The sarco(endo)plasmic reticulum Ca2+-ATPases (SERCAs) play a crucial role in regulating free cytosolic Ca2+ concentration in diverse cell types. It has been shown that recombinant SERCA3, when measured in heterologous systems, exhibits low apparent affinity for Ca2+; however, Ca2+ affinity of native SERCA3 in an endogenous setting has not been examined. Such a measurement is complicated, because SERCA3 is always coexpressed with the housekeeping isoform SERCA2b. We used a fluorescence-based assay for monitoring continuous Ca2+ uptake into microsomes to examine the properties of endogenous human SERCA3 and SERCA2b. The kinetic parameters were derived using a cooperative two-component uptake model for Ca2+ activation, and the values assigned to SERCA3 were confirmed using the highly specific human SERCA3 inhibitory antibody PL/IM430. First, using recombinant human SERCA3 and SERCA2b proteins transiently expressed in HEK-293 cells, we confirmed the previously observed low apparent Ca2+ affinity for SERCA3 compared with SERCA2b (1.10 ± 0.04 vs. 0.26 ± 0.01 μM), and using mixtures of recombinant protein isoforms, we validated the two-component uptake model. Then we determined apparent Ca2+ affinity for SERCA proteins present endogenously in cultured Jurkat T lymphocytes and freshly isolated human tonsil lymphocytes. The apparent Ca2+ affinity in these two preparations was 1.04 ± 0.07 and 1.1 ± 0.2 μM for SERCA3 and 0.27 ± 0.02 and 0.26 ± 0.01 μM for SERCA2b, respectively. Our data demonstrate, for the first time, that affinity for Ca2+ is inherently lower for SERCA3 expressed in situ than for other SERCA isoforms.
- calcium pump
- calcium uptake
- continuous fura 4F fluorescence assay
the spatial and temporal control of intracellular Ca2+ homeostasis is crucial for regulating diverse Ca2+-dependent physiological events. The sarco(endo)plasmic reticulum Ca2+-ATPases (SERCAs) play a critical role in regulating the cytosolic free Ca2+ concentration ([Ca2+]), as well as the levels of Ca2+ in the endoplasmic reticulum stores required for controlling complex dynamics of Ca2+ signals. SERCA proteins are encoded by a family of structurally related and alternatively spliced transcripts from three distinct genes (SERCA1, SERCA2, and SERCA3) expressed in a tissue-dependent and/or developmentally regulated manner. Alternative splicing of the ATP2A1 gene generates transcripts encoding SERCA1a and SERCA1b isoforms, which are expressed exclusively in adult and neonatal fast-twitch skeletal muscle, respectively. The ATP2A2 gene encoding the SERCA2 protein is alternatively spliced and processed in a tissue-specific manner, giving rise to three isoforms: SERCA2a, which is expressed predominantly in cardiac/slow-twitch skeletal muscle; SERCA2b, which is expressed ubiquitously; and SERCA2c, which has been reported to have a broader expression pattern than that typically attributed to SERCA2a. The level of SERCA2c protein relative to coexpressed SERCA2a or SERCA2b is not known (13, 19, 35). SERCA2b has a higher apparent affinity for Ca2+ and turns over at a rate twofold slower than SERCA2a. SERCA2c has a lower apparent Ca2+ affinity and a slower turnover rate than SERCA2a. The functional differences are due to replacement of the COOH-terminal sequence AILE (amino acids 994–997) in SERCA2a, by a 49- to 50-amino acid hydrophobic tail in SERCA2b (32, 34), and by the 6 amino acids (i.e., VLSSEL) in SERCA2c (13). SERCA1 and SERCA2a are subject to regulation by phospholamban and sarcolipin, which modulate their activity by decreasing their apparent affinity for Ca2+ (36), so that the values for this parameter determined in the recombinant system do not match those observed in the endogenous setting.
The ATP2A3 gene is alternatively spliced to generate six different transcripts encoding the isoforms SERCA3a, SERCA3b, SERCA3c, SERCA3d, SERCA3e, and SERCA3f, which differ in the length and sequence of their COOH-terminal ends (8). SERCA3 is expressed abundantly in only a select number of adult tissue and cell types, predominantly of nonmuscle origin. More interestingly, SERCA3 is always coexpressed with the ubiquitous isoform SERCA2b. The highest SERCA3 expression is observed in cells of hematopoietic lineage (platelets, mast cells, and lymphocytes), embryologically related endothelial cells, secretory epithelial cells (large and small intestine, thymus, and trachea), and Purkinje neurons in the cerebellum (1, 5, 10, 39, 50). The expression of SERCA3 has also been detected in the mammalian retina and human myocardium (14, 27). In addition, the expression of SERCA3 mRNA during ontogeny in the rat indicates a tissue-specific and developmentally regulated pattern in the cardiovascular system (2). Physiological roles for SERCA3 have been suggested in various systems: studies on SERCA3-deficient mice have indicated that lack of SERCA3 alters endothelium-dependent relaxation in vascular smooth muscle (30) and epithelium-dependent relaxation in tracheal smooth muscle (24). Downregulation of SERCA3 expression is suggested to be responsible for the impaired glucose responses in the islets of Langerhans in diabetic mouse and rat models of non-insulin-dependent diabetes mellitus, a metabolic disease associated with abnormal insulin secretion (43). In humans, missense mutations in the SERCA3 gene are thought to render patients more susceptible to type 2 diabetes mellitus (48).
Comparative functional studies conducted on recombinant SERCA isoforms expressed in heterologous systems indicate three distinct characteristics of SERCA3. 1) Recombinant SERCA3a displayed lower apparent affinity for Ca2+ when assayed for Ca2+-ATPase activity and E-P formation (K0.5 ∼1.1 μM vs. ∼0.3 μM for SERCA2b) than other SERCA isoforms (34). More recently, with use of recombinant proteins expressed heterologously, it was shown that all six SERCA3 isoforms display similarly lower apparent affinity for Ca2+ (8, 37). 2) SERCA3 differed from other isoforms, in that its pH optimum was 7.2–7.4, whereas that for SERCA1 and SERCA2 pumps was 6.8–7.0. 3) Apparent affinity for vanadate inhibition was markedly (10-fold) higher for SERCA3 than for the other SERCA isoforms (34).
A major question that has not been fully answered is the physiological significance of structural and functional variability among the different SERCA isoforms. This is of particular relevance for SERCA3, because it has the highest isoform diversity, is always coexpressed with SERCA2b, and is only expressed in a select number of cell/tissue types. As mentioned above, one of the biochemical hallmarks of recombinant SERCA3 function measured in a heterologous expression system is its unusually low apparent affinity for Ca2+. Such a value would have important implications for understanding how Ca2+ is regulated in those cells where SERCA3 expression is high. However, the properties of SERCA3 have not been evaluated in an endogenous setting, and it has only been assumed that endogenous SERCA3 would display the same properties as recombinantly expressed protein. Therefore, in this study, we undertook the task of deriving reliable kinetic parameters from systems that express SERCA3 endogenously, a task that is challenging, because SERCA3 is always coexpressed with the ubiquitous isoform SERCA2b. With use of cultured Jurkat T lymphocytes and freshly isolated human tonsil lymphocytes, it was determined that the endogenous SERCA3 in each of these cell types indeed displayed low apparent affinity for Ca2+, indicating, for the first time, that the distinct properties are inherent in the SERCA3 protein and not likely the result of regulation by a modulatory protein.
MATERIALS AND METHODS
All molecular procedures were performed essentially according to standard protocols (4) or the directions of reagent manufacturers, unless indicated otherwise. All chemicals were of the highest-quality analytic grade available and were obtained from BDH, ICN (via Fisher Scientific, Ottawa, ON, Canada), VWR International (Mississauga, ON, Canada), or Sigma-Aldrich (Oakville, ON, Canada), unless indicated otherwise. The fluorescent dye fura 4F and Calcium Calibration Kit No. 2 were purchased from Molecular Probes/Invitrogen (Burlington, ON, Canada); DNA purification kits from Qiagen (Mississauga, ON, Canada); monoclonal anti-SERCA3 antibody PL/IM430 from Research Diagnostics (Fitzgerald Industries, Concord, MA); polyclonal anti-SERCA3 antibody PA1-910 and monoclonal anti-SERCA2b antibody IID8 from Affinity BioReagents (Golden, CO); and Western blot chemiluminescence solutions from Pierce (via VWR International). A clonal derivative of the human embryonic kidney cell line HEK-293 was originally obtained from Ron Kopito (Stanford University) (47). The Jurkat T cell line was obtained from the University of Calgary Hybridoma and Cell Bank. All cell culture media and supplements were purchased from Invitrogen.
Recombinant and Endogenous Expression of SERCA
Recombinant expression of all SERCA isoforms was investigated in HEK-293 cells, as described previously (11). HEK-293 cells were transfected with plasmid DNA encoding human SERCA2b and SERCA3 and harvested 48 h later.
The human T lymphocytic Jurkat cell line was maintained in a humidified atmosphere at 37°C and 5% CO2 in RPMI 1640 medium supplemented with 10% FBS, 4 mM GlutaMAX-I, and 1× antibiotic-antimycotic. Cells were continuously maintained in log phase and were routinely seeded at 2.5 × 105 cells/ml and grown to a density of ∼1 × 106 cells/ml. Cell viability was assayed by trypan blue exclusion and was always >95%.
Human tonsils were obtained from the Alberta Children's Hospital (Calgary, AB, Canada) after tonsillectomy procedures, in accordance with the regulations as approved by the research ethics board of the Faculty of Medicine at the University of Calgary. Tonsils were placed in RPMI 1640 medium after the surgical procedure and used within 2–3 h for lymphocyte isolation, essentially according to published methods (12). Briefly, the surgically removed tonsils were rinsed and gently minced in RPMI 1640 medium to release lymphocytes. Tissue pieces were allowed to settle, and lymphocytes were collected from the supernatant by brief centrifugation to harvest a pellet. Erythrocytes were lysed by resuspension of the pellet in 0.2% NaCl for 20 s followed by addition of NaCl to a final concentration of 1.6%. The resulting lymphocyte cell pellet was rinsed three times in PBS and used for microsome preparation.
Microsomes were prepared essentially as described previously (11, 39). HEK-293 cells from five 10-cm dishes were swollen in hypotonic medium and lysed using a Dounce homogenizer. The resulting cellular homogenate was separated by differential centrifugation to isolate a postmitochondrial particulate fraction enriched in endoplasmic reticulum that is referred to as “microsomes.” Protein concentration was determined using a Coomassie dye-binding assay (Bio-Rad Laboratories, Mississauga, ON, Canada) with bovine γ-globulin as the standard. Microsomes were stored in aliquots at −80°C. The method used to isolate microsomes from Jurkat and human tonsil lymphocytes was essentially as described above, with the following exceptions: ∼2 × 108 Jurkat T cells or the cells isolated from one tonsil were swollen in hypotonic medium as described above, but for a longer duration (20–30 min), and cells were lysed using a Dounce or a Potter-Elvehjem homogenizer and subjected to differential centrifugation, as described above.
Three different SERCA-specific antibodies were used for immunoblot experiments: PL/IM430 monoclonal antibody, which recognizes only human SERCA3 (39); clone IID8 monoclonal antibody, which recognizes human SERCA2a and SERCA2b (9); and PA1-910 polyclonal antibody, which recognizes SERCA3 from various species (39). Proteins were resolved on SDS-PAGE (7.5%) and electrophoretically transferred to 0.2-μm nitrocellulose membranes at 100 V for 1 h in ice-cold transfer buffer.
All immunoblot procedures were conducted at room temperature (23°C). The membrane was blocked for 30 min in 5% (wt/vol) dried skim milk dissolved in a standard PBS solution with 0.1% Tween 20 (PBS-T) and then incubated with the primary antibody in PBS-T for 1 h. The membranes were washed three times in PBS-T for 5 min each, incubated with the appropriate horseradish peroxidase-conjugated secondary antibody for 30 min, and then washed again as described above. The membranes were developed using ECL Super Signal Plus reagent and visualized on Kodak-Max Bio Light film.
Quantification of Relative SERCA2b and SERCA3 Levels by Immunoblotting
The relative levels of SERCA2b and SERCA3 in microsomes prepared from various cell lines were estimated using recombinant SERCA2b and SERCA3 samples of known activity as follows. Samples (30–60 μg) of microsomes isolated from Jurkat, canine spleen, or human tonsil lymphocytes were loaded on SDS-polyacrlyamide gels adjacent to lanes containing varying amounts of microsomes isolated from HEK-293 cells transfected with human SERCA2b or SERCA3. Since small quantities (0.25–1.5 μg) of HEK-293 microsomes were used, 0.5 mg/ml bovine serum albumin was added to these samples as a carrier. Immunoblotting was conducted as described above using PL/IM430 to detect SERCA3 or IID8 to detect SERCA2b. Developed films were scanned and analyzed by densitometry using NIH ImageJ software (http://rsb.info.nih.gov/ij). The intensity of the SERCA band in the lymphocyte microsomes was compared with a standard curve generated with the transfected HEK-293 cell microsomes containing amounts of SERCA chosen to produce signals with levels of intensity that bracket those in the lymphocyte microsomes. The lymphocyte immunoblot signal was converted to equivalent SERCA activity units based on the quantity of SERCA activity in the transfected HEK-293 microsomes, determined previously using the spectrophotometric assay (see below).
Functional Analysis of SERCA Proteins
Spectrophotometric ATP hydrolysis assay.
The rate of ATP hydrolysis was determined using an enzyme-coupled spectrophotometric assay, as described previously (34). In this ATP-regenerative system, hydrolysis of ATP by SERCA is coupled to the oxidation of NADH, which was measured as a decrease in absorption at 340 nm in a spectrophotometer (model DU-640, Beckman). The data from ATP hydrolysis measurements were fit to the general Michaelis-Menten-type cooperative binding model for substrate activation (Eq. 1), where vmax is the maximum activity reached, [Ca2+]free is free [Ca2+], K0.5 is the substrate concentration that gives an ATPase velocity equal to half of vmax, and n is the equivalent of the Hill coefficient (nH). GraphPad Prism software (www.graphpad.com) was used for nonlinear curve fitting. (1)
Fluorescence-based Ca2+ uptake assay.
Ca2+ uptake into microsomes was monitored continuously using a modification of a fura 2-based fluorescence assay (25) to measure the kinetics of SERCA activity in microsomes prepared from HEK-293 cells expressing SERCA2 isoforms (26). The Ca2+-selective fluorescent dye fura 4F was used to monitor ATP-dependent Ca2+ uptake by microsomes prepared from HEK-293 cells and various lymphocyte cell types. The Ca2+ dissociation constant (Kd) for fura 4F was determined using a set of Ca2+ calibration buffers (Calcium Calibration Buffer Kit No. 2, Molecular Probes/Invitrogen). When calibrated under these assay conditions (30°C, MOPS buffer, pH 7.0), Kd determined for fura 4F was 750 nM, which is between the K0.5 values expected for SERCA2b and SERCA3 [200 nM–1 μM, based on recombinant studies using an NADH-coupled enzymatic assay (34)], indicating that it would be a good choice for the measurement of Ca2+ uptake by microsomes expressing SERCA2b and/or SERCA3.
The Ca2+ uptake reactions were investigated in an assay buffer containing 100 mM KCl, 4 mM MgCl2, 20 mM MOPS, 10 mM oxalate, 1.25 mM ATP, 1.25 mM creatine phosphate, and 0.4 U/ml creatine phosphokinase (pH 7.0). Oxalate was included in the buffer to precipitate Ca2+ within the microsomes to maintain a Ca2+ gradient across the membranes, and creatine phosphokinase was included to regenerate ATP. Fura 4F was added outside the microsomal vesicles to a final concentration of 1.7 μM. For microsomes isolated from transfected HEK-293 cells, ∼50–100 μg of microsomal protein were used per reaction, whereas ∼300–400 μg were used for microsomes isolated from Jurkat and human tonsil lymphocytes. The uptake reactions were initiated by the addition of sufficient Ca2+ to the cuvette containing microsomes, reaction buffer, and fura 4F to reach ∼4 μM free Ca2+. The solution in the cuvette was constantly stirred to ensure adequate mixing, and all experiments were conducted at 30°C. Fluorescence was measured using a fluorometer (QuantaMaster DeltaRAM V, Photon Technologies, Lawrenceville, NJ), with the excitation wavelength alternated between 347 and 380 nm and fluorescence emission monitored at 510 nm.
The 347 and 380 nm vs. time curves were corrected for light scatter (by subtraction of the signal measured in the assay buffer containing microsomes alone); then the 347 nm-to-380 nm (347/380) fluorescence ratio (R) was calculated at 1-s intervals. Ca2+ uptake by microsomes results in a decline in the 347/380 ratio with time. As described below, the Ca2+ uptake kinetics can be determined from a single aliquot of microsomes from the 347/380 ratio-time curve. The 347/380 ratio-time curves were smoothed by a moving-window average calculation according to the equation described by Kargacin and Kargacin (25) using 10–20 point window widths.
Free [Ca2+] in the cuvette at each time point was calculated from the 347/380 fluorescence ratio according to Eq. 2 (20). Kd was determined to be 750 nM, and β was 17.89 under these experimental conditions, and the minimum and maximum R values (Rmin and Rmax) were measured for each experiment (2)
For each time point, the extravesicular total Ca2+ concentration ([Ca2+]total) was calculated from free [Ca2+] and binding of Ca2+, Mg2+, and H+ to buffer components ATP, creatine phosphate, oxalate, and fura 4F (25). Ca2+ uptake velocity (μmol·min−1·mg membrane protein−1) is given by Eq. 3 (3) where −d[Ca2+]total/dt (≈ −Δ[Ca2+]total/Δt) is the negative derivative of the decline in extravesicular total [Ca2+] measured over the time interval Δt (in min), vol is the volume of solution in the cuvette, and wt is the amount of microsomal protein in the sample (in mg).
Kinetic parameters were determined from the calculated Ca2+ uptake velocity vs. free [Ca2+] by nonlinear curve fitting with GraphPad Prism software. Data from HEK-293 cell microsomes expressing SERCA3 or SERCA2b were fit to a simple single-component cooperative model (Eq. 1). Data from Jurkat and human tonsil lymphocytes and from mixtures of HEK-293 cell microsomes expressing SERCA3 or SERCA2b were fit using two uptake components (i.e., one due to SERCA2b and one due to SERCA3), each of which represented an independent cooperative Michaelis-Menten model, as given by the following equation (4)
Inhibition of SERCA3 Activity by PL/IM430 Antibody
The SERCA3 inhibitory antibody PL/IM430 (11) was used in this functional assay as a means to distinguish SERCA2b from SERCA3 activity. In experiments where this antibody was used, the microsomes were preincubated with the antibody or an equal volume of antibody buffer for 5 min at room temperature (∼25°C). The Ca2+ uptake activity was measured as described above. SERCA3 and SERCA2b activity in the microsome samples in the absence and presence of the inhibitory antibody PL/IM430 was estimated using Eq. 4. For all the curve fits, nH = 1.8 and 2.0, directly determined from HEK-293 cell microsomes expressing SERCA3 and SERCA2b, respectively, was used, and the K0.5 values determined for Jurkat and tonsil lymphocyte microsomes (see Table 1) in the absence of the antibody were used in the curve fits for data obtained in the presence of the antibody.
HEK-293 Cells Expressing Recombinant SERCA2b and SERCA3
The first objective was to confirm the Ca2+ affinity differences reported in previous studies with use of recombinant human SERCA isoforms expressed in HEK-293 cells. As depicted in Fig. 1, the Ca2+ dependence of ATP hydrolysis activity indicated a fourfold difference in apparent Ca2+ affinity between human SERCA2b and SERCA3: for SERCA2b, K0.5 = 0.26 ± 0.01 μM and nH = 1.75 ± 0.07 (n = 4); for SERCA3, K0.5 = 1.04 ± 0.02 μM and nH = 1.60 ± 0.04 (n = 4). These values are consistent with those obtained previously: K0.5 = 0.27 ± 0.03 μM and nH = 1.7 ± 0.3 for SERCA2b and K0.5 = 1.1 ± 0.1 μM and nH = 1.8 ± 0.2 for SERCA3 (34).
These differences in apparent affinity between SERCA3 and SERCA2b were confirmed using the continuous fluorescence-based assay system for monitoring Ca2+ uptake into microsomes (25). This assay was chosen to determine the kinetic parameters of Ca2+ uptake for cells expressing the two SERCA isoforms endogenously with low combined SERCA activity (3–5 nmol·min−1·mg−1). The continuous fluorescence assay is preferred over 45Ca2+ uptake, because the latter requires multiple samples to generate a single Ca2+ dependence curve, whereas the former requires only one sample for the measurement of a complete uptake curve from which velocity and Ca2+ dependence of uptake can be determined (Fig. 2). Figure 2A, a representative trace of fura 4F fluorescence (347/380) ratio as a function of time, shows a reduction in fluorescence as Ca2+ is taken up into microsomes from HEK-293 cells transfected with human SERCA3. For each time point, the fura 4F ratio is converted to free [Ca2+] and, subsequently, to total [Ca2+] (see materials and methods). Figure 2B shows total [Ca2+] as a function of time derived from the data of Fig. 2A. These data were then used to derive a Ca2+ uptake velocity-free [Ca2+] plot, as illustrated in Fig. 2C for SERCA3 and Fig. 2D for SERCA2b. The data from 10–12 independent experiments with different microsome preparations were then fit to Eq. 1 (see materials and methods), resulting in K0.5 = 1.10 ± 0.04 μM and nH = 1.81 ± 0.04 for SERCA3 and K0.5 = 0.26 ± 0.01 μM and nH = 2.00 ± 0.05 for SERCA2b. Plots of the residuals (Fig. 2, C and D, insets) obtained from the difference between the curve fits and the actual data for each SERCA isoform demonstrate that the data were well fit at all free [Ca2+] values. These values are in close agreement with those determined previously (34) and with those obtained in the present study using the ATP hydrolysis assay (see above). In both of these functional assays, the rate of ATP hydrolysis and the reduction in free [Ca2+] were inhibited >99% by thapsigargin (11, 33), demonstrating that these measurements correspond to SERCA activity (Fig. 2A).
The next objective was to determine the kinetic parameters for uptake for each of the two different SERCA enzymes (SERCA2b and SERCA3) when they were present in a simulated coexpression system. This was done by mixing, in various proportions in vitro, microsomes from HEK-293 cells separately expressing recombinant SERCA3 or SERCA2b with previously determined maximum SERCA activities. Fura 4F fluorescence measurements were used to measure Ca2+ uptake into the mixture samples, and the relative activity for each SERCA isoform was used to calculate the ratio of maximal velocity values. Finally, these experimentally determined activity ratios were compared with the theoretical values.
For each mixture sample, the contribution (activity ratio) from the two isoforms was derived by computer fit of the data to the two-component model given by Eq. 4, in which the apparent Ca2+ affinities for each SERCA isoform were set to the values experimentally determined previously (0.26 and 1.10 μM for SERCA2b and SERCA3, respectively). Figure 3A shows a representative uptake curve for a mixture of SERCA3 and SERCA2b with an expected theoretical activity ratio of 4.2. Examination of the residuals (Fig. 3A, inset) indicates that the data are well fit by the model across the full range of [Ca2+] measured. As illustrated in Fig. 3B, the derived SERCA3-to-SERCA2b activity ratio matched very closely the expected activity ratios over a wide range of generated theoretical ratios, and, consequently, the slope of the relation between predicted and measured activity ratios was very close to 1. These data demonstrate the reliability of our method for extracting activity ratios from mixtures of SERCA3 and SERCA2b. This principle was then applied to measurement of SERCA2b and SERCA3 activities in cellular systems where these two proteins are coexpressed endogenously.
Cultured Lymphocytes Endogenously Expressing SERCA2b and SERCA3
Microsomes from three different human cell lines (human T lymphocyte cell lines Jurkat and HPB-ALL and the human colonic epithelial cell line HT-29) were initially tested by semiquantitative immunoblotting to estimate the relative level of SERCA2b and SERCA3 protein expression. Of the three cell lines tested, Jurkat cells contained the highest relative level of SERCA3 (data not shown). Therefore, Jurkat cells were chosen as the first system from which to derive the kinetic properties of endogenous SERCA3.
The proportion of SERCA3 to SERCA2b protein activity in microsomes from Jurkat lymphocytes was first estimated by semiquantitative immunoblotting using microsomes isolated from HEK-293 cells transfected with SERCA2b or SERCA3 as standards (see materials and methods) (Fig. 4A). In general, in Jurkat cell microsomes, SERCA3 activity was estimated to be approximately twofold higher than SERCA2b activity, although activity varied slightly (1.6- to 1.9-fold) between different preparations of Jurkat cell microsomes.
Figure 4B illustrates the Ca2+ uptake activity measured in microsomes isolated from Jurkat T lymphocytes. Thapsigargin inhibited >99% of the observed Ca2+ uptake, indicating that activity measured was due exclusively to SERCA enzymes (data not shown). The overall endogenous SERCA activity in microsomes isolated from Jurkat cells was very low compared with SERCA activity in heterologous systems (<5% of the activity in recombinant microsomes from HEK-293 cells) or in muscle membrane preparations (typically <1% of the activity measured under the same assay conditions for sarcoplasmic reticulum isolated from cardiac or skeletal muscle tissues). This low activity, however, was sufficient to derive reliable kinetic parameters because of the highly sensitive nature of this assay system. As summarized in Table 1, analysis of the uptake data (Fig. 4B) revealed separate kinetic components, consistent with the existence of two functional SERCA isoforms with high and low apparent Ca2+ affinity. Experiments conducted with five independent Jurkat cell microsome preparations indicate that, on average, the value of the component with lower Ca2+ affinity was 1.04 ± 0.07 μM and that of the higher-affinity component was 0.27 ± 0.02 μM. These values are very close to those determined for the individual SERCA3 and SERCA2b isoforms expressed in HEK-293 cells (1.10 ± 0.04 and 0.26 ± 0.01 μM, respectively). However, there was a discrepancy between the derived activity ratio of lower- to higher-Ca2+-affinity components (0.8 ± 0.2) and the SERCA3-to-SERCA2b activity ratio based on semi-quantitative immunoblotting (1.8 ± 0.1). This difference may be explained by the presence in Jurkat cells of one wild-type and one mutated nonfunctional SERCA3 allele (see below).
Freshly Isolated Lymphocytes Endogenously Expressing SERCA2b and SERCA3
Jurkat T cells are a transformed and immortalized cell line; therefore, these experiments were also repeated in microsomes prepared from lymphocytes isolated from freshly excised human tonsils. Once again, the relative proportion of the two different SERCA isoforms expressed in these tonsil lymphocytes was estimated by semiquantitative Western blotting using the same approach described above for Jurkat cells, resulting in a SERCA3-to-SERCA2b activity ratio of 4.2 ± 0.5 (n = 4), as illustrated in Fig. 5A and summarized in Table 1.
The endogenous SERCA activity in these tonsil lymphocytes was then measured using fura 4F. Although the overall endogenous SERCA activity in these cells was also very low compared with that in recombinant expression systems or native cardiac/skeletal sarcoplasmic reticulum (see above), it was sufficient to derive reliable kinetic parameters. Figure 5B depicts a Ca2+ uptake velocity-free [Ca2+] curve for a single preparation of human tonsil lymphocyte microsomes. Analysis of the uptake data obtained with four independent tonsil lymphocyte microsome preparations revealed two separate kinetic components, with an activity ratio of 3.5 ± 0.4 (lower to higher affinity; SERCA3-to-SERCA2b). This derived ratio is consistent with the immunoblot SERCA3-to-SERCA2b ratio of 4.2 ± 0.5 (Table 1). These kinetic components exhibited apparent Ca2+ affinities of 1.1 ± 0.2 and 0.26 ± 0.01 μM, which are very close to values determined for the individual SERCA3 and SERCA2b isoforms expressed recombinantly (1.10 ± 0.04 and 0.26 ± 0.01 μM, respectively).
SERCA3-Specific Inhibitory Antibody PL/IM430
The data presented above confirm that endogenous SERCA activity in lymphocytes displays two distinguishable components with kinetic properties consistent with those obtained from individual recombinantly expressed SERCA2b and SERCA3 isoforms. To determine whether the component of activity with low apparent Ca2+ affinity corresponds to SERCA3, the monoclonal antibody PL/IM430 was used. PL/IM430 has been demonstrated to be highly specific to human SERCA3 and to reduce human SERCA3 activity by 80% (11). The idea was to use PL/IM430 to significantly inhibit SERCA3 activity and then to determine the kinetic parameters of the remaining SERCA activity, which should correspond mostly to SERCA2b. The functional specificity of PL/IM430 inhibition was first confirmed under the present assay conditions using recombinant SERCA3 and SERCA2b, respectively. As depicted in Fig. 6, A and B, the addition of PL/IM430 resulted in a significant reduction (by ∼75%) of SERCA3 Ca2+ transport activity, whereas it had no significant effect on SERCA2b activity.
When tested on Jurkat microsomes, the addition of PL/IM430 resulted in a significant reduction in the total SERCA activity compared with the control (Fig. 6C). This reduction was due to a selective inhibition of the component with lower apparent Ca2+ affinity (84% inhibition) compared with 7% inhibition of the higher-affinity component (n = 3; Table 1). PL/IM430 antibody incubation with human tonsil lymphocytes induced similar selective inhibition of the low-affinity component (68% vs. 16% for the higher-affinity component, n = 2; Fig. 6D, Table 1). These results confirm that, in Jurkat and tonsil lymphocyte preparations, the activity of the component with lower apparent Ca2+ affinity corresponded to functional SERCA3.
Although the SERCA3-to-SERCA2b activity ratio in the different lymphocyte preparations ranged from ∼0.95 in Jurkat to 4 in tonsil lymphocytes, the derived apparent Ca2+ affinity for the two SERCA components from these two lymphocyte preparations was, in all cases, very close to values determined for heterologously expressed individual SERCA isoforms. The validity of the assignment of these components to SERCA isoform activity is supported by the fact that the calculated ratio of activities corresponded closely with expectations based on immunoblot data and by the selective inhibition of the component with lower apparent Ca2+ affinity by the human SERCA3-specific antibody PL/IM430. From these observations, we conclude that SERCA isoforms endogenously expressed in lymphocytes display kinetic properties identical to those determined for individual isoforms expressed in heterologous systems.
Detailed characterization of SERCA3 function has thus far been restricted to recombinant protein expressed heterologously. Although a number of tissues express high levels of the SERCA3 pump, it is almost always coexpressed with SERCA2b; thus, determination of the properties of SERCA3 in its endogenous environment has been a challenging proposition. This study employed two distinct cellular systems that express varying levels of SERCA3 compared with SERCA2b and an assay system that is highly sensitive to overcome this hurdle. Our data clearly indicate, for the first time, that endogenously expressed SERCA3 displays low apparent affinity for Ca2+, as has been observed previously for the heterologously expressed isoform (34). The close match between the apparent Ca2+ affinity values of endogenous and exogenously expressed SERCA3 suggests that the native protein is not likely to be regulated by an accessory protein, in contrast to the other two SERCA isoforms, which are regulated by phospholamban and/or sarcolipin (36). In fact, SERCA3 lacks the residues that are critical for functional association with phospholamban (45, 46).
A highly sensitive fura 4F-based assay was necessary to measure reliably the low Ca2+ transport activities in microsome samples prepared from cell types expressing SERCA3 and to distinguish between the activities of the coexpressed SERCA2b and SERCA3 pumps. The principal advantage of the fluorescence-based continuous monitoring assay is the ability to determine a complete Ca2+ uptake velocity-free [Ca2+] curve and the kinetic parameters for the Ca2+ uptake from a single microsome sample (25) (Fig. 2). In comparison, 45Ca2+ uptake requires a large number of independent measurements to generate the same data and, thus, is more time and material consuming and, most importantly, increases the risk of introducing intra- and interexperimental variability.
The assay system and subsequent analysis were confirmed in several ways. 1) The kinetic parameters derived from recombinant microsomes isolated from HEK-293 cells transfected with human SERCA2b and SERCA3 were very similar to those reported previously (34). 2) With use of mixtures of microsomes from transfected HEK-293 cells containing various SERCA3-to-SERCA2b activity ratios, the assay reliably predicted two kinetic components with parameters consistent with those obtained from the individual components of the mixture. 3) Treatment with the highly specific anti-human SERCA3 inhibitory antibody PL/IM430 resulted in ∼70% diminution of the low-affinity component of the uptake activity, clearly associating these kinetic parameters with SERCA3 activity and confirming the assignment of the kinetic components and parameters.
It is interesting to note the discrepancy in Jurkat cells between the derived SERCA3-to-SERCA2b activity ratio obtained from the uptake assay (0.8 ± 0.2, n = 5) and the activity ratio based on semiquantitative immunoblotting (1.8 ± 0.1, n = 5). These data suggest that there is twice as much SERCA3 protein as activity. This result may be explained by the previous observation by two different groups that Jurkat lymphocytes express two SERCA3 alleles, one of which gives rise to a nonfunctional isoform as a result of the substitution of a highly conserved residue, Met817, by Ile817 (39, 51). Equal expression of active and inactive SERCA3 alleles at the protein level would explain the observed activity ratio being half the protein ratio. Although other explanations are possible, for example, SERCA3 is more susceptible than SERCA2b to inhibition and/or inactivation during membrane isolation or SERCA3 is present in a compartment with greater passive Ca2+ permeability than SERCA2b, there is no evidence to support such contentions. Moreover, such properties would have to be selective to Jurkat cells, because no discrepancy between activity and protein was observed in the tonsil lymphocytes.
The data presented here clearly indicate, for the first time, that endogenous SERCA3 expressed in Jurkat T cells and human tonsil lymphocytes is inherently a low-affinity Ca2+ pump. Additional experiments were also performed in freshly isolated rabbit and dog spleen lymphocytes. The derived kinetic parameters from those studies were also consistent with a two-component fit to the uptake data with a low-affinity (K0.5 = 1.1–1.5 μM) and a high-affinity (K0.5 = 0.3–0.4 μM) component with relative activities suggesting a SERCA3-to-SERCA2b activity ratio of 0.6–1.0 (data not shown). However, the lack of an inhibitory antibody specific to SERCA3 prevented validation of these data (note that PL/IM430 is highly specific to only human SERCA3). Nonetheless, because of the conserved nature of SERCA3 amino acid sequence among many species (human, mouse, rat, rabbit, dog, chicken, and cow) and because all six human SERCA3 isoforms display lower apparent affinity for Ca2+ (8, 37), it is highly likely that SERCA3 pumps also display low apparent Ca2+ affinity in these animal species. Indeed, studies with recombinant rat and rabbit SERCA3 proteins have shown that those enzymes display lower apparent affinity for Ca2+ as expected (15, 44).
An interesting observation of the present study was the relative expression of the two SERCA pumps in the different cellular systems. Jurkat T cells and rabbit and dog lymphocytes expressed comparable levels of SERCA3 and SERCA2b or lower levels of SERCA3 than SERCA2b, but the relative SERCA3 level in freshly isolated human tonsil lymphocytes was at least three- to fivefold higher than that of SERCA2b. Since the tonsils were obtained from individuals who underwent tonsillectomy procedures, it is feasible that SERCA3 is upregulated during lymphocyte activation to generate the immune response that led to the inflammation that eventually resulted in tonsillectomy. This speculation is consistent with a study showing that chronic stimulation of endothelial cells with histamine resulted in the upregulation of SERCA3 and modification of intracellular Ca2+ homeostasis (22).
We have shown that endogenous SERCA3 pumps display lower apparent affinity for Ca2+. It is important to note that the actual Ca2+ transport binding residues in the transmembrane region are conserved among all three SERCA isoforms. Studies with chimeric protein constructs have also shown that a combination of the COOH-terminal hydrophobic region and a large portion of the central cytoplasmic loop is necessary to define isoform-specific apparent Ca2+ affinity (44). Kinetic studies have demonstrated several major differences in the rates of partial reactions between SERCA3 and SERCA1, which can account for the SERCA3-type low apparent affinity via an overall stabilization of the lower-affinity E2 conformation in SERCA3 (15). These data indicate that complex global conformational differences between SERCA3 and the other isoforms involving quite large stretches of the protein underlie the difference in apparent Ca2+ affinity.
Given that SERCA3 is always coexpressed with SERCA2b, it is interesting to consider the physiological implications of expressing two SERCA pumps with different Ca2+ affinities. Because SERCA3 is selectively expressed in a range of specialized cell types, such as endothelial cells, lymphocytes, and pancreatic β-cells, it seems likely that its function is important for the regulation of specialized Ca2+ signaling pathways and for the maintenance of Ca2+ homeostasis in these cell types. Several studies of Ca2+ signaling mechanisms in such cell types suggest an important physiological role for a SERCA pump with high and low Ca2+ affinity. For example, studies using pancreatic β-cells isolated from SERCA3-knockout mice indicate that SERCA3 becomes active when cytosolic [Ca2+] rises and appears to be required for normal Ca2+ oscillations in response to glucose, a response mediated by store-operated Ca2+ (SOC) entry (3, 6, 7). SERCA3 has also been shown to be directly involved in maintaining SOC channel-mediated Ca2+ entry in platelets, where the two SERCA isoforms reside in two different stores: SERCA2b in Ca2+ stores associated with the dense tubular system and SERCA3 in lysosome-associated, acidic Ca2+ stores. Depletion of the acidic stores leads to the apparent formation of macromolecular complexes involving transient receptor potential proteins, type II inositol trisphosphate receptors, and SERCA3 (42). Furthermore, stromal interaction molecule-1 appears to regulate the refilling of these acidic stores by directly interacting with SERCA3 (31). These data indicate that SERCA3 forms a major Ca2+ sequestration pathway in these nonmuscle cell types. It is also interesting to note the existence of four separate Ca2+ pools or compartments in Jurkat lymphocytes (21) and the possible rearrangement of specific compartments during lymphocyte activation (40). Although previous evidence suggests codistribution of SERCA2b and SERCA3 proteins (39) in Jurkat cells, it is possible that SERCA3 and SERCA2b are recruited to different compartments during cellular activation, thus contributing to isoform-selective distinct functions for these two SERCA proteins. Polarized distribution of SERCA pumps has been observed in exocrine epithelial cells, where SERCA2b is expressed at the apical pole and SERCA3 at the basal pole, where SOC entry occurs (28, 38).
On the basis of such observations, it is tempting to speculate that, in lymphocytes where SERCA3 is abundantly coexpressed with SERCA2b, the two pumps would have distinct roles, particularly in light of unique features of T cell Ca2+ signaling: the rise in [Ca2+] is never explosive (as in muscle cells); rather, it is a slow increase that is sustained for a period of minutes to hours and, potentially, days; the amount of Ca2+ in intracellular stores is very low, and the lymphocyte Ca2+ stores can remain empty for prolonged periods to accommodate Ca2+ entry via SOC channels, which is the primary mode of Ca2+ influx in lymphocytes; and finally, the repetitive release and reuptake of Ca2+ generate oscillations, the frequency of which governs the efficiency and specificity of gene expression and cell differentiation in T cells (16, 29, 41, 49). In this scenario, the high-affinity SERCA2b isoform would play the housekeeping role, i.e., maintain intracellular stores and restore cytosolic [Ca2+] under normal basal physiological conditions, when the resting [Ca2+] in lymphocytes is ∼100 nM (29). The low-affinity SERCA3 isoform would be the ideal candidate to refill specialized stores and maintain cytosolic Ca2+ under the sustained conditions of high intracellular [Ca2+] that accompany T cell activation and typically range over 1 μM (29). These SERCA3 stores would be located in the vicinity of high-[Ca2+] microdomains (or rearrange after T cell receptor activation) corresponding to sites of local Ca2+ entry via SOC channels, where the local [Ca2+] can easily far exceed 1 μM (17). SOC channels undergo rapid feedback inhibition by high local intracellular [Ca2+] (52), and SERCA3 can thus promote their sustained activity by buffering the Ca2+ in the vicinity of SOC entry sites. Furthermore, SERCA3 pumps and pools may act as insulating elements that prevent Ca2+ flow toward the SERCA2b-filled stores, thereby ensuring that these stores remain unloaded to allow sustained Ca2+ entry via SOC channels.
It has been assumed, but never demonstrated, that endogenous SERCA3 would display the same characteristics as recombinantly expressed SERCA3. This study provides the first unequivocal evidence that low Ca2+ affinity is an inherent property of SERCA3 and likely not the result of regulation by a modulatory protein. In addition to having distinct biochemical properties and physiological functions, SERCA3 also has the highest isoform diversity (6 alternatively spliced isoforms compared with only 2 for SERCA1 and SERCA2) among SERCA gene products. Further studies are clearly necessary to deduce how these isoforms are regulated and how their concerted or differential actions influence cellular Ca2+ signaling.
This publication was made possible by grants from the Heart and Stroke Foundation of Alberta, Northwest Territories, and Nunavut to J. Lytton and M. E. Kargacin.
Present address of P. C. Chandrasekera: Department of Physiology, Wayne State University, 421 East Canfield St., Detroit, MI 48201.
- Copyright © 2009 the American Physiological Society