We identified the human ortholog of soluble adenylyl cyclase (hsAC) in a locus linked to familial absorptive hypercalciuria and cloned it from a human cDNA library. hsAC transcripts were expressed in multiple tissues using RT-PCR and RNA blotting. RNA blot analysis revealed a predominant 5.1-kb band in a multiple human tissue blot, but three splice transcript variants were detected using RT-PCR and confirmed by performing sequence analysis. Immunoblot analysis showed 190- and 80-kDa bands in multiple human cell lines from gut, renal, and bone origins in both cytosol and membrane fractions, including Caco-2 colorectal adenocarcinomas, HEK-293 cells, HOS cells, and primary human osteoblasts, as well as in vitro induced osteoclast-like cells. The specificity of the antiserum was verified by peptide blocking and reduction using sequence-specific small interfering RNA. Confocal immunofluorescence cytochemistry localized hsAC primarily in cytoplasm, but some labeling was observed in the nucleus and the plasma membrane. Cytoplasmic hsAC colocalized with microtubules but not with microfilaments. To test the function of hsAC, four constructs containing catalytic domains I and II (aa 1–802), catalytic domain II (aa 231–802), noncatalytic domain (aa 648–1,610), and full-length protein (aa 1–1,610) were expressed in Sf9 insect cells. Only catalytic domains I and II or full-length proteins showed adenylyl cyclase activity. Mg2+, Mn2+, and Ca2+ all increased adenylyl cyclase activity in a dose-dependent manner. While hsAC had a minimal response to HCO3− in the absence of divalent cations, HCO3− robustly stimulated Mg2+-bound hsAC but inhibited Mn2+-bound hsAC in a dose-dependent manner. In summary, hsAC is a divalent cation and HCO3− sensor, and its HCO3− sensitivity is modulated by divalent cations.
- bicarbonate sensor
- calcium homeostasis
using genome-wide linkage analysis of three kindreds, our group mapped a locus and cloned a novel gene linked to absorptive hypercalciuria (AH), a syndrome of intestinal Ca2+ hyperabsorption, hypercalciuric Ca2+ nephrolithiasis, and low bone mineral density (26, 27). The predicted amino acid sequence of one gene in this locus is homologous to rat soluble adenylyl cyclase (sAC) (4). The human ortholog of sAC (hsAC) has 33 exons and spans 104 kb, and the full-length cDNA of the intestine has 5,085 nucleotides (GenBank accession no. AF331033). The polypeptide sequence predicts two adenylyl cyclase catalytic sites. hsAC is highly polymorphic, with at least 17 sites of allelic sequence variation. Although some of these sequence variations can be found in apparently healthy individuals, seven of them occur with higher frequency in patients with AH than in healthy individuals and can potentially be pathogenic mutations. The functional significance of these base changes is not yet defined. We first proceeded to characterize the wild-type hsAC protein.
cAMP is a second messenger that transduces signals to intracellular effectors. Classically, cAMP comes from G protein-coupled transmembrane adenylyl cyclase (tmAC) (22, 33–35). Buck and coworkers (4, 9, 23, 37) purified, cloned, and characterized sAC from rat testis. sAC is distinct from the traditional G protein-regulated tmACs. sAC is soluble and is located in the cytoplasm and the intracellular organelles (40), enabling cAMP to be generated directly inside the cell and compartmentalized in the vicinity of its targets (4, 39). sAC does not respond to the heterotrimeric G protein regulators and forskolin, but rather is stimulated by HCO3− ion. It is considered a HCO3− sensor in the rat and plays the essential role in mammalian sperm biology (9, 30, 37, 41). Homozygous sAC-deficient mice are infertile because of a severe sperm motility defect, but this phenotype can be rescued by provision of cell-permeant cAMP (12). Besides the testis and sperm cells, sAC is also found in the kidney, the choroid plexus (9), and multiple cell lines (39). The broad expression pattern of sAC suggests that sAC may subserve a multitude of functions throughout the body, including Ca2+ homeostasis. The human ortholog has not been characterized extensively to date.
MATERIALS AND METHODS
Human embryonic kidney (HEK)-293 cells derived from human kidney, Caco-2 cells derived from human colorectal adenocarcinoma, HOS cells derived from human osteogenic sarcoma, NRK cells derived from normal rat kidney, OKP cells derived from opossum kidney, LLCPK cells derived from pig kidney, and RAW264.7 cells derived from mouse monocytes were obtained from the American Type Culture Collection (Manassas, VA). Cells were seeded onto 100-mm culture dishes with in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37°C in humidified 5% CO2 in air. Culture medium was replaced every 48 h. Human primary osteoblasts were purchased from Cambrex Bioproducts (Walkersville, MD) and cultured in the osteoblast growth medium (Cambrex Bioproducts). Osteoclast-like cells were induced from murine monocytic RAW264.7 cells with 50 ng/ml receptor activator of NF-κβ ligand (RANKL) and 5 ng/ml monocyte colony-stimulating factor (M-CSF) (15, 28, 29, 31, 38). Osteoclasts were fixed and stained for tartrate-resistant acid phosphatase (Kamiya Biomedical, Seattle, WA) after 7–10 days as a marker for differentiation. To examine functional resorptive activity, induced osteoclasts were plated on culture plates coated with apatite [Ca10(PO4)6(OH)2·6H2O; Kamiya Biomedical], and resorption pits were detected by staining the apatite with silver nitrate (Sigma, St. Louis, MO). The negatively stained resorption pits were visualized under a digitized microscope.
Cloning of hsAC and hsAC mRNA analysis.
hsAC cDNA was PCR amplified from a human testis marathon ready cDNA library (Clontech, Palo Alto, CA) using forward primer 5′-GAACATGAACACTCCAAAAGAAGAAT-3′ and reverse primer 5′-GAAATGATTGTCCACGGTATTAGC-3′ derived from human. The PCR product was TA cloned (Invitrogen, Carlsbad, CA) and sequenced. The putative amino acid sequence of hsAC was analyzed for functional domains using the SMART program (http://smart.embl-heidelberg.de/). hsAC mRNA was analyzed using RT-PCR. Primers spanning exons 3–6 of hsAC were as follows: forward, 5′-CAGCAGTGCCATGTACATGGA-3′, and reverse, 5′-AGCTGCCAGCAGTTTGGTGA-3′. For HOS and Caco-2 cDNA, total RNA was isolated from HOS and Caco-2 cells using TRIzol reagent (Sigma), and RT-PCR was performed using the Gold RNA PCR kit (PE Biosystems, Foster City, CA) and Expand High Fidelity PCR system (Roche Diagnostics, Indianapolis, IN). PCR products were TA cloned and sequenced. Human testis, kidney, and small intestine cDNA libraries were purchased from Clontech.
RNA blots were performed using a single-stranded riboprobe hybridized against a multiple tissue human polyA+ RNA blot (Clontech) according to the manufacturer's recommendations. Riboprobes were used because uniformly [32P]dCTP-labeled (specific activity >108 cpm/μg), double-stranded hsAC cDNA probes failed to yield satisfactory hybridization signals. To generate the cRNA probe, hsAC (nt 181–2,161) was subcloned into pBluescript SK(−) (Stratagene, La Jolla, CA), sequenced, and used as a template. [32P]UTP-labeled (PerkinElmer, Wellesley, MA) antisense hsAC cRNA was synthesized using Strip-EZ RNA strippable probe synthesis kit (Ambion, Austin, TX). The filters were stripped and reprobed with β-actin for internal control of loading.
A peptide (SLSEGDALLA) corresponding to the extreme NH2 terminus of hsAC (variant V1) was selected based on its antigenicity and lack of homology with other protein sequences using a BLAST homology search. The location of the immunogenic peptide is shown in Fig. 1D. Rabbits were immunized with this peptide to generate anti-hsAC serum (Invitrogen). For immunoblots, cells were washed twice with ice-cold phosphate-buffered saline (PBS), and we collected them by scraping them from the culture wells. Harvested cells were suspended in 1 ml of lysis buffer (10 mmol/l Tris, pH 7.4, 1 mmol/l EDTA, 1% glycerol, 1 μg/ml leupeptin, 1 μg/ml pepstatin, 50 μg/ml aprotinin, and 0.5 mmol/l phenylmethylsulfonyl fluoride) on ice and disrupted by two 10-s bursts from a Tekmar sonic disrupter (Fisher Scientific, Hampton, NH) at power setting 60. Protein concentrations were determined using a bicinchoninic acid protein assay (Pierce, Rockford, IL). Equal amounts of sample proteins were electrophoretically separated on 10% sodium dodecyl sulfate-polyacrylamide gel. Proteins were electrically transferred to nitrocellulose membrane and the membrane was incubated for 1 h in blocking solution (10% nonfat dry milk in PBS, pH 7.4). Subsequently, anti-hsAC serum were added for 3 h, after which the blot was washed several times in PBS containing Tween 20 and then incubated 1 h with a 1:5,000 dilution of goat anti-rabbit IgG horseradish peroxidase-conjugated secondary antibody (Amersham, Piscataway, NJ) in PBS containing 5% nonfat dry milk. Reactive bands were visualized using enhanced chemiluminescence (Amersham). For competitive immunoblotting experiments, 2 μl of anti-hsAC serum was incubated with 670 μg of either the immunogenic peptide or BSA overnight at 4°C before immunoblotting.
Crude membrane and cytosolic fractions were prepared from multiple cell lines. Cells were scraped and disrupted by performing sonication twice for 10 s each time in 50 mM Tris·HCl (pH 7.2) and 150 mM NaCl (pH 7.2) with complete protease inhibitor cocktail (Roche Diagnostics), followed by centrifugation at 1,000 g for 10 min to remove nuclei and debris. Cell lysates were centrifuged at 100,000 g for 1 h, and the crude membrane pellet was suspended by sonicating it two times for 2 s each time in lysis buffer.
siRNA for knockdown hsAC.
The 21-nucleotide small interfering RNA (siRNA) targeting hsAC was custom synthesized (Ambion, Austin TX) to obtain the following sequences: sense, 5′-AUGUAGCCUGGAGAUCCAUUU-3′, and antisense, 5′-AUGGAUCUCCAGGCUACAUUU-3′. Oligonucleotides were annealed according to the manufacturer's instructions. Approximately 24 h before transfection, HOS cells were plated in six-well culture plates at an appropriate cell density so that they were ∼70% confluent the next day. For the complex formation and transfection, 6 μl of the siPORT Amine transfection reagent (Ambion, Austin, TX) were mixed with 200 μl of serum-free medium and incubated at room temperature for 20 min, and then siRNA (100 nM) was added to the complex and incubated at room temperature for an additional 20 min. The mixture was applied to HOS cells, and cells were incubated with the mixture for 5 h and then switched to regular cell culture medium (DMEM with 10% FBS). Cell culture medium was replaced every 2 days, and HOS cells were harvested every 24 h up to 5 days for immunoblot analysis.
hsAC confocal immunofluorescence microscopy.
Human primary osteoblasts were grown to 20% confluence on glass-bottomed culture dishes. Cells were fixed with 3.7% paraformaldehyde for 10 min at room temperature, quenched with 100 mM glycine for 5 min at room temperature, permeabilized in 0.1–0.3% Triton X-100 for 5 min on ice, blocked with PBS-10% milk for 30 min at room temperature, incubated in anti-hsAC for 4°C overnight, and then incubated in FITC-labeled goat anti-rabbit IgG (Molecular Probes, Eugene, OR) for 30 min at room temperature. For colocalization with microfilaments, total actin was stained with rhodamine-phalloidin (Sigma) for 30 min. For colocalization with microtubules, cells were incubated in anti-β-tubulin for 4°C overnight and then in Alexa Fluor 633 F(ab′)2 fragment of goat anti-mouse IgG (Molecular Probes) for 30 min at room temperature. Confocal immunofluorescence microscopy was performed with a Zeiss LSM 410 confocal microscope using a krypton/argon laser. Images were analyzed using LSM 5 Image Browser software.
Expression of hsAC in Sf9 cells.
Human sAC was expressed using the Bac-to-Bac Baculovirus Expression System in Spodoptera frugiperda (Sf9) cells (Invitrogen). Briefly, either full-length or the indicated putative functional domains were subcloned into pFastBac HT plasmid (Invitrogen). This involved PCR amplification of specific domains using Pfu DNA polymerase, followed by double digestion of plasmid and PCR fragments with EheI and SalI and ligation with T4 DNA ligase. All constructs were partially sequenced in both directions to confirm that the inserts were in frame with the His epitope. Sf9 cells grown at a density of ∼2 × 106 cells/ml were infected with baculovirus. After infection, Sf9 cells were collected every 24 h for up to 5 days. Recombinant hsAC was confirmed on the basis of immunoblotting, confocal immunocytochemistry, and enzymatic activity. For recombinant hsAC expression in Sf9 cells, actin was stained with Oregon Green 488 phalloidin (Molecular Probes) and hsAC was detected using anti-hsAC serum.
Adenylyl cyclase assay.
Adenylyl cyclase activity was measured as described by Alvarez and Daniels (1, 2). An assay was performed in 200 μl of total reaction volume using ∼800 μg of Sf9 cell lysate in the presence of 50 mM Tris·HCl, pH 7.5, substrate [α-32P]ATP (0.1 μCi/assay), 2 mM cAMP, 0.1 mM GTP, 20 mM phospho(enol)pyruvate, 3 U of pyruvate kinase, and 2 mM ATP. The appropriate concentrations of MgCl2, MnCl2, and/or HCO3− were added. The assay was started by the addition of cell lysate, and the mix was incubated for 15 min at 37°C and stopped by the addition of 2.2 M HCl containing [3H]cAMP, heating the solution to 95°C for 4 min, and then chilling it on ice. [32P]cAMP generated by the reaction and [3H]cAMP were recovered with the use of an acidic alumina column (ICN Biomedical, Irvine, CA). [3H]cAMP and [32P]cAMP content were determined by performing liquid scintillation spectrometry using dual isotope windows for [3H] and [32P]. [3H]cAMP was used to calculate the efficiency of cAMP recovery. To assess the dose-response relationship, hsAC activity was assayed as a function of varying Mg2+, Mn2+, and Ca2+ concentrations. For HCO3− sensitivity, hsAC was assayed as a function of varying HCO3− in the presence of either 5 mM MgCl2 or 5 mM MnCl2. Curve fitting was performed using SigmaPlot software (SPSS, Chicago, IL).
With regard to the HCO3− dose-response relationship in the presence of Mn2+, one needs to consider that the addition of HCO3− might decrease free ionic Mn2+ concentration in the assay solution because of the formation of MnHCO complex and subsequent lowering of the stimulatory effect of Mn2+. To compensate for this complication, we increased the total Mn2+ concentration in the assay medium as we increased the HCO3− concentration so that free ionic Mn2+ stayed constant. We performed separate experiments and empirically measured the association constant of MnHCO in vitro on the basis of the reduction of free HCO3− by increasing Mn2+ at three pH levels (pH 7.1–7.7) and obtained a value of 5 × 10−3. We used this value to clamp the free Mn2+ constant while varying HCO3−.
Ionic Ca2+ in the cyclase assay reaction was determined using Nova 8 (Nova Biomedical, Waltham, MA) for the range from 0.2 to 6 mM. For the lower range of 0–1,000 nM, ionic Ca2+ was measured fluorometrically with Photomultiplier Detection System 814 (Photon Technology International, Lawrenceville, NJ) using 10 μM fura-FF (Molecular Probes) as the fluorescence indicator (λexcitation, 340/380 nm; λemission, 510 nm) (14). After we determined ionized Ca2+ in the cell lysate using fura-FF ratiometric fluorometry, we manipulated Ca2+ in either direction by adding either EGTA or CaCl2. The final ionic Ca2+ concentration in each reaction was determined for each sample.
Data are presented as means ± SE. Comparisons were performed using a paired Student's t-test. P < 0.05 was considered statistically significant in all analyses.
hsAC transcripts in human tissues and cell lines.
Using a RNA probe corresponding to the hsAC nucleotide 181–2,161 (aa 1–659) against a human multiple tissue blot containing 1 μg of polyA+ RNA, we detected a predominant 5.1-kb transcript in brain, heart, skeletal muscle, thymus, kidney, liver, placenta, lung, and peripheral blood leukocyte (Fig. 1A). A faint doublet was discernible in the kidney, liver, and heart, which may suggest multiple transcripts. Although transcripts appear to be ubiquitous using RT-PCR, labeling on RNA blots for small intestine and colon were visible only after prolonged exposure. Different regions of the hsAC cDNA were examined using nested primers. With the use of a pair of primers spanning exons 3–6 of hsAC, PCR reactions revealed three bands in commercially purchased cDNA from normal human tissue, testis, kidney, and small intestine, and in cDNA we performed reverse transcription from HOS and Caco-2 cells (Fig. 1B). PCR products from each tissue or cell were subcloned and sequenced for each clone (Fig. 1C). Sequence analysis revealed that the top band had an extra 37 5′-nucleotides to exon 5. The bottom band had no exon 5 sequence at all (Fig. 1D). On the basis of the primary cDNA sequence, the presence of an extra 37 nucleotides before exon 5 and the omission of the entire exon 5 both result in new start codons in the hsAC sequence. We arbitrarily named these three variants V1, V2, and V3 (Fig. 1, B and C). None of these splice variants led to frame shift beyond exon 6, giving rise to identical predicted COOH termini. In V1, the extra 37 nucleotides had an extra C (indicated by asterisks in Fig. 1D) that resulted in a new ATG start codon. No splice variations were detected in the remaining exons of hsAC.
hsAC expression and cellular distribution.
We next analyzed the expression of the hsAC protein. The specificity of the anti-hsAC serum was authenticated on the basis of competitive inhibition by the immunogenic peptide and siRNA knockdown (Fig. 2). No HEK-293 and Caco-2 cell signals were detected with competitive inhibition by preincubating the immune serum with the immunogenic peptide (Fig. 2A). The introduction of 100 nM hsAC siRNA caused ∼73% decrease in hsAC protein expression in HOS cells by day 5 after transfection (Fig. 2B).
Immunoblot analysis of kidney-derived (HEK-293, OKP, NRK, and LLCPK1), gut-derived (Caco-2), and bone-derived (primary osteoblast, HOS) cells revealed two bands at ∼190 kDa and ∼80 kDa (Figs. 2 and 3), with the 80-kDa band being dominant. Both the 190- and 80 kDa-bands were present in the pellet and the supernatant after 100,000 g centrifugation of HOS, Caco-2, and HEK-293 cells. The 190-kDa band signal was relatively weak in HOS and HEK-293 cells. Immunoblot analysis of kidney cell lines from pig, rat, and opossum revealed a diverse expression pattern (Fig. 3C). In LLCPK cells, a 135-kDa band was detected in the cytosol and an 80-kDa band was observed in the membrane fraction. In NRK cells, a 128-kDa band was detected in the membrane fraction and a 70-kDa band was detected in both the cytosol and the membrane fraction. In OKP cells, 190- and 85-kDa bands were detected only in the cytosol fraction.
To further define the intracellular distribution of hsAC, we performed immunofluorescence microscopy of primary cultured human osteoblasts (Fig. 4A) and in vitro induced osteoclast-like cells (Fig. 4B). In human osteoblasts, hsAC was located primarily within the cells. About 30% of the cells demonstrated strong fluorescence at the plasma membrane. About 50% of the cells showed fluorescence inside the nucleus. No signal was detected in preimmune serum or in the IgG control. All staining was reduced >80% by the immunogenic peptide. Inside the cell, hsAC exhibited a fibrillar distribution. Colocalization with actin or tubulin showed that these fibrils were associated not with the microfilament but rather with the microtubules (Fig. 4B). Osteoclast-like cells were induced from RAW264.7 cells in the presence of RANKL and M-CSF. Multinuclear cells typically form from days 5 to 7 after induction and are functionally absorptive as evidenced by resorptive pit formation on the apatite-coated culture plates. hsAC was localized in the cytosol in osteoclasts and highly concentrated in the perinuclear region (Fig. 4C).
Four expression plasmids were constructed to produce recombinant NH2-terminal hexahistidine-tagged hsAC in Sf9 cells. Each construct contained either the full-length hsAC or various predicted functional domains of cyclase (Fig. 5). hsAC1–802 spans catalytic domains I and II. hsAC231–802 includes only catalytic domain II. hsAC648–1610 includes the entire COOH terminus lacking both catalytic domains. hsAC1–1610 is the full-length hsAC polypeptide. After plasmid transfection, Sf9 cells expressed recombinant hsAC on day 2, peaked on day 3 or 4, and then declined on day 5 (data not shown). A similar expression time course also was observed in immunoblot analysis using anti-hexahistidine antibodies (data not shown). The various constructs had different distribution patterns in Sf9 cells. The full-length and noncatalytic domains were located close to the plasma membrane and exhibited more of a “ring” pattern. In addition to the ring pattern, the second catalytic or both catalytic domains were located inside the cytosol. This suggests that in Sf9 cells, hsAC is largely a membrane-associated protein. The noncatalytic COOH terminus may harbor the sequence for membrane association, whereas the catalytic domain is distributed more abundantly in the cytoplasm.
hsAC: regulation by divalent cations and HCO3−.
We next examined adenylyl cyclase activity of recombinant hsAC in Sf9 cells. Whole Sf9 cell lysate from day 3 postinfection was used because of the higher hsAC antigenic expression. Compared with Sf9 cells alone, both full-length hsAC1–1610 and the double-catalytic domain hsAC1–802 constructs demonstrated adenylyl cyclase activity (Fig. 6). The addition of Mg2+ or Mn2+ significantly increased cyclase activity 2.7- or 78.5-fold, respectively (Fig. 6). No significant cyclase activity was detected in the second catalytic domain (hsAC231–802) or in the noncatalytic domain of hsAC (hsAC648–1610). This indicates that both catalytic domains are needed for cyclase activity. In all four constructs in the absence of the divalent cation, HCO3− had no effect on adenylyl cyclase activity. However, in the presence of Mg2+ and Mn2+, the enzyme was sensitive to HCO3−. HCO3− significantly increased the Mg2+-bound cyclase activity 7.2-fold for the catalytic domains (hsAC1–802) and 6.6-fold for full-length hsAC (hsAC1–1610) (Fig. 6). In contrast, HCO3− significantly decreased the Mn2+-bound cyclase activity 2.4- or 2.1-fold for the catalytic or full-length hsAC, respectively. In all of the assay conditions, no statistically significant difference was found for the full-length and catalytic domains of hsAC. It is noteworthy that there was some endogenous adenylyl cyclase activity in Sf9 cells that could be stimulated by both Mg2+ and Mn2+ (Fig. 6).
Next, we investigated the dose-response relationship of the Mg2+, Mn2+, and HCO3− (Fig. 7). Only the catalytic domain (hsAC1–802) and the full-length (hsAC1–1610) constructs were used in these experiments. No statistically significant differences in the dose responses were observed between the full-length (hsAC1–1610) and catalytic domains (hsAC1–802) (data not show) of hsAC. Both Mg2+ and Mn2+ increased cyclase activity in a dose-dependent manner (Fig. 7). Mg2+ stimulation fitted well with first-order kinetics, with Km of 2.41 mM (Fig. 7A). Mn2+ also rapidly increased cyclase activity from 1 to 5 mM, peaking at 20 mM but with a 20-fold higher Vmax. The Mn2+ dose-response curve displayed a sigmoid shape (Fig. 7B), with half-maximal activation at 3.9 mM and a Hill coefficient of 2.37.
As expected on the basis of the data shown in Fig. 6, the HCO3− dose response was opposite, depending on the ambient divalent cation for both the catalytic and full-length hsAC. HCO3− robustly increased the Mg2+-bound adenylyl cyclase (Fig. 7C) but inhibited the Mn2+-bound adenylyl cyclase (Fig. 7D). The half-maximal effect for both the stimulation and inhibition are shown at comparable HCO3− concentrations (8.7 mM for Mg2+ vs. 3.9 mM for Mn2+). Because HCO3− addition may complex Mn2+ and decrease ionic Mn2+ stimulation, we conducted the experiment shown in Fig. 7D with increasing total Mn2+ as we increased the HCO3− clamping free Mn2+ constant.
In addition to Mg2+ and Mn2+, we studied the effect of Ca2+ on cyclase activity. For catalytic domain and full-length hsAC in the absence of Mg2+ and HCO3−, Ca2+ weakly stimulates cyclase activity (data not shown). To better visualize the dose response of Ca2+, we stimulated the basal level of cyclase activity by adding a small amount of Mg2+ and HCO3− in the Ca2+ study as previously described by other investigators (17, 23). With the presence of 2 mM Mg2+ and 20 mM HCO3−, Ca2+ stimulated cyclase activity in a dose-dependent manner from 0.12 to 1.31 mM (Fig. 8A). At more physiological cytosolic concentrations from 2 to 1,200 nM ionized Ca2+, cyclase activity also increased in a dose-dependent manner (Fig. 8B). No significant differences between the catalytic and full length hsAC were observed at either higher or physiological concentrations of ionic Ca2+ (Fig. 8, A and B).
HCO3−-regulated processes are pervasive in biology, and sAC may be the link between ambient HCO3− concentration and known cell signaling molecules. Because of the importance of HCO3− concentration in osteoblast and osteoclast function (3, 6, 7), and because sequence variation in hsAC is associated with lower bone density and increased intestinal Ca2+ absorption (24, 26), we were interested in the role of hsAC in human Ca2+ homeostasis.
In this study, we have demonstrated that hsAC is a functional adenylyl cyclase that responds to divalent ions and HCO3− ion regulation. Transcripts are ubiquitous as shown by RT-PCR and RNA blots of organs of Ca2+ homeostasis. hsAC is of low abundance in human tissue and is detectable only with a single-stranded riboprobe on a 1-μg polyA+ RNA gel. The signal in the colon and small bowel is visible only after extremely long exposure. In contrast to its rodent ortholog, hsAC appears to be more complex in that it has multiple transcripts. While variant transcripts of such similar lengths are difficult to discern with the resolution of gel electrophoresis, they are reproducibly detected in multiple tissues and human cell lines with the use of RT-PCR. On the basis of prediction from the primary sequence, these transcripts should translate to three polypeptides with three different NH2 termini. However, the predicted mobility difference is not expected to be detectable using SDS-PAGE. These splice variations do not give rise to frame shifts, and we do not find any splice variation in other regions of hsAC. This method will predict identical COOH termini for the three variants. At present, it is unknown whether these variant transcripts actually give rise to three distinct polypeptides.
A specific anti-hsAC antiserum labeled hsAC antigen in HEK-293, Caco-2, HOS, and primary cultured human osteoblast cells. Two bands of ∼190- and ∼80-kDa mass are detected in human kidney, small intestine, and bone cell lines, with the 80-kDa band being the predominant one. If the transcript variants do give rise to different polypeptides, because of the locale of the epitope, one would expect this antiserum to label definitively the V1 variant and perhaps the V2 variant (Fig. 1D). The relationship between the three transcripts and protein bands is not clear. The ∼190-kDa band is compatible with the full-length translation open-reading frame. The ∼80-kDa band may be a posttranslationally processed protein containing the NH2-terminal epitope. Multiple putative proteolytic cleavage sites are predicted from the primary amino acid sequence of hsAC, the rodent sAC is thought to be processed by cleavage, and the cleaved COOH terminus has been postulated to be an autoinhibitory domain (18).
hsAC is detected in both cytosolic and membrane fractions of the cell lysate after 100,000 g centrifugation. The relative distribution differs slightly from one cell line to another, but the presence in both pellet and supernatant is consistent, except in OKP cells, in which very little or no hsAC antigen is present in the membrane pellet. Although the primary sequence predicts possible putative transmembrane regions, native hsAC is unlikely to be a true transmembrane protein, because it can be released easily into the supernatant by weak nonionic detergents. In immunocytochemistry, hsAC has different distribution in different cells. We examined native hsAC in two human bone cells: osteoblasts and osteoclasts. Native hsAC in primary cultured human osteoblasts shows intracellular, plasma membrane, and nuclear localization patterns. The intracellular pattern is definitely fibrillar, and by colocalization, the enzyme is associated with microfibrils rather than microfilaments. hsAC is present in osteoclasts as well as in its uninduced progenitor RAW cells. The pattern is distinct from osteoblasts in that it has a more diffuse nonfibrillar intracellular staining with accentuation in the perinuclear region. We next examined the expression pattern of recombinant hsAC consisting of different regions of the protein in Sf9 cells. Intracellular staining was evident in all constructs. The full-length (hsAC1–1610) and noncatalytic domain (hsAC648–1610) proteins demonstrated a ring pattern denoting plasma membrane association. hsAC harboring catalytic domains I and II (hsAC1–802) or catalytic domain II alone (hsAC231–802) showed primarily intracellular staining in Sf9 cells. These results suggest that there may be signals between aa 802 and aa 1,610 that confer membrane association.
A series of seminal papers demonstrated that the rat sAC is a soluble adenylyl cyclase that responds to HCO3− stimulation and thus is considered the HCO3− sensor (9, 17, 30, 37, 41). hsAC is also an adenylyl cyclase, but it is distinct from the rat ortholog. When expressed in Sf9 cells, full-length hsAC as well as hsAC containing the catalytic domains yielded adenylyl cyclase activity. We specifically addressed the second catalytic domain because the putative protein translated from the V3 transcript should lack the first catalytic domain. No cyclase activity was evident in the second catalytic domain or in the noncatalytic domain of hsAC. This suggests that both catalytic domains are necessary for a functional adenylyl cyclase. If the V3 transcript indeed translates to polypeptide in native tissue, it may perform functions other than those of adenylyl cyclase. No difference is found between the double-catalytic domain (hsAC1–802) and full-length hsAC (hsAC1–1610) when expressed in insect cells. This finding appears to be distinct from the rat sAC, in which the catalytic domain has 10 times the cyclase activity of the full-length domain. It is most likely that the catalytic domain of hsAC is proteolytically cleaved in Sf9 cells to a form that resembles hsAC1–802. One cannot detect a 180-kDa band in immunoblots of Sf9 lysates from cells infected with the catalytic domain (hsAC1–802) or the full-length hsAC (hsAC1–1610). This complete processing of the full-length protein may explain the lack of difference in cyclase activity between hsAC1–802 and hsAC1–1610.
We have shown that three divalent cations activate hsAC, with Mn2+ being the most potent, followed by Mg2+ and Ca2+. This finding was previously described for rat sAC (16, 17, 23). The finding that only the catalytic region of hsAC is sufficient for these interactions has been described in the rat protein (36). The range of concentrations tested for Mg2+ and Ca2+ are both within the expected normal intracellular range, so hsACs could very well be Mg2+ and Ca2+ sensors. It is interesting that Mg2+ and Mn2+ have different kinetics, suggesting that they may bind to different sites on the protein. Although Mn2+ has been shown to activate the rat sAC, no kinetic studies have been performed (23). Although the overall homology of the protein is low, catalytic domains I and II of hsAC resemble the catalytic site of the tmAC. The mechanism of generating cAMP by tmAC has been studied extensively (10, 11, 19, 36). The enzymatic core (catalytic domains I and II) uses a two-metal-ion catalytic mechanism in which, upon metal binding, the enzyme undergoes a conformational transformation. It is proposed that Mg2+ and Mn2+ bind to two distinct sites on tmAC (36); both of which lead to activation of the enzyme.
In addition to HCO3− sensitivity in the recombinant state, the role of sAC as a HCO3− sensor previously was elegantly shown in intact mammalian cells (25). Our present study has shown that the cyclase activity of hsAC is also HCO3− sensitive. However, HCO3− has divergent effects on hsAC, depending on the identity of the divalent cation. In the presence of 5 mM Mg2+ (∼Km of Mg2+), HCO3− stimulated adenylyl cyclase potently, >20-fold, with a Km of 7.8 mM. This clearly enables hsAC to function as a HCO3− sensor in the intact cell. Surprisingly, in the presence of 5 mM Mn2+, HCO3− inhibited cyclase activity with a virtually identical Km (3.9 mM). This effect is not due to complexing of Mn2+ by HCO3−, because we clamped the free Mn2+ concentration constant. The physiological significance of this finding is unclear, because millimolar quantities of Mn2+ are not usually encountered in the cytoplasm. However, this finding may uncover interesting mechanisms of HCO3− interaction with the protein. In the tmAC, Zn2+ and Mg2+ appear to bind to identical sites (36), but while Mg2+ stimulates the enzyme, Zn2+ inhibits it. The interaction between HCO3− and hsAC may be quite different, depending on whether hsAC binds to Mg2+ or Mn2+. Cann et al. (8) demonstrated that a point mutation in the active site (Lys646) of a class III adenylyl cyclase reduced catalytic activity by 95% and also reduced HCO3− activation, and they proposed that the HCO3− binding site might be situated in the vicinity of that residue.
Many biological processes are modulated by ambient HCO3− concentrations. In bone, high HCO3− and alkaline pH inhibit bone resorption, and low HCO3− and acidic pH stimulate it (5, 6). Low HCO3− stimulates and high HCO3− inhibits osteoclastic activity (3, 13, 20, 21). Metabolic alkalosis stimulates osteoblasts (6). It is thought that systemic pH and HCO3− alter bone mineral content by direct physicochemical means or biologically via osteoblasts and osteoclasts, but the signaling pathway between HCO3− and bone cells remains unclear. In the kidney, there are multiple HCO3−-sensitive processes. One important one is the stimulation of Ca2+ absorption in the distal nephron by luminal HCO3− (32). hsAC is expressed in this part of the nephron and can potentially be the mediator of this response. Although the precise function of hsAC in bone and kidney remains to be determined, the clinical association of hsAC to kidney stones and low bone density (26, 27) and the presence of hsAC in both kidney and bone cells suggest that hsAC may provide a link for HCO3− sensing.
In conclusion, hsAC is a human ortholog of the rat sAC that is expressed in multiple tissues and cells involved in Ca2+ homeostasis. hsAC has a multiply spliced transcript with an unknown biological significance at present but has potential roles in mediating HCO3−-regulated cell function in bone and kidney cells. The hsAC protein exists freely as a cytosol as well as in membrane-associated forms. The cyclase catalytic activity as well as divalent ion and HCO3− sensitivity are fully preserved in the catalytic NH2-terminal half of the protein.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants R01 DK-48481 and P01 DK-20543 and by the Department of Veterans Affairs Research Service. W. Geng is partially supported by seed funds from the Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center at Dallas.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society