Molecular identification of anion exchanger isoform AE2 in articular chondrocytes. Trans Am Orthop Res Soc 22: 10. G CS, T AH, J, 1992. Mechanisms of activation of Na + /H + exchange in human osteoblast-like SaOS-2 cells. Biochem J 288: 137–143. G J, K CR, 1992. Role of calcium and cAMP messenger systems in intracellular pH regulation of osteoblastic cells. Am J Physiol 262: C111–C121. G AE, H JN, A JE, 1988. Differentiation of muscle, fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell population: effect of dexamethasone. J Cell Biol 106: 2139–2151. G AE, H JN, A JE, 1990. Continuously growing bipotential and monopotential myogenic, adipogenic, and chondrogenic subclones isolated from
562
the multipotential RCJ 3.1 clonal cell line. Dev Biol 142: 313–318. G S, R D, M MJ, 1989. Na + /H + exchanger and growth factor-induced cytosolic pH change. Role for cellular proliferation. Biochim Biophys Acta 988: 73–91. G S, W M, W TK, D GP, O J, P J, W DC, F JK, 1993. Focal localization of the NHE-1 isoform of the Na + /H + antiport: assessment of effects on intracellular pH. EMBO J 12: 5209–5218. G R, B JN, 1995a. The isolation and culture of cells from explants of human trabecular bone. Calcif Tissue Int 56 (Suppl 1): S8–S10. G R, J CJ, T JT, 1995b. Human bone tissue formation in diffusion chamber culture in vivo by bone-derived cells and marrow stromal fibroblastic cells. Bone 16: 597–601. G A, S CJ, B WF, 1994. Effects of CGRP, forskolin, PMA, and ionomycin on pHi dependence of Na-H exchange in UMR-106 cells. Am J Physiol 266: C1083–C1092. H CB, B ME, M AK, 1990. Subcellular distribution of Na + /H + antiport activity in rat renal cortex. Kidney Int 37: 707–716. K RR, L HF, 1985. Primary structure and transmembrane orientation of the murine anion exchange protein. Nature 316: 234–238. K RR, L BS, S DM, L AE, M CW, S K, 1989. Regulation of intracellular pH by a neuronal homolog of the erythrocyte anion exchanger. Cell 59: 927–937. K KE, S GE, 1989. Primary structure of the rat kidney band-3 anion exchange protein deduced from a cDNA. J Biol Chem 264: 8185–8192. K KE, N PR, S GE, 1990. cDNA cloning and tissue distribution of mRNAs for two proteins that are related to the band 3 Cl /HCO3 exchanger. J Biol Chem 265: 462–471. M JF, A A, P J, 1997. Molecular cloning and characterization of the human AE2 anion exchanger (SLC4A2) gene. Genomics 39: 74–85. M A, G S, P AE, H AC, U JPG, E JC, F MJO, 1997. Molecular characterisation of isoforms of the Na + , K + -ATPase in primary human bone derived osteoblasts. Trans Am Orthop Res Soc 22: 85. O J, G S, 1997. Na + /H + exchangers of mammalian cells. J Biol Chem 272: 22,373–22,376.
Cell Biology International, Vol. 22, No. 7/8, 1998
P AM, 1984. The cellular basis of bone remodelling: the quantum concept re-examined in light of recent advances in the cell biology of bone. Calcif Tissue Int 36: S37–S45. R CR, B PF, 1988. Control of intracellular pH in rat calvarial osteoblasts: coexistence of both chloridebicarbonate and sodium-hydrogen exchange. Calcif Tissue Int 42: 237–242. R IR, C R, A LV, H KA, 1988. Parathyroid hormone depresses cytosolic pH and DNA synthesis in osteoblast-like cells. Am J Physiol 255: E9–E15. S C, C L, F A, P J, 1990. Growth factors induce phosphorylation of the Na + /H + antiporter, a glycoprotein of 110 kD. Science 247: 723–726. S I, L RS, M T, K RR, 1995. Sulfate transport mediated by the mammalian anion exchangers in reconstituted proteoliposomes. J Biol Chem 270: 11,251–11,256. S-T A, S C, P J, S MA, B D, A SL, 1994. Immunolocalization of anion exchanger AE2 and cation exchanger NHE-1 in distinct adjacent cells of gastric mucosa. Am J Physiol 266: C559–C568. S T, K J, F M, F T, 1992. Second messenger signaling in the regulation of cytosolic pH and DNA synthesis by parathyroid hormone (PTH) and PTHrelated peptide in osteoblastic osteosarcoma cells: role of Na + /H + exchange. J Cell Physiol 152: 28–34. T JT, 1980. In: Urist MR, ed. Fundamental and Clinical Bone Physiology. J.B. Lippincott Company, 45–82. T C-M, L SA, Y CHC, B SR, N S, P J, D M, 1994. Moelcular Properties, Kinetics and Regulation of Mammalian Na + / H + Exchangers. Cell Physiol Biochem 4: 282–300. T C-M, M AI, Y VW, W AJM, P J, S C, P J, D M, 1991. Molecular cloning of cDNA encoding the rabbit ileal villus cell basolateral membrane Na + /H + exchanger. EMBO J 10: 1957–1967. V D RW, 1995. Na + /H + exchange modulates acidification of early rat liver endocytic vesicles. Am J Physiol 269: C943–C954. W S, S M, P J, 1997. Molecular physiology of vertebrate Na + /H + exchangers. Physiol Rev 77: 51–74. W RJ, H AC, 1992. Measurement of intracellular pH in isolated bovine articular chondrocytes. Exp Physiol 77: 521–524.
EXPRESSION OF CATION EXCHANGER NHE AND ANION EXCHANGER AE ISOFORMS IN PRIMARY HUMAN BONE-DERIVED OSTEOBLASTS A. MOBASHERI1,2,3, S. GOLDING2, S. N. PAGAKIS2, K. CORKEY2, A. E. POCOCK3, B. FERMOR3, M. J. O’BRIEN2, R. J. WILKINS2, J. C. ELLORY2 and M. J. O. FRANCIS3* 1
Department of Biomedical Sciences, School of Biosciences, University of Westminster, 115 New Cavendish Street, London W1M 8JS; 2University Laboratory of Physiology, University of Oxford, Parks Road, Oxford OX1 3PT; and 3Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Headington, Oxford OX3 7LD, U.K. Received 11 July 1997, accepted 26 August 1998
The authors used isoform-specific antibodies against cation (NHE) and anion (AE) exchange isoforms in order to establish their specific expression and localization in dispersed human bone-derived cells. Immunocytochemical preparations of permeabilized osteoblasts probed with polyclonal antibodies were optically analysed by conventional immunofluorescence and confocal laser scanning microscopy. These techniques demonstrated the abundant presence of epitopes of the cation exchangers NHE1 and NHE3 and the anion exchanger AE2 in these cells. The NHE1 and NHE3 isoform proteins were predominantly located in subplasmalemmal and nucleoplasmic vesicles. The AE2 isoform was densely localized to a subcellular location characteristic of the Golgi complex. The molecular identity of the AE and NHE isoforms was investigated by RT-PCR that confirmed the presence of NHE1 and NHE3 transcripts in addition to NHE4. RT-PCR and diagnostic restriction analysis of amplified AE cDNA established preferential AE2 expression. Since AE2 has been shown to act as a sulfate transporter at low pH, it is possible that it performs this function in the osteoblast Golgi complex where sulfation reactions occur post-translationally on numerous extracellular matrix macromolecules prior to secretion and mineralization. The Na + /H + exchanger proteins are regulated by mitogenic and non-mitogenic stimuli in the osseus environment and are involved in the large fluxes of ions and protons that necessarily occur during bone formation and resorption and thus play an important role in intracellular ion homeostasis in osteoblasts. 1998 Academic Press
K: osteoblast; Na + /H + exchanger; NHE; anion exchanger; AE; isoform; immunofluorescence; confocal microscopy; RT-PCR; alkaline phosphatase; bone mineralization, pH regulation
INTRODUCTION Bone, the skeletal support system of the vertebrate body, is a dynamic tissue that is continually formed, resorbed and remodelled. Osteoblasts are highly specialized mononuclear cuboidal cells responsible for the formation of organic bone matrix and its subsequent mineralization (Triffitt, *To whom correspondence should be addressed: Dr Martin J. O. Francis, University Lecturer in Biochemistry, Nuffield Department of Orthopaedic Surgery, University of Oxford, Nuffield Orthopaedic Centre, Headington, Oxford OX3 7LD; E-mail: [email protected] 1065–6995/98/070551+12 $30.00/0
1980; Parfitt, 1984). They originate from multipotential mesenchymal cells, which also give rise to closely related fibroblasts and chondroblasts and somewhat less related myocytes and adipocytes (Grigoriadis et al., 1988, 1990). Osteoblasts actively synthesize and secrete an extracellular matrix consisting of type I collagen, proteoglycan ground substance and non-collagenous proteins. The mineralization process is initiated by osteoblasts through the polarized secretion of matrix vesicles loaded with alkaline phosphatase, an enzyme responsible for calcification of the matrix. 1998 Academic Press
552
We are interested in the molecular identity and isoforms of active and passive transport proteins involved in pH regulation and ion homeostasis in osteoblasts. In this study, we have concentrated our efforts on candidate plasma membrane transport systems involved in pH regulation in primary human bone-derived osteoblasts. Most mammalian cells possess multiple mechanisms of pH homeostasis responsible for maintaining a relatively stable intracellular pH (pHi)7.3. Under physiological conditions these may be divided into the acid and base extruder groups. For instance, in the archetypal cell, sodium independent Cl /HCO3 exchangers also known as band 3-related anion exchangers participate in base extrusion (Alper, 1994) whereas sodium dependent Na + /H + or cation exchangers are involved in acid efflux (Tse et al., 1994). Active transport mechanisms such as the H + ,K + -ATPase pump (Chow and Forte, 1995) complement the combined activity of the cation and anion exchangers in some cell types. In addition to intracellular pH and ion homeostasis, cation and anion exchangers play a significant role in cell volume regulation (Grinstein et al., 1989; Alper, 1994) and in epithelial cells cation exchangers are responsible for acid secretion or Na + reabsorption (Wakabayashi et al., 1997). Four cation exchanger proteins (NHE1–4) and three anion exchanger proteins (AE1–3) have been identified and characterized to date (Alper, 1994; Tse et al., 1994). Immunocytochemical studies suggest that NHE1 is the ubiquitous housekeeping isoform expressed by most epithelial and nonepithelial cells and functional studies have demonstrated its involvement in intracellular pH regulation. However, NHE2, NHE3 and NHE4 are more restricted in their distribution and are predominantly expressed in the epithelial cells of the stomach, kidney and the intestines (Wakabayashi et al., 1997). AE1 commonly referred to as band-3, is the prototype member of the anion exchange family found in the erythropoietic system and the kidney (Kopito and Lodish, 1985; Kudrycki and Shull, 1989). Transcripts for AE2 have been detected in almost all tissues examined whereas mRNA for the third member of the anion exchanger family AE3 appears to be expressed most abundantly in heart and brain (Kudrycki et al., 1990). It has recently become apparent that a number of hormones and growth factors modulate cytosolic pH (pHi) in osteoblasts, and there is some evidence that this in turn influences DNA synthesis and cellular proliferation (Reid et al., 1988). Parathyroid hormone (PTH) and parathyroid hormone-
Cell Biology International, Vol. 22, No. 7/8, 1998
related peptide (PTHRP) are important regulators of normal and pathological bone remodelling. A consequence of PTH/PTHRP action in osteoblasts is intracellular acidification followed by activation of the Na + /H + exchanger system that has been shown to play a central role in modulating pHi recovery (Redhead and Baker, 1988). PTH has been demonstrated to decrease pHi and growth in UMR-106 osteoblast-like cells (Reid et al., 1988). Depression of pHi has important consequences for bone metabolism. In osteoblasts the regulation of pHi is a critical component of hormone stimulated bone remodelling where local fluctuations occur in the osseus extracellular microenvironment (Redhead and Baker, 1988). At the cellular level, Na + /H + exchanger activity is regulated by a variety of stimuli that activate diverse signal transduction systems including protein kinase A, protein kinase C, tyrosine kinase and Ca2+ /calmodulin dependent protein kinase II (Wakabayashi et al., 1997). This extensive functional and regulatory diversity appears to be generated by tissue-specific expression of Na + /H + exchanger isoforms in a variety of tissues. A recent report by Azarani and co-workers (1995) suggests that the established rat osteoblastic cell line UMR-106 expresses only one isoform of the Na + /H + exchanger, namely NHE1, that is activated by PTH/PTHRP via a cyclic AMP dependent pathway. Other investigators working on similar and unrelated cell lines have suggested that more diverse signalling pathways are in operation in bone derived cells (Reid et al., 1988; Sugimoto et al., 1992; Green and Kleeman, 1992; Gupta et al., 1994). The existence of various signal transduction pathways may be reflected by the bone cell-specific expression of multiple NHE isoform proteins differentially responsive to distinct signalling pathways. The conflicting hypotheses regarding the expression and regulation of Na + /H + exchange in bone cells prompted us to investigate the expression of Na + /H + exchanger (and anion exchanger) systems in human bone-derived osteoblasts at the protein and mRNA levels to gain an insight into the process of human osteoblast pH regulation. To this end, we used isoform-specific antibodies to determine which cation and anion exchange systems may be important to these cells in the context of human bone mineralization and investigated their expression and cellular localization by immunofluorescence confocal laser scanning microscopy and the reverse transcriptase polymerase chain reaction (RT-PCR). There may be important species specific differences in the expression of these systems in bone and thus the work described here aims to unravel the
Cell Biology International, Vol. 22, No. 7/8, 1998
physiological complexities of pH regulation in human bone-derived osteoblastic cells at the molecular level. MATERIALS AND METHODS Cell preparation and culture Human bone derived osteoblasts were grown out from trabecular bone chips and cultured essentially as previously described (Gundle et al., 1995a,b). The approval of the Hospital Medical Staff Ethics Committee was obtained for all the experiments described. Briefly, samples of trabecular bone were obtained from the greater trochanter of patients undergoing surgery for hip replacement. Primary cultures of human bone derived osteoblasts were obtained by outgrowth from bone chip explants in DMEM supplemented with 10 m NaHCO3, 25 m HEPES, 2 m -glutamine, 10% heat treated foetal calf serum, 30 units/ml benzylpenicillin and 15 µg/ml streptomycin sulfate, 100 n -ascorbic acid-2-phosphate (Asc-2-P) and 10 n dexamethasone pH 7.3 in 80-cm2 culture flasks in a humidified incubator at 37C and 5% CO2, 95% air until they reached confluence. They were then detached using sequential collagenase and trypsin digestion and plated into 24 well plates containing microscope glass coverslips. For comparing alkaline phosphatase activities, primary cultures of human anterior cruciate ligament (ACL) cells were obtained by outgrowth from explants of normal ACL obtained at knee replacement for osteoarthritis in Dulbecco’s minimum essential medium (DMEM) supplemented with 10 m NaHCO3, 25 m HEPES, 2 m -glutamine, 10% heat treated foetal calf serum, 30 units/ml benzylpenicillin and 15 µg/ml streptomycin sulfate 100 n -ascorbic acid-2-phosphate (ASC-2-P), pH 7.3 in 80-cm2 culture flasks in a humidified incubator at 37C and 5% CO2/95% air until they reached confluence. They were then detached using sequential collagenase and trypsin digestion and plated into 24-well plates. Isoform-specific antibodies A monoclonal antibody raised against the human erythrocyte anion exchange AE1 isoform was purchased from Sigma (B9277). Dr Seth Alper, Harvard Medical School, Boston donated affinitypurified rabbit polyclonal antibodies against the C-terminal peptide of AE1 and AE2. This antibody was originally raised against the C-terminal peptide
553
of AE2 but cross-reacts with AE1 (Stuart-Tilley et al., 1994). Polyclonal antiserum against NHE1 was obtained from Dr Jacques Pouyssegur, Centre National de la Recherche Scientifique, Nice. Polyclonal antisera against NHE2 and NHE3 were obtained from Dr Ming Tse and Dr Mark Donowitz, Johns Hopkins University Medical School, Baltimore. A monoclonal antibody against the human erythrocyte Mg2+ -dependent Ca2+ ATPase (Clone 5F10) was purchased from Sigma (A7952) and the abundant expression of this protein in human bone derived cells was used as a marker for the osteoblastic phenotype (Borke et al., 1988) in addition to quantitative alkaline phosphatase expression. Confocal image acquisition and immunofluorescence microscopy Osteoblasts grown on circular glass coverslips were fixed in freshly prepared 3.7% paraformaldehyde in PBS for 10 min before permeabilizing in 1% SDS in PBS (Brown et al., 1996) for a further 5 min. Non-specific binding sites in the permeabilized cells were blocked overnight with 10–20% normal goat serum (NGS; Sigma) in PBS at 4C. The cells were then probed with primary isoform-specific antibodies and secondary anti-mouse and anti-rabbit IgG (Sigma) conjugated to crystalline tetramethylrhodamine isothiocyanate (TRITC) appropriately diluted in PBS containing 1% NGS before mounting in non-fluorescent mounting media (Hydromount , National Diagnostics). The cells were examined using a Leica confocal laser scanning microscope fitted with an oil immersion objective (100). For excitation of TRITC, an argon–krypton laser was used. With the pinhole filter combination, laser power setting and TRITC filter, optical images were recorded with intervals of approximately 0.5 ìm in the z direction. The images were median filtered (33) to remove noise and saved in either a 512512 pixel TIFF (tagged image file format) format or in a 256256 pixel TIFF format using the public domain NIH Image Analysis software for Macintosh. Filtered images were printed on a Codonics dye sublimation photographic quality printer with a minimum resolution of 600 dpi. Total RNA isolation, cDNA synthesis and RT-PCR RT-PCR was necessary to distinguish between AE1 and AE2 expression as the polyclonal antibody available cross-reacts with both protein isoforms.
554
Cell Biology International, Vol. 22, No. 7/8, 1998
Table 1. DNA sequence of oligonucleotide primers for RT-PCR. Sequences are from 5 to 3 direction and the estimated size of the RT-PCR product is indicated. In the AE row, isoform-independent oligonucleotide primers complementary to mRNA sequences within membrane spanning domains 1 and 6 of AE are shown. The predicted size of all expected AE isoforms is given in parentheses. The neutral base inosine (I) was used in the AE forward primer due to degeneracy in codon usage between AE isoforms Isoform NHE1 NHE2 NHE3 NHE4 AE â-actin
Sequence
PCR Product predicted size (bp)
TCTGCCGTCTCAACTGTCTCTA (forward) CCCTTCAACTCCTCATTCACCA (reverse) GCAGATGGTAATAGCAGCGA (forward) CCTTGGTGGGGGCTTGGGTG (reverse) ACGTCCAGGACCCCTACATC (forward) TGGACCTCCTCAAACACGGCCAG (reverse) GGCTGGGATTGAAGATGTATGT (forward) GCTGGCTGAGGATTGCTGTAA (reverse) TTC ATC TAC TTT GCI GCC CTG (forward) CGA CAG GAG GGC CGT GTT GGG (reverse) TTC AAC TCC ATC ATG AAG AAG TGT GAC GTG (forward) CTA AGT CAT AGT CCG CCT AGA AGC ATT (reverse)
Total RNA was prepared from the osteoblasts using a Qiagen RNeasy kit (Qiagen). Before PCR amplification an oligo-dT primed cDNA library was prepared from total RNA using SuperScript II reverse transcriptase (Life Technologies). PCR of AE was performed as previously described (Golding et al., 1997) using isoform-independent oligonucleotide primers complementary to mRNA sequences within membrane spanning domains 1 and 6 of AE. The primer sequences used were as follows: AEM1-F(orward) primer 5 TTC ATC TAC TTT GCI GCC CTG-3 and AEM6-R(everse) primer 5 -CGA CAG GAG GGC CGT GTT GGG-3 . Inosine (I), a neutral base, was used in the primers due to degeneracy in codon usage between AE isoforms. The following â-actin primer sequences were used as a positive PCR control: 5 -TTC AAC TCC ATC ATG AAG AAG TGT GAC GTG-3 (Forward) and 5 -CTA AGT CAT AGT CCG CCT AGA AGC ATT-3 (Reverse). The PCR cycling protocol consisted of 1 cycle of 95C3 min, then 40 cycles of 95C1 min, 55C1 min, 72C2 min, followed by 1 final cycle of 72C5 min. Isoformspecific primers and cycle conditions for NHE PCR were as described in Borensztein et al. (1995) (Table 1), but with the NHE3 primer sequences replaced by 5 -ACG TCC AGG ACC CCT ACA TC-3 (forward) and 5 -TGG ACC TCC TCA AAC ACG GCC AG-3 (reverse) as the primers described were found to be rat-specific. PCR was also performed using cDNA prepared in the absence of reverse transcriptase as a negative control. Following amplification, AE PCR products
422 310 460 501 339 (AE1), 573 (AE2), 457 (AE3) 310
were purified and concentrated using the Wizard PCR Purification System (Promega, U.K.), and resuspended in 18 ìl of nuclease-free water to which was added 2 ìl 10 restriction enzyme buffer (supplied) and 5 U AvaII restriction enzyme (Boehringer Mannheim, U.K.). Samples were then incubated at 37C for 3 h before separation on a 4% NuSieve GTG agarose (FMC, Flowgen, U.K.) gel in TAE buffer (40 m Tris/acetate, 1 m EDTA) for 4 h at 40 V. DNA assay. Solubilized DNA was measured fluorometrically. DNA isolated from cells was solubilized with 10 m EDTA, adjusted to pH 12.3 with NaOH and incubated at 37C for 30 min. The pH was adjusted to 7.0 with 1 KH2PO4. Hoechst 33258 (300 ng) was added and fluorescence measured using a Perkin Elmer LS 30 fluorimeter with excitation and emission wavelengths set at 355 nm and 450 nm and compared to that of known standard DNA samples. Alkaline phosphatase activity. Alkaline phosphatase activity was determined by measuring the release of p-nitrophenol from p-nitrophenyl phosphate disodium as described in Sigma technical bulletin no. 104. For quantitative studies, the amount of p-nitrophenol released was determined using a Gilford System 2600 spectrophotometer at 410 nm. Sites of alkaline phosphatase activity were stained red for qualitative studies and microscopic evaluation.
Cell Biology International, Vol. 22, No. 7/8, 1998
RESULTS Immunofluorescence and confocal analysis of NHE and AE isoform expression in human osteoblasts Immunofluorescence and confocal laser scanning microscopy was utilized to investigate the specific expression and subcellular localization of epitopes of cation and anion exchange proteins abundantly expressed in osteoblasts. Isoform specific polyclonal antibodies and secondary TRITCconjugated anti-rabbit IgG were used to stain cation and anion exchange proteins in permeabilized osteoblasts in which antigenic sites were retrieved by SDS pre-treatment (Brown et al., 1996). We used a confocal microscope in addition to a conventional fluorescence microscope since it allows 3-dimensional visualization of the protein of interest and reveals more information about its morphological and subcellular localization. The immunofluorescence data revealed abundant expression of the Mg2+ -dependent Ca2+ -ATPase (PMCA1 isoform) in the plasma membrane and subplasmalemmal vesicles of paraformaldehyde fixed, SDS permeabilized osteoblasts (Fig. 2A) The expression of the alkaline phosphatase (Fig. 1) was used as a bone specific marker for the osteoblastic phenotype. Comparison of alkaline phosphatase expression levels in human bone-derived osteolasts and anterior cruciate ligament cells provided convincing evidence for a true osteoblastic phenotype in the bone-derived osteoblasts (see Table 2). The cation exchangers NHE1 and NHE3 were both present as indicated by the intense fluorescent staining of both proteins in osteoblasts (Fig. 2B, E, respectively). The fluorescent immunostaining for the NHE1 isoform was predominantly nucleoplasmic with nucleolar exclusion; the most abundant staining was present in a particulate intracellular pattern as observed by optical sectioning (Fig. 2B). This finding is concordant with the concept of ubiquitous NHE1 isoform expression but perhaps not consistent in a straightforward manner with a housekeeping acid efflux role in dispersed osteoblasts. In contrast to NHE1 the fluorescent immunostaining of the NHE3 isoform was plasmalemmal and nucleoplasmic. The monoclonal anti-human erythrocyte AE1 did not recognize any epitopes in osteoblasts (results not shown). Although no evidence for AE1 expression was found the anion exchanger AE2 isoform was abundantly expressed in osteoblasts (Fig. 2C). In contrast to the NHE proteins, most of the staining for AE2 was predominantly intracellular and polar suggesting that the bulk of the AE2 protein is in an intracellular compartment characteristic of the
555
Golgi apparatus. It is important to note that the intracellular AE2 protein was not detectable without prior permeabilization and antigen retrieval with SDS and non-permeabilized cells exhibited almost no immunoreactivity when probed with the AE2 specific antibodies. This finding is consistent with the established notion that newly biosynthesized AE2 protein is translated in the endoplasmic reticulum and further processed in the cis, medial and trans-Golgi compartments before trafficking to the plasma membrane (Alper, 1994). However, the high intracellular abundance of the AE2 isoform in osteoblasts suggests that most of the AE2 protein may not be targeted to the plasma membrane but intended for prolonged transit in the organelles. In all experiments performed, cells probed with secondary antibody alone exhibited negligible background staining even when the laser power was set to maximum (results not shown). RT-PCR analysis of AE and NHE expression in human osteoblasts We have previously used RT-PCR to demonstrate the presence of transcripts encoding AE2 in isolated articular chondrocytes (Golding et al., 1997). Here we used an identical approach to detect transcripts encoding AE2 in human osteoblasts. This was carried out to confirm the results obtained by immunofluorescence and confocal microscopy. Although the anti-erythrocyte AE1 monoclonal antibody did not recognize epitopes in osteoblasts, the polyclonal AE antibody can recognize shared epitopes on the AE1 and AE2 isoforms. Therefore, experiments were designed and performed to demonstrate the molecular identity of the AE system present in these cells. PCR amplification of cDNA between membrane spanning regions 1 and 6 of AE produced one cDNA product of a size (570 bp) consistent with the amplification of AE2 cDNA rather than AE1 which, based on other cloned sequences, would have been amplified to give a cDNA product 100 bp smaller in size (Fig. 3) (Golding et al., 1997). As further proof, the osteoblast PCR product was digested with the restriction enyme AvaII. Using the Webcutter v2.0 program (http://www.firstmarket.com/cutter/) and the published mRNA sequences for rat kidney AE1 (Kudrycki and Shull, 1989), human kidney AE2 (Medina et al., 1997) and mouse brain AE3 (Kopito et al., 1989) between MEM1 and MEM6 it was deduced that digestion by AvaII would produce cDNA fragments of 16 bp, 140 bp and 183 bp (AE1); 66 bp, 240 bp and 267 bp (AE2); and 68 bp, 277 bp and 112 bp (AE3). As can be seen in Fig. 4,
556
Cell Biology International, Vol. 22, No. 7/8, 1998
Fig. 1. A: High-power phase-contrast light micrograph of cultured human bone-derived osteoblasts. B: Expression of alkaline phosphatase by the cells demonstrated by histochemical staining. Original magnification=400, reproduced at 95%.
Fig. 2. Confocal laser scanning microscope (CLSM) optical sections (0.5 ìm, in the z direction) through osteoblasts probed with polyclonal antibodies against the Mg2+ -dependent Ca2+ -ATPase, the cation exchanger NHE1 and anion exchanger AE2 isoforms. A: The polar expression of the Mg2+ dependent Ca2+ -ATPase (PMCA1 isoform) in human primary bone derived osteoblasts is one marker for the osteoblastic phenotype (N=nucleus; bar=10 ìm). B: Cation exchange (NHE1 isoform) distribution in human primary bone derived osteoblasts. The staining is predominantly nucleoplasmic with nucleolar exclusion. C: Distribution of the band-3 like anion exchanger (AE2 isoform) in human primary bone derived osteoblasts. The pattern of immunofluorescent staining is typically organellar and the bulk of the AE2 protein is most likely in the Golgi complex (arrows indicate intracellular compartment). D: Control cells where primary antibody was omitted. E: Conventional immunofluorescence micrograph of osteoblasts probed with a polyclonal NHE3 specific antibody and secondary TRITC-conjugated IgG.
558
Cell Biology International, Vol. 22, No. 7/8, 1998
Table 2. Expression of alkaline phosphatase by human bone derived osteoblasts (HBDO) compared with nonosteoblastic anterior cruciate ligament (ACL) cells compared over a period of 12 days. These quantitative measurements show roughly a seven-fold difference in alkaline phosphatase expression in HBDO cells compared to ACL cells on day 6 and a two-fold increase on day 12. S.D., Standard deviation of the mean; n=9. Units of alkaline phosphatase expression shown are per ìg of culture DNA Days in culture
ACL
..
HBDO
..
3 6 9 12
15.64 18.50 23.66 56.03
4.46 0.62 0.17 6.70
77.0 137.80 114.45 123.80
6.50 11.54 8.74 11.51
AvaII digestion of the osteoblast PCR product produced fragments of a size consistent with the original product being AE2. Using RT-PCR, we have also been able to amplify NHE isoforms 1, 3 and 4 from osteoblasts (Fig. 5) using protocols described elsewhere (Borensztein et al., 1995). However, we were unable to detect transcripts for NHE2 by RT-PCR (not shown). Therefore, taken together these results indicated that human primary derived osteoblasts transcribe the AE2 isoform of AE and NHE isoforms 1, 3 and 4. Overall, the results presented here suggest that human bone derived cells possess at least three distinct Na + /H + exchangers (NHE1, NHE3 and NHE4) and at least one anion exchanger (AE2). DISCUSSION In this study we have demonstrated that human bone derived osteoblasts express the NHE1 and
Fig. 3. RT-PCR amplified a 570 bp cDNA fragment from cDNA library prepared from the human bone derived osteoblast that corresponds to regions between membrane spanning regions 1 and 6 of the AE2 isoform.
NHE3 isoforms of the Na + /H + exchanger system at the protein level and the NHE4 isoform at the mRNA level. Although we have presented no functional evidence for an acid efflux ‘housekeeping’ role for NHE1, our data indicates that this Na + / H + exchanger is present in human osteoblasts. The expression of NHE1 in osteoblasts was expected as Northern blot analysis in UMR-106 osteoblastic cells (Azarani et al., 1995) and immunocytochemistry in almost all other cells and tissues examined have found this protein (Tse et al., 1991). The extensive literature on NHE1 expression clearly indicates that this isoform is ubiquitously expressed and plays a ‘housekeeping’ role in almost all cells (for reviews see Orlowski and Grinstein, 1997 and Wakabayashi et al., 1997). Physiologically, the presence of cation exchange mechanisms has been previously demonstrated by numerous groups in osteoblast-like cells (Redhead and Baker, 1988; Reid et al., 1988; Sujimoto et al., 1988; Dascalu et al., 1992; Graham and Tashjian Jr, 1992). The NHE1 exchanger is activated by changes in cell volume (particularly hyperosmolarity or cell shrinkage) and is entirely inhibited by amiloride and its analogues in osteoblasts and its immunostaining in human osteoblasts suggests that this protein may participate in pH regulation and acid efflux in these cells. Our data also suggests that the NHE1 and NHE3 proteins are differently distributed in osteoblasts. The pattern of NHE1 staining appears to be particulate and predominantly intracellular whereas NHE3 is present in the plasma membrane and endomembrane compartments. Immunolocalization and subcellular fractionation studies in other tissues have shown that NHE1 and other antiporters are not distributed homogeneously or even exclusively in the plasma membrane (Orlowski and Grinstein, 1997). NHE1 has been found along the borders of lamellipodia (Grinstein et al., 1993) and in endomembrane compartments of renal (Van Dyke, 1995) and hepatic (Hensley et al., 1990) endosomes. Indeed, NHE3 has also been found in internalized membranes, subapical membrane vesicles and juxtanuclear clusters of renal epithelial cells (Orlowski and Grinstein, 1997). The presence of antiporters in endosomes and subapical membrane vesicles suggests that intracellular stores could serve as a functional reservoir of spare Na + /H + exchangers that may be called upon under certain physiological circumstances. Thus, in human osteoblasts NHE proteins are compartmentalized in distinct subcellular locations. Whether they perform specific intracellular functions or remain quiescent during storage remains to be determined. It is
Fig. 4. Restriction digestion of the osteoblast AE PCR product. The PCR product shown in Fig. 3 was digested with AvaII, producing 2 detectable fragments of 240 bp and 270 bp which is more consistent with the sizes of the predicted digestion products for AE2 than for other AE isoforms. A third predicted band of 66 bp was undetectable under the gel conditions used.
Cell Biology International, Vol. 22, No. 7/8, 1998 559
560
Fig. 5. Evidence for the presence of mRNA for isoforms 1, 3 and 4 of the cation exchanger (NHE). PCR using isoform specific NHE primers and â-Actin primers demonstrated the presence of NHE1, NHE3, NHE4 and â-Actin (positive control) mRNA. PCR products of approximately 422 bp (NHE1), 460 bp (NHE3), 501 bp (NHE4) and 310 (â-Actin) were amplified from a first strand cDNA library of human bone-derived osteoblasts.
important to discuss the possible biological roles of the various NHE isoforms in bone. In mineralizing bone Ca2+ transport and resulting changes in local Ca2+ concentrations are essential for matrix calcification. The elevated presence of hormones and growth factors and increases in intracellular Ca2+ concentrations may activate protein kinases such as protein kinase C and the Ca2+ /calmodulin dependent protein kinase II, eventually leading to phosphorylation and activation of NHEs (particularly NHE1; Sardet et al., 1990). The calcium binding protein calmodulin has been shown to bind strongly to the NHE1 cytoplasmic domain (Bertrand et al., 1994). Of the five known Na + /H + exchangers, four of them (NHE1–4) have been proposed to contain conserved calmodulin binding domains; these Na + /H + exchangers bind calmodulin in a Ca2+ dependent manner (Wakabayashi et al., 1997). Therefore, Ca2+ /calmodulin-mediated activiation of NHE1 and other Na + /H + exchanger isoforms in human bone cells (NHE3 and NHE4) may play an important role in alkalinization of bone fluid compartment during matrix mineralization. We have previously demonstrated that human bone derived osteoblasts abundantly express four isoforms of the Na + ,K + -ATPase (á1, á3, â1 and â2; Mobasheri et al., 1997). Therefore, under physiological conditions, the osteoblast plasma membrane Na + ,K + -ATPase which consists of 4 different heterodimeric protomers (á1/â1, á1/â2, á3/â1 and á3/â2) generates an inwardly directed Na + gradient to provide a constant driving force for H + extrusion by the Na + /H + exchange system.
Cell Biology International, Vol. 22, No. 7/8, 1998
The presence of three isoforms of NHE (NHE1, NHE3 and NHE4) in osteoblasts suggests that these cells exploit the steep inward Na + gradient across the plasma membrane to regulate pHi more effectively. The sensitivity of each of the NHE isoforms present to mitogenic (hormones: PTH/PTHRP; growth factors; TGF-â) and nonmitogenic stimuli (impact loading of bone, osmotic changes, shear stress, cell spreading and migration) suggests that the regulation of pHi in human bone derived osteoblasts is subject to finely tuned regulatory mechanisms and multiple second messenger pathway systems. Furthermore, the fact that human bone derived osteoblast express several distinct forms of the Na + /H + exchange system compared to established osteoblast like cell lines that express only one isoform of the Na + /H + exchange system (NHE1) suggest that cell lines do not represent ideal physiological models for studies of pH regulation and ion transport in bone. The specific expression of the anion exchanger AE2 in osteoblasts was suspected before this study was initiated. We have previously detected abundant expression of the AE2 protein in related chondrocytes where the AE1 and AE3 isoforms were undetectable at the mRNA and protein levels (Golding et al., 1997). The absence of AE1 and AE3 was not a surprising discovery as the AE1 isoform appears to be unique to the erythropoietic system and the basolateral membrane of collecting cells in the kidney and AE3 is mainly expressed by cardiomyocytes and neuronal cells (Alper, 1994). In this study the AE2 isoform was abundantly detected in cultured human osteoblasts. However, the subcellular ‘organellar’ localization of AE2 was markedly different from that of the NHE proteins. The concentration of negatively charged anions in the extracellular matrix of bone and cartilage is relatively low and thus it is unlikely that any AE protein would function as a plasma membrane Cl /HCO3 exchanger (Wilkins and Hall, 1992) or participate in base extrusion. Studies performed by Redhead and Baker (1988) strongly indicate that sodium-independent AE systems exist in cultured rat calvarial osteoblasts but they do not participate directly in pHi regulation; recovery from intracellular acid loading generated by exposure to NH4Cl in these cells was Na + dependent and blocked by the Na + /H + exchanger inhibitor amiloride but unaffected by the anion exchange inhibitors 4-acetamido-4 -isothiocyanatostilbene-2,2 -disulfonic acid (SITS) and 4,4 -diisothiocyanato-stilbene-2,2 -disulfonic acid (DIDS). Due to its low abundance in the plasma membrane of human osteoblasts, it is more likely
Cell Biology International, Vol. 22, No. 7/8, 1998
that the AE2 isoform in osteoblasts is modified for the uptake of sulfate (SO42 ) and concentrating this anion in the Golgi complex where sulfation reactions routinely occur on numerous extracellular matrix macromolecules. This isoform of AE has been shown to mediate SO42 transport in reconstituted proteoliposomes and its steep pH dependence of SO42 transport; maximal at pH 5.5 and reduced to less than 10% (of the value at pH 5.5) at pH 8.5 suggests that AE2 participates in the pH regulation of SO42 transport (Sekler et al., 1995). This would explain the absence of the AE2 isoform in the plasma membrane of human osteoblasts and the reports of non-participation of anion exchange mechanism in pH regulation in other osteoblast models. In summary, our findings suggest that one AE and three NHE proteins are expressed in adult human bone derived osteoblasts. Although it is unlikely that other isoforms of AE are present in osteoblasts, the expression of novel isoforms of NHE cannot be ruled out at this stage. Exposure to pharmacologically active compounds such as amiloride and its analogues and a number of growth factors may affect the expression, abundance and turnover of the NHE proteins in bone as has been shown in other tissues. The presence of multiple NHE isoforms in human osteoblasts suggests that intracellular pH regulation in these cells may be achieved more effectively by several isozymes of Na + /H + exchange. The abundant intracellular localization of AE2 in osteoblasts is an intriguing observation and biochemical studies are required to determine its precise role in endomembrane compartments. If the abundant AE2 present in internal membranes of the osteoblast is truly functional, it is tempting to speculate that this protein is primarily involved in organellar sulphate transport and may play a role in organellar anion exchange where transport and accumulation of SO43 is important for sulfation of secreted matrix macromolecules. ACKNOWLEDGEMENTS The authors wish to thank Dr Seth L. Alper (Molecular Medicine and Renal Units, Beth Israel Hospital, Harvard Medical School, Boston) and Dr Jacques Pouyssegur (Centre de Biochemie, Centre National de la Recherche Scientifique, Nice) for their generous gifts of polyclonal antibodies against AE2 and NHE1 respectively. The generosity of Dr Mark Donowitz and Dr Chung-Ming Tse (Departments of Medicine and Physiology, The
561
Johns Hopkins University School of Medicine, Baltimore) is acknowledged for providing polyclonal antibodies against NHE2 and NHE3. The Nuffield Department of Orthopaedic Surgery, Nuffield Orthopaedic Centre, Oxford University supported most of the work described in this paper through grants to M.J.O.F. A Wellcome Trust programme grant to J.C.E. and a scholarship to A.M. from the Arthritis Research Campaign U.K. also funded this study. REFERENCES A SL, 1994. The band 3-related AE anion exchanger gene family. Cell Physiol Biochem 4: 265–281. A A, O J, G D, 1995. Parathyroid hormone and parathyroid hormone-related peptide activate the Na + /H + exchanger NHE-1 isoform in osteoblastic cells (UMR-106) via a cAMP-dependent pathway. J Biol Chem 270: 23,166–23,172. B B, W S, I T, P´ J, S M, 1994. The Na + /H + exchanger isoform 1 (NHE 1) is a novel member of the calmodulin-binding proteins. Identification and characterization of calmodulinbinding sites. J Biol Chem 269: 13,703–13,709. B JL, E EF, M J, K P, M KG, P JT, R BL, K R, 1988. Epitopes of the human erythrocyte Ca2+ -Mg2+ -ATPase pump in human osteoblast-like cell plasma membranes. J Clin Endocrinol Metab 67: 1299–1304. B P, F M, L K, B M, P M, 1995. RT-PCR analysis of Na + /H + exchanger mRNAs in rat medullary thick ascending limb. Am J Physiol 268: F1224–F1228. B D, L J, ML M, S-T A, T R, A S, 1996. Antigen retrieval in cryostat tissue sections and cultured cells by treatment with sodium dodecyl sulfate (SDS). Histochem Cell Biol 105: 261–267. C DC, F JG, 1995. Functional significance of the beta subunit for heterodimeric P-type ATPases. J Exp Biol 198: 1–17. D A, N Z, K R, 1992. Hyperosmotic activation of the Na + -H + exchanger in a rat bone cell line: temperature dependence and activation pathways. J Physiol 456: 503–518. G S, M A, W RJ, E JC, 1997. Molecular identification of anion exchanger isoform AE2 in articular chondrocytes. Trans Am Orthop Res Soc 22: 10. G CS, T AH, J, 1992. Mechanisms of activation of Na + /H + exchange in human osteoblast-like SaOS-2 cells. Biochem J 288: 137–143. G J, K CR, 1992. Role of calcium and cAMP messenger systems in intracellular pH regulation of osteoblastic cells. Am J Physiol 262: C111–C121. G AE, H JN, A JE, 1988. Differentiation of muscle, fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell population: effect of dexamethasone. J Cell Biol 106: 2139–2151. G AE, H JN, A JE, 1990. Continuously growing bipotential and monopotential myogenic, adipogenic, and chondrogenic subclones isolated from
562
the multipotential RCJ 3.1 clonal cell line. Dev Biol 142: 313–318. G S, R D, M MJ, 1989. Na + /H + exchanger and growth factor-induced cytosolic pH change. Role for cellular proliferation. Biochim Biophys Acta 988: 73–91. G S, W M, W TK, D GP, O J, P J, W DC, F JK, 1993. Focal localization of the NHE-1 isoform of the Na + /H + antiport: assessment of effects on intracellular pH. EMBO J 12: 5209–5218. G R, B JN, 1995a. The isolation and culture of cells from explants of human trabecular bone. Calcif Tissue Int 56 (Suppl 1): S8–S10. G R, J CJ, T JT, 1995b. Human bone tissue formation in diffusion chamber culture in vivo by bone-derived cells and marrow stromal fibroblastic cells. Bone 16: 597–601. G A, S CJ, B WF, 1994. Effects of CGRP, forskolin, PMA, and ionomycin on pHi dependence of Na-H exchange in UMR-106 cells. Am J Physiol 266: C1083–C1092. H CB, B ME, M AK, 1990. Subcellular distribution of Na + /H + antiport activity in rat renal cortex. Kidney Int 37: 707–716. K RR, L HF, 1985. Primary structure and transmembrane orientation of the murine anion exchange protein. Nature 316: 234–238. K RR, L BS, S DM, L AE, M CW, S K, 1989. Regulation of intracellular pH by a neuronal homolog of the erythrocyte anion exchanger. Cell 59: 927–937. K KE, S GE, 1989. Primary structure of the rat kidney band-3 anion exchange protein deduced from a cDNA. J Biol Chem 264: 8185–8192. K KE, N PR, S GE, 1990. cDNA cloning and tissue distribution of mRNAs for two proteins that are related to the band 3 Cl /HCO3 exchanger. J Biol Chem 265: 462–471. M JF, A A, P J, 1997. Molecular cloning and characterization of the human AE2 anion exchanger (SLC4A2) gene. Genomics 39: 74–85. M A, G S, P AE, H AC, U JPG, E JC, F MJO, 1997. Molecular characterisation of isoforms of the Na + , K + -ATPase in primary human bone derived osteoblasts. Trans Am Orthop Res Soc 22: 85. O J, G S, 1997. Na + /H + exchangers of mammalian cells. J Biol Chem 272: 22,373–22,376.
Cell Biology International, Vol. 22, No. 7/8, 1998
P AM, 1984. The cellular basis of bone remodelling: the quantum concept re-examined in light of recent advances in the cell biology of bone. Calcif Tissue Int 36: S37–S45. R CR, B PF, 1988. Control of intracellular pH in rat calvarial osteoblasts: coexistence of both chloridebicarbonate and sodium-hydrogen exchange. Calcif Tissue Int 42: 237–242. R IR, C R, A LV, H KA, 1988. Parathyroid hormone depresses cytosolic pH and DNA synthesis in osteoblast-like cells. Am J Physiol 255: E9–E15. S C, C L, F A, P J, 1990. Growth factors induce phosphorylation of the Na + /H + antiporter, a glycoprotein of 110 kD. Science 247: 723–726. S I, L RS, M T, K RR, 1995. Sulfate transport mediated by the mammalian anion exchangers in reconstituted proteoliposomes. J Biol Chem 270: 11,251–11,256. S-T A, S C, P J, S MA, B D, A SL, 1994. Immunolocalization of anion exchanger AE2 and cation exchanger NHE-1 in distinct adjacent cells of gastric mucosa. Am J Physiol 266: C559–C568. S T, K J, F M, F T, 1992. Second messenger signaling in the regulation of cytosolic pH and DNA synthesis by parathyroid hormone (PTH) and PTHrelated peptide in osteoblastic osteosarcoma cells: role of Na + /H + exchange. J Cell Physiol 152: 28–34. T JT, 1980. In: Urist MR, ed. Fundamental and Clinical Bone Physiology. J.B. Lippincott Company, 45–82. T C-M, L SA, Y CHC, B SR, N S, P J, D M, 1994. Moelcular Properties, Kinetics and Regulation of Mammalian Na + / H + Exchangers. Cell Physiol Biochem 4: 282–300. T C-M, M AI, Y VW, W AJM, P J, S C, P J, D M, 1991. Molecular cloning of cDNA encoding the rabbit ileal villus cell basolateral membrane Na + /H + exchanger. EMBO J 10: 1957–1967. V D RW, 1995. Na + /H + exchange modulates acidification of early rat liver endocytic vesicles. Am J Physiol 269: C943–C954. W S, S M, P J, 1997. Molecular physiology of vertebrate Na + /H + exchangers. Physiol Rev 77: 51–74. W RJ, H AC, 1992. Measurement of intracellular pH in isolated bovine articular chondrocytes. Exp Physiol 77: 521–524.