c-Myb–Dependent Cell Cycle Progression and Ca2+ Storage in Cultured Vascular Smooth Muscle Cells
Abstract Considerable controversy surrounds the role of the c-myb proto-oncogene in vascular smooth muscle cells (VSMCs). Previous investigations using antisense approaches have suggested a relationship between c-myb expression, cell cycle progression, and cytoplasmic Ca2+ concentration ([Ca2+]cyt). However, the ability of certain antisense oligonucleotides to bind and inactivate growth factors allows alternative explanations. To define more specifically the role of c-Myb in cultured VSMCs (SVE and A10 cell lines), we have generated stable cell clones expressing a dominant-negative c-Myb lacking critical elements of the DNA binding domain (Δ5-SVE) and transiently transfected cell populations (GRE-MEn-SVE and GRE-MEn-A10) expressing a glucocorticoid-inducible chimeric protein that targets the Drosophila Engrailed repressor domain to c-Myb–responsive promoters. The Δ5-SVE clones and GRE-MEn cell populations exhibit a 60% reduction in mean intracellular c-Myb activity, as measured by cotransfection assays with a c-Myb–responsive reporter, a 42% decrease in the mean S phase entry of growth-arrested (G0) cells after serum stimulation, and a 36% inhibition of mean cell proliferation over 4 days. These cells also display 28% (34-nmol/L) and 30% (42-nmol/L) reductions in mean [Ca2+]cyt at G0 and at the G1/S interface, respectively, as well as significant reductions in the peak [Ca2+]cyt responses to thapsigargin (5 μmol/L) and caffeine (10 mmol/L). These latter reductions in operationally defined Ca2+ pools were observed both at different stages of the cell cycle and after transient induction of the dominant-interfering construct, suggesting that c-Myb regulates these releasable Ca2+ stores independent of its effects on cell cycle progression.
The proto-oncogene c-myb is homologous to the transforming gene product of the avian myeloblastosis virus and encodes a 75-kD transcription factor (c-Myb) present in diverse cell types (reviewed in References 1 and 21 2 ). c-Myb expression occurs at low levels in quiescent cells, increases rapidly as cells begin to proliferate, and peaks in the late G1 phase of the cell cycle.3 4 5 In some systems, c-Myb plays a critical role in regulating cell growth and differentiation.6 Mice homozygous for the c-myb null allele die before day 15 of fetal life because of the absence of hepatic hematopoiesis.7 Inhibition of c-Myb function interferes with proliferation of certain myeloid leukemic cells as well as the development of T cells.8 9 The growth of VSMCs may also depend on c-Myb activity.5 10 11
These actions of c-Myb have been attributed to transactivation of other cell cycle–associated genes, such as c-myc, proliferating cell nuclear antigen, cdc2 kinase, and DNA polymerase-α.12 13 14 15 However, specifically timed alterations in the levels of [Ca2+]cyt are also necessary for passage of cells through cell cycle checkpoints (reviewed in Reference 1616 ). Although it has been hypothesized that Ca2+-dependent cell cycle events require the release of Ca2+ from intracellular stores,17 18 19 20 little is known about the regulation of such stores during cell cycle progression in VSMCs.
Previous investigations using antisense approaches have suggested a relationship between expression levels of c-myb, proliferation, and [Ca2+]cyt in VSMCs.10 11 21 22 However, these studies did not examine c-Myb–dependent effects on intracellular Ca2+ stores. Moreover, several recent studies have shown that the biological effects of antisense phosphorothioate oligonucleotides to c-myb (AS-c-myb) can be due to the binding and inactivation of growth factors.23 24 25 Thus, considerable controversy presently surrounds the role of c-Myb in VSMCs.26 27 To address these issues, we now report on the development of two independent approaches to reducing c-Myb–dependent gene activity and define more specifically the effects of these manipulations on cell cycle progression and intracellular Ca2+ levels in VSMCs.
The transcription factor c-Myb binds to its consensus hexanucleotide DNA sequence [5′-PyAAC(G/Py)G-3′] as a monomer.28 29 30 This interaction is critically dependent on the integrity of the second (R2) and third (R3) of three imperfect 51 amino acid repeats that constitute the DNA binding domain of the transcription factor.31 32 Homodimerization of c-Myb through a leucine zipper motif in the negative regulatory domain inhibits both DNA binding and transactivation.30 33 This informed the design of a construct encoding a dominant-negative form of c-myb (Δ5-Myb), which lacks an intact DNA binding domain but still complexes with endogenous c-Myb and/or other cofactors involved in transactivation.34 35 Therefore, constitutive overexpression of this construct should inhibit the function of the endogenous transcription factor (see Fig 1⇓).
The DNA binding region of c-Myb can also be coupled to the repressor domain of the Drosophila Engrailed transcription factor (MEn). This chimeric protein has previously been used to suppress c-Myb–dependent gene transcription in vitro and in vivo.9 This construct (MEn) was rendered inducible by insertion of a glucocorticoid responsive element in its promoter (GRE-MEn) and was used to evaluate the effects of acute repression of c-Myb–dependent gene activity at defined stages of the VSMC cell cycle (see Fig 1⇑).
Materials and Methods
All manipulations and preparations of plasmid DNA were carried out by standard methods. All constructs were checked by restriction mapping, and in-frame deletions and fusions were evaluated by standard double-strand sequencing techniques.
Construction of the Dominant-Negative Expression Vector pΔ5
The full-length murine c-myb expression vector p3090 (a gift of Dr Michael Kuehl, Bethesda, Md) was digested with BstXI-EcoRI and blunted. Successful relegation deleted base pairs 357 to 588 of c-myb cDNA. The resulting gene product lacks amino acid residues 109 to 185 which constitute critical portions of the R2 and R3 regions of the c-Myb DNA binding domain but retains transactivating and negative-regulatory domains.
Construction of the Glucocorticoid-Inducible MEn Expression Vector pGRE-MEn
pUCMEnT (a generous gift of Dr Kathleen Weston, London, UK) contains, at the BamHI site of pUC12, c-myb sequences encoding amino acid residues 71 to 200 of the murine c-Myb DNA binding domain, ligated in-frame with Drosophila Engrailed sequences encoding amino acids 2 to 298 of the alanine-rich repressor domain, as well as sequences encoding the c-myc 9E10 epitope tag. A HindIII site precedes the 9E10 epitope sequences in pUCMEnT.9 The glucocorticoid-inducible eukaryotic expression vector pGRE5-2Neo (Dr Elizabeth Harrington, Boston, Mass) was created by cloning a 1509-bp blunt-ended EcoRI-BamHI fragment encoding the SV40 promoter–Neo ORF–SV40 polyA assembly from pRc/CMV (Invitrogen) into a blunt-ended Sal I site of pGRE5-2 (United States Biochemical). pGRE5-2Neo is composed of five GREs placed upstream from the adenovirus 2 major late promoter TATA region and a Neo resistance cassette. Initially, the entire MEnT chimeric assembly was transferred as a BamHI fragment from pUCMEnT into the BamHI site of pUHG10-3 (Dr Hartmut Weiler, Cambridge, Mass), and the resulting plasmid was designated ptetMEnT. Since the BamHI site of pUHG10-3 is preceded by a Sac II site, a Sac II digestion of the appropriately oriented ptetMEnT, followed by blunting and further HindIII digestion, allowed forced cloning of an Engrailed repressor (MEn) encoding fragment into the Bgl II–blunt-HindIII site of pGRE5-2Neo. The resulting construct was designated pGRE-MEn.
Construction of the Glucocorticoid-Inducible c-Myb Expression Vector pGRE-Myb
Similarly, a full-length murine c-myb fragment encoding the BamHI–blunt-HindIII fragment from p3090 was directionally cloned into the Xho I–blunt-HindIII site of pGRE5-2Neo. The resulting plasmid was designated pGRE-Myb.
The SV40–large-T immortalized rat VSMC line (SVE) was a gift of Dr Christopher Reilly (Merck Sharpe, and Dohme, West Point, Pa36 ). A10 rat embryonic aortic smooth muscle cells were purchased from the American Type Culture Collection cell repository (CRL No. 147637 ). The two cell lines were cultured at 37°C, under 5% CO2, in DMEM supplemented with 10% heat-inactivated (55°C for 30 minutes) FBS and in 1% penicillin G (104 U/mL)/streptomycin (104 μg/mL). Cell lines were trypsinized (0.5% trypsin-EDTA) and passaged at least twice weekly. All studies of stable clones were performed beyond passage 12; experiments involving transiently transfected cells were carried out after 4 days, but within 10 days of transfection. All tissue culture reagents were provided by JRH Biosciences, and all chemical reagents were of the highest chemical purity available.
Transfection and Clonal Selection
Supercoiled pΔ5 and the Hy resistance vector tgCMV/HyTK (Targeted Genetics) were cotransfected in a 10:1 molar ratio into SVE cells using the calcium phosphate precipitation technique as previously described.21 Two separate transfections of 106 SVE cells were carried out, and Hy selection (400 μg/mL) was maintained for >6 weeks. This led to the isolation of 32 Hy-resistant clones, only four of which were found to express Δ5-Myb mRNA. These were isolated by limiting dilution, and the two with the highest mutant mRNA levels were used for all subsequent experiments. Both SVE and A10 cells were also studied after transient transfections with either 5 μg of supercoiled pGRE-MEn, pGRE-Myb, or pGRE5-2Neo. Successful transfection of >30% of the cell population was documented in cotransfection experiments with pCMV-β-Gal followed by LacZ staining as previously described.38 The transiently transfected cell populations were split after 24 to 48 hours, and cell cycles were synchronized as outlined below.
Induction of GRE-Driven Expression Constructs
Before experimental studies, cell cycle–synchronized cell populations were exposed to cell cycle stage-appropriate medium either containing 0.5 μmol/L Dex or lacking Dex. Northern blots and functional assays conducted after 4 hours of Dex treatment showed strong induction of exogenous mRNA and protein activity (see “Results”). For experiments assessing the effect of the MEn repressor on proliferation, synchronized cell populations were exposed to medium with or without Dex for the first 16 hours of the 96-hour assay. In determining the effects of MEn expression on cell cycle progression, the duration of medium±Dex exposure equaled that of serum stimulation unless otherwise specified in the text. For mim-CAT assays, and [Ca2+]cyt determinations, the period of Dex exposure was 4 to 6 hours.
The Quantification of c-Myb Function
Each of the wt cells, stable clones, and transiently transfected cell populations described above were also transiently cotransfected, by calcium phosphate precipitation in a 10:1 molar ratio, with the c-Myb responsive construct p5 mim-CAT (gift of Dr Kathleen Weston, London, UK9 ) (see Fig 1⇑), and the control construct pCMV-β-Gal (Clontech). After 24 hours, the cells were split, and total cell lysates from a fraction of the cells were subjected to both CAT and β-Gal enzyme assays as per the commercial kit manufacturer’s specifications (Promega). Enzyme activity was quantified by liquid scintillation counting. After correction for transfection efficiency (LacZ), individual CAT activities were normalized to control values (ie, wt or control activity=1) and expressed as relative CAT activity. The data represent the mean±SEM of at least two experiments.
Cell Cycle Progression
Each clone or transiently transfected cell population was plated at medium density (≈5×105 cells per 25-cm2 tissue culture flask) in multiple parallel flasks and allowed to attach overnight in 10% FBS-DMEM. The next day, cells were washed twice in PBS and synchronized via growth arrest by serum starving them in 0.25% FBS-DMEM for 48 to 72 hours. We have demonstrated that this protocol is highly effective in arresting ≈90% of all VSMCs in the G0/G1 phase of the cell cycle.22 Parallel cultures of growth-arrested cells were serum-stimulated with 10% FBS-DMEM and then serially harvested and ethanol-fixed at 4- to 8-hour intervals (G0, G0+8, G0+12, G0+16, G0+24, and G0+32). This process involved two washes in PBS, trypsinization, centrifugation for 10 minutes at 500g, and resuspension of the cell pellet at a density of 4×106 cells/mL in serum-free DMEM containing 0.1% sodium azide. With gentle vortexing, an equal volume of 100% ethanol was slowly added to the suspension. Cells were stored at 4°C until all specimens were similarly fixed. The entire batch was pelleted as outlined above and resuspended at 2×106 cells/mL in PBS with 0.1% Triton X-100. Cells were then incubated for 1 hour at 37°C with 100 μg/mL of RNase A. After another centrifugation, cells were resuspended in PBS containing 50 μg/mL propidium iodide and incubated overnight at 4°C in the dark. DNA quantification was performed on a FACStar Flow cytometer. By analyzing raw cell cycle data with a MODFIT computer program (Verity Software), the proportion of cells in different stages of the cell cycle was determined for each of the above time points. Each stable clone, or transfected cell population, was studied on at least two separate occasions.
Each clonal cell line or transiently transfected cell population was plated at 20 000 cells per well in separate 24-well cluster plates (Sarstedt Inc). The cells were allowed to attach overnight in 10% FBS-DMEM, washed twice with PBS, and then growth-arrested as described above. Of note, the plating efficiency (the number of cells counted after overnight attachment) did not differ between the various experimental and control cell populations. Every 24 hours after serum stimulation, cells from six of the initial 24 wells were washed once with PBS, trypsinized, and counted with a Coulter Counter (Coulter Electronic). The mean number of cell doublings over 4 days was calculated by the following equation: n=ln(x/y)/ln(2), where n is the number of doublings, x is the cell count at G0+96, and y is the cell count at G0.
Determination of Free Cytoplasmic Ca2+ Levels with Fura 2
Cells were plated at a low density (<50 000 cells) on 25-mm circular glass coverslips, allowed to attach overnight, and growth-arrested. At 4- to 8-hour intervals after serum stimulation, cells were loaded for 30 minutes at 37°C and 5% CO2 in media supplemented with 4 μmol/L of the cell-permeant fluorescent Ca2+ indicator dye fura 2-AM (Molecular Probes). Coverslips were washed in a Ca2+-measurement buffer (mmol/L: NaCl 137, KCl 2.7, CaCl2 0.68, MgCl2 0.49, and glucose 5.5, pH 7.4) and mounted on a modified Leiden chamber in which the coverslip constituted the bottom and to which 1 mL of Ca2+-measurement buffer was added. [Ca2+]cyt was measured by ratio imaging using an Image-1 digital ratio imaging system (Universal Imaging) equipped with an Olympus IMT-2 inverted microscope, a Dage-MTI CCD7 series video camera, a Genisys image intensifier, a Pinnacle REO-650 optical disk drive, and a color video monitor/printer as described previously.39 Fura 2 fluorescence images were monitored and acquired at 510-nm emission with alternating 340- and 380-nm excitation. Single cell images and the corresponding cytoplasmic 340/380 ratios, calculated on a pixel-by-pixel basis, were collected for data processing. With the same experimental settings for the imaging system, fura 2 ratio values were calibrated in vitro to known Ca2+ concentrations.40 Briefly, fura 2 free acid (2 μmol/L) was dissolved in a Ca2+-free HEPES buffer solution (mmol/L: KCl 110, NaCl 10, MgCl2 1, HEPES 25, and EGTA 1.5, pH 7.0), and variable total Ca2+ was added in quantities calculated to yield free-Ca2+ concentrations ranging between 36 nmol/L and 40 μmol/L. Rmin was determined at zero Ca2+ (free Ca2+ <10 nmol/L), and Rmax was determined at 40 μmol/L. Kd was determined by fitting the experimental R values at various levels of free [Ca2+] using the equation [Ca2+]free=Kd(Sf2/Sb2)[(R−Rmin)/(Rmax−R)], where the factor Sf2/Sb2 corrects for fura 2 ion sensitivity at 380 nm. A similar procedure was used for in situ calibration, where 2 μmol/L of the nonfluorescent analogue of the Ca2+ ionophore 4-BrA23187 (Molecular Probes) was used to collapse Ca2+ gradients during 15-minute incubations of fura 2–loaded VSMCs in a series of EGTA/Ca2+ buffers with free [Ca2+] ranging from 36 to 1270 nmol/L. Kd was determined to be 218 nmol/L under these conditions. Calculated values of resting [Ca2+]cyt determined by in vitro and in situ calibration were statistically indistinguishable (P>.2).
To analyze releasable intracellular Ca2+ stores, the Ca2+ measurement buffer (0.68 mmol/L CaCl2) was aspirated and replaced with a Ca2+-free buffer (0 mmol/L CaCl2, 1.5 mmol/L EGTA) containing either the inhibitor of SERCA, Tg (5 μM), or the agonist of Ca2+ store release, caffeine (10 mmol/L). The [Ca2+]cyt responses to these agents were monitored over 20-minute periods by ratio image acquisition at intervals between 6 to 120 seconds. Images of entire fields were stored to optical disk for subsequent replay and data sampling from large numbers of individual cells. Data processing was performed with Image-1 software.
Changes in [Ca2+]cyt, as well as absolute [Ca2+]cyt values, are reported with the understanding that a portion of the cytoplasmic signal could represent organellar Ca2+. Complete in situ calibrations were performed with each cell type and in each experimental condition. There was little difference between calibration curves for different growth conditions and cell types. Autofluorescence at both 340 and 380 nm was undetectable at the gain settings used.
Statistical significance is defined as P<.05 unless otherwise given as a specific P value or 95% confidence interval. Statistical tests were performed using the Primer software program (Version 1.0, McGraw Hill). The various tests used include Student’s t test (mim-CAT, number of cell doublings, and [Ca2+]cyt data) and the comparison test for proportions (cell cycle distributions).
Considerable difficulty was encountered in producing stable VSMC clones that expressed c-Myb–interfering molecules. After large-scale transfection attempts, only 4 of 32 antibiotic-resistant SVE clones were found to express Δ5-Myb mRNA. The two with the highest constitutive expression of the Δ5-Myb construct were selected for further study (Δ5-SVE). Despite multiple attempts with varied transfection and selection protocols, SVE cells could not be stably transfected with the inducible c-Myb-Engrailed repressor construct (pGRE-MEn). However, transient transfections with this construct were successfully carried out, in which >30% of cells were shown to be cotransfected with pCMV-β-Gal (as determined by LacZ staining data not shown). To assess the effects of c-Myb–modifying molecules in the absence of the immortalizing SV40–large-T antigen,36 transfection experiments were also conducted in the A10 VSMC cell line derived from the embryonic rat aorta.37
Dominant-Interfering c-Myb Constructs Decrease c-Myb Function, G1/S Progression, and Cell Proliferation
Fig 2⇓ depicts the relationship between c-Myb function, G1 to S phase cell cycle progression, and overall cell proliferation in SVE cells stably expressing the Δ5-Myb construct, SVE cell populations transiently transfected with the inducible Myb-Engrailed repressor (GRE-MEn-SVE), and their respective controls. Fig 2A⇓ presents the mean±SEM normalized CAT activities obtained after transient transfections with a c-Myb–responsive reporter construct; Fig 2B⇓ expresses the data for cell cycle progression as the percentage (±coefficient of variance) of growth-arrested (G0) cells entering S phase after 16 hours of serum stimulation (G0+16 hours). Of note, this time point represents the cell cycle stage at which the maximum percentage of wt-SVE cells are either in or are about to enter S phase.21 22 Fig 2C⇓ shows the mean±SD number of cell doublings that have occurred 4 days after serum stimulation and demonstrates whether alterations in G1/S-phase progression translate into persistent effects on cell proliferation.
In a representative Δ5-SVE clone, a 40% reduction in mim-CAT activity (Fig 2A⇑, bar 1 versus bar 2) is associated with a marked 47% decrease in serum-stimulated S-phase entry (Fig 2B⇑, 38% versus 20% of cells enter S phase; 95% confidence limits for the absolute 18% difference were 16% to 20%). This alteration in S-phase progression leads to a significant inhibition in the overall serum-stimulated growth rate after 4 days (Fig 2C⇑, 3.95±0.22 versus 2.85±0.3 cell doublings, ie, a 28% reduction in growth; P<.01).
Fig 2⇑ also summarizes data from GRE-MEn-SVE cell populations and control GRE-Neo-SVE cell populations. Fig 2A⇑ reveals that Dex-induced expression of the MEn repressor decreased c-Myb–dependent mim-CAT activity by ≈55% compared with control cells treated with this glucocorticoid (P<.02). Of note, GRE-MEn-SVE cell populations, in the absence of Dex, also exhibit slightly lower mim-CAT activity compared with wt-SVE cells, which indicates some leakiness of the GRE promoter (data not shown). The reductions in mim-CAT activity achieved by transient MEn expression are associated with a significant 36% decrease in S-phase entry (Fig 2B⇑, 36% versus 23%, 95% confidence limit for difference is 11% to 15%) and a 40% reduction in overall proliferation (Fig 2C⇑, 3.8±0.2 versus 2.3±0.3 cell doublings, P<.03). Given that these flow cytometric and proliferation studies were conducted on entire cell populations, it is noteworthy that the magnitudes of these effects equaled or exceeded the LacZ determined transfection efficiency. However, LacZ staining is unlikely to have detected cells with subthreshold levels of constitutive β-Gal activity, whereas glucocorticoid exposure would enable even such transfectants to express GRE-driven constructs at much higher (detectable) levels.
For the Δ5-Myb dominant-negative and MEn repressor constructs, similar effects on proliferation and cell cycle progression were documented in at least one other Δ5-SVE clone and in several transiently transfected A10 cell populations (GRE-MEn-A10).
Dominant-Interfering c-Myb Constructs Decrease Cytoplasmic Ca2+ Levels
Previous studies from our own laboratory using constitutive overexpression of c-Myb21 and phosphorothioate antisense oligonucleotides against the proto-oncogene (AS-c-myb)22 suggested a relationship between the levels of c-Myb and [Ca2+]cyt. Although the associated growth-inhibitory effect of AS-c-myb was also observed in primary VSMCs,10 11 the relationship between c-Myb overexpression and elevated [Ca2+]cyt has not been examined in cells other than those expressing the immortalizing SV40 large-T antigen.21 22 Moreover, the above studies failed to quantify the functional activity of the c-Myb transactivator. Fig 3⇓ summarizes data from growth-arrested A10 cells in which c-myb and MEn constructs are induced by 4-hour exposures to Dex (0.5 μmol/L) and functional levels of c-Myb are compared with [Ca2+]cyt. wt-A10 cells and transiently transfected control A10 cells (GRE-Neo-A10) exhibit levels of mim-CAT activity (arbitrarily set at 1.0) that increase slightly after exposure to Dex (Fig 3B⇓, 1.2±0.2 for wt-A10+Dex and 1.3±0.2 for GRE-Neo-A10+Dex). The resting levels of [Ca2+]cyt in these cell populations are also minimally elevated by exposure to Dex ([Ca2+]cyt, 107±3 nmol/L for wt-A10 versus 116±4 nmol/L for wt-A10+Dex and 115±5 nmol/L for GRE-Neo-A10 versus 121±3 nmol/L for GRE-Neo-A10+Dex). The transient transfection of A10 cells with pGRE-Myb and the subsequent exposure of this growth-arrested cell population to glucocorticoid induce augmented expression of c-myb mRNA (Fig 3A⇓). This increase in c-myb expression, which results in a significant elevation of mim-CAT activity (Fig 3B⇓, 2.9±0.2 nmol/L for GRE-Myb-A10+Dex versus 1.3±0.1 nmol/L for GRE-Neo-A10+Dex, P<.05), is coincident with a rise in [Ca2+]cyt (Fig 3C⇓, 145±5 versus 121±3 nmol/L, P<.05). These observations are in agreement with previous studies using constitutive overexpression of c-myb.21 22 GRE-Myb-A10 cells not exposed to Dex also exhibit slightly enhanced expression of c-myb mRNA (Fig 3A⇓) but possess only marginally augmented levels of mim-CAT activity (Fig 3B⇓, 1.55±0.15 versus 1.0±0.1, P<.09) and [Ca2+]cyt (Fig 3C⇓, 120±3 versus 107±3 nmol/L, P<.04) compared with wt-A10 cells.
Growth-arrested GRE-MEn-A10 cell populations treated with Dex exhibit an augmented expression of the very low abundance endogenous c-myb message (Fig 3A⇑). Despite this possible feedback-induced upregulation of the endogenous transactivator, the potent suppression of c-Myb–dependent transcription dramatically reduces mim-CAT activity by >80% (Fig 3B⇑, 0.2±0.05 nmol/L for GRE-MEn-A10+Dex versus 1.3±0.1 nmol/L for GRE-Neo-A10+Dex, P<.01) and decreases [Ca2+]cyt by ≈38% (Fig 3C⇑, 75±3 nmol/L for GRE-MEn-A10+Dex versus 121±4 nmol/L for GRE-Neo-A10+Dex, P<.01). Of note, the relative changes in mim-CAT activity and [Ca2+]cyt seen with c-myb or MEn induction in G1/S synchronized A10 populations (G0+16 hours of serum stimulation) were virtually identical to those described above.
Dominant-Interfering c-Myb Constructs Reduce Intracellular Ca2+ Pools
Having observed a consistent correlation between the functional levels of c-Myb and [Ca2+]cyt, we sought to determine whether pharmacologically releasable intracellular Ca2+ pools might also be affected by dominant-interfering c-Myb molecules. To this end, Tg, the potent SERCA inhibitor known to deplete SERCA-maintained Ca2+ stores, was used to estimate the size of these stores as a function of c-Myb activity and cell cycle stage.
Fig 4A⇓ shows that a near G1/S synchronized Δ5-SVE clone (72 hours of growth arrest followed by 16 hours of serum stimulation) exhibits a significantly smaller peak [Ca2+]cyt after exposure to Tg (5 μmol/L) than do wt-SVE cells at a similar stage of the cell cycle (276±31 versus 440±21 nmol/L, P<.01). A nearly identical effect was observed in a second Δ5-SVE clone and, most important, also held true for both clones under conditions of growth arrest (G0, peak [Ca2+]cyt after Tg, 103±14 nmol/L for Δ5-SVE versus 162±17 nmol/L for wt-SVE, P<.01). Although peak [Ca2+]cyt after Tg is reached faster in Δ5-SVE cells than in wt-SVE cells (Fig 4A⇓, 4.6 versus 8.0 minutes), the calculated rate of increase in [Ca2+]cyt in these mutants is actually lower (≈35 nmol/L per minute for Δ5-SVE versus ≈40 nmol/L per minute for wt-SVE). Given that these experiments were conducted in a buffer free of extracellular Ca2+, Tg-elicited elevations in [Ca2+]cyt represent a balance between the leakage of SERCA-maintained Ca2+ pools into the cytoplasm and the concurrent efflux of cytoplasmic Ca2+ into the extracellular space. Because Tg is not known to perturb the Ca2+ permeability of sarcoplasmic/endoplasmic reticular membranes, reductions in Tg-elicited Ca2+ responses are likely due to decreased SERCA-maintained Ca2+ stores, to increased rates of Ca2+ efflux, or to both.
In Fig 4B⇑, where only the mean±SEM peak [Ca2+]cyt is shown, growth arrested GRE-MEn-SVE cell populations treated for 4 hours with 0.5 μmol/L Dex possess significantly decreased resting and post-Tg [Ca2+]cyt compared with identically treated controls (GRE-Neo-SVE) (Fig 4B⇑, G0 panel: before Tg, 83±7 versus 114±11 nmol/L, P<.01; after Tg, 88±6 versus 157±17 nmol/L, P<.01). This effect was also evident after 16 hours of serum stimulation in which only the last 4 hours included 0.5 μmol/L Dex (Fig 4B⇑, G1/S panel: before Tg, 117±7 versus 147±11 nmol/L, P<.01; after Tg, 197±20 versus 283±18 nmol/L, P<.001). Of note, these data show that Dex-treated GRE-MEn-SVE cell populations compared with identically treated GRE-Neo-SVE cell populations at the same stage in the cell cycle experience a greater proportional reduction in Tg-releasable Ca2+ stores than in resting [Ca2+]cyt. For example, Fig 4B⇑ shows that the acute transient expression of MEn at G0 causes a 27% (31-nmol/L) decrease in resting [Ca2+]cyt (from 114±6 to 83±7 nmol/L) and a 44% (69-nmol/L) decrease in the Tg-sensitive peak [Ca2+]cyt (from 157±17 to 88±6 nmol/L). Similarly, MEn expression at the G1/S interface decreased resting [Ca2+]cyt by 20% (30 nmol/L from 147±11 to 117±7 nmol/L) but reduced peak [Ca2+]cyt after Tg exposure by 30% (86 nmol/L from 283±18 to 197±20 nmol/L).
Importantly, parallel control (GRE-Neo-SVE) and experimental (GRE-MEn-SVE) cell populations were found to have similar cell cycle distributions just before the above-described 4-hour Dex exposure in late G1 (ie, at G0+12 hours, 37% G0/G1 phase and 44% S phase for GRE-Neo-SVE versus 40% G0/G1 phase 40% and 40% S phase for GRE-MEn-SVE, P>.10). This suggests that the highly significant reductions in resting and post-Tg peak [Ca2+]cyt mediated by MEn expression are unlikely to reflect retarded progression through the cell cycle. Indeed, this observation suggests that acute MEn expression can affect Ca2+ regulation independent of its effects on cell cycle progression. An examination of the cell cycle distribution of GRE-MEn-SVE cells after the above 4 hours of Dex exposure in late G1 shows only a strong trend toward reduced G1 to S transitions (ie, at G0+12 hours+4-hour Dex, 32% G0/G1 phase and 47% S phase for GRE-Neo-SVE versus 39% G0/G1 phase and 41% S phase for GRE-MEn-SVE, P<.1). Possible explanations for why a more robust inhibition of S-phase entry was not found include the following: (1) nontransfected and imperfectly synchronized cells were present, (2) the critical time point for Ca2+-dependent cell cycle events precedes G0+12 hours+4-hour Dex, and (3) earlier and more prolonged interference with c-Myb–dependent functions may be required to produce maximal inhibition of this cell cycle transition (see Fig 2⇑, G0+16-hour Dex).
In addition to the Dex-treated cell populations shown in Fig 4B⇑, the levels of resting and Tg-elicited peak [Ca2+]cyt for cell populations not treated with Dex were also examined as cells moved from G0 to the G1/S boundary. GRE-Neo-SVE cells not exposed to Dex exhibit a 50% increase in resting [Ca2+]cyt as they move from G0 to G1/S (GRE-Neo-SVE, 102±10 nmol/L at G0 phase versus 153±12 nmol/L at G1/S phase, P<.01) but a nearly 90% increase in peak Tg-elicited [Ca2+]cyt (GRE-Neo-SVE+Tg, 170±15 nmol/L at G0 phase versus 320±25 nmol/L at G1/S phase). By comparing these data with the resting and post-Tg peak [Ca2+]cyt of GRE-Neo-SVE cells exposed to Dex, we note that the glucocorticoid itself has no significant effect on Ca2+ homeostasis in these cells (GRE-Neo-SVE, 102±10 nmol/L for G0 versus 114±6 nmol/L for G0+Dex, 170±15 nmol/L for G0+Tg versus 157±17 nmol/L for G0+Dex+Tg, 153±12 nmol/L for G1/S versus 147±11 nmol/L for G1/S+Dex, and 320±25 nmol/L for G1/S+Tg versus 283±18 nmol/L for G1/S+Dex+Tg; P>.2 for all comparisons).
To examine further the effects of inhibited c-Myb function on intracellular Ca2+ stores, we exposed our clonal and transfected cell populations to the direct Ca2+ agonist caffeine. A typical experiment involving G1/S synchronized control (GRE-Neo-SVE) and MEn-expressing cell populations (GRE-MEn-SVE) is shown in Fig 5A⇓. After a 4-hour exposure to 0.5 μmol/L Dex, caffeine (10 mmol/L) evoked a significantly smaller rise in [Ca2+]cyt in MEn-expressing cells compared with identically treated control cells (Fig 5A⇓, peak [Ca2+]cyt, 182±20 versus 259±21 nmol/L, P<.01). In Fig 5B⇓, which compares peak [Ca2+]cyt after caffeine exposure in wt-SVE and Δ5-SVE cells, we confirm that decreases in functional c-Myb levels are associated with significantly diminished caffeine-evoked [Ca2+]cyt responses in both G0 (142±15 versus 103±11 nmol/L, P<.05) and G1/S (311±24 versus 163±28 nmol/L, P<.01) synchronized cells. Moreover, as observed with Tg-elicited [Ca2+]cyt, caffeine-evoked [Ca2+]cyt also increased during G0 to S cell cycle progression (wt-SVE, 141±15 nmol/L for G0 versus 311±21 nmol/L for G1/S, P<.01; Δ5-SVE, 103±11 nmol/L for G0 versus 163±28 nmol/L for G1/S, P<.05).
It is important to note that unlike the slowly rising [Ca2+]cyt observed after Tg exposure, caffeine-evoked elevations in [Ca2+]cyt are rapid (time to peak [Ca2+]cyt, 4.6 to 8 minutes for Tg in Fig 4A⇑ versus 20 seconds for caffeine in Fig 5A⇑) and are unlikely to be as affected by altered rates of Ca2+ efflux. This suggests that the reduced caffeine-mediated (and Tg-mediated) peak [Ca2+]cyt in Δ5-Myb and MEn-expressing cells reflects a true diminution of releasable Ca2+ stores rather than an enhanced rate of plasmalemmal Ca2+ extrusion alone.
In summary, the above data show that resting [Ca2+]cyt can be modulated by either endogenous or exogenous alterations in c-Myb activity from ≈75 to ≈150 nmol/L, whereas peak Tg-elicited [Ca2+]cyt, under the same manipulations, ranges from ≈90 to 450 nmol/L. Thus, either alterations in resting [Ca2+]cyt or releasable Ca2+ stores could contribute to regulation by c-Myb of Ca2+-dependent cell cycle events. The data also demonstrate that irrespective of cell cycle stage, and independent of cell cycle effects, dominant-interfering c-Myb constructs reduce releasable Ca2+ stores in cultured VSMCs.
We have previously demonstrated that in vitro suppression of c-myb levels with antisense oligonucleotides significantly reduces cell cycle progression and proliferation of VSMCs.10 Other investigators using similar approaches have reached similar conclusions in other cell types.6 We have also shown that synchronized VSMCs exhibit a rise in [Ca2+]cyt at the G1/S interface, which is immediately preceded by a 2-fold increased expression of c-myb.21 This cell cycle–dependent elevation of [Ca2+]cyt is abolished by the addition of AS-c-myb.21 22 Furthermore, growth-arrested stable VSMC clones with 2- to 4-fold overexpression of c-myb possess higher [Ca2+]cyt compared with control cells.21 22 Taken together, these results suggested that c-Myb, rather than other components of the cell cycle machinery, is responsible for the increased [Ca2+]cyt at the G1/S interface.
The above investigations also suggested that the elevation of [Ca2+]cyt at G1/S is dependent on cell cycle–associated alterations in the rates of plasmalemmal Ca2+ influx and efflux. The increased rate of cellular Ca2+ entry at G1/S, compared with G0, depends on augmented secretion of IGF-1, as evidenced by the reduction of Ca2+ influx to baseline levels after treatment with neutralizing anti–IGF-1 antibodies or AS-IGF1R.22 In contrast, treatment with AS-c-myb had no effect on the rate of Ca2+ entry into the cell.22 Moreover, several investigations with other cell types support the postulated relationship between the levels of IGF-1/IGF-1 receptor and Ca2+ influx.41 42 43 On the other hand, a reduction in the rate of Ca2+ extrusion at the G1/S interface appeared to depend on increased expression of c-myb, based on the dramatic elevation of the Ca2+ efflux rate produced by AS-c-myb but not by neutralizing anti–IGF-1 antibodies or AS-IGF1R.22 Thus, these studies implied that the increase in [Ca2+]cyt in VSMCs at the G1/S interface is achieved through an upregulation of cellular Ca2+ entry by the IGF-1/IGF-1 receptor system and a downregulation of cellular Ca2+ extrusion by c-Myb.
This model of c-Myb–dependent cell growth and Ca2+ regulation has been questioned because of nonspecific interactions of the antisense oligonucleotides used in the above studies.23 24 25 Moreover, the model did not consider the possibility that shifts in resting [Ca2+]cyt could be due to opposing shifts in intracellular Ca2+ stores, ie, that a redistribution of intracellular Ca2+ may underlie c-Myb–dependent effects. To address both of these issues, we have now used two independent dominant-interfering constructs to reduce c-Myb–dependent transcription in two different VSMC cell lines and have determined the effects of these manipulations on the regulation of releasable intracellular Ca2+ pools.
The first construct encodes a c-Myb protein without a DNA binding domain, but capable of interacting with endogenous c-Myb and ancillary transacting cofactors, to generate inactive heterodimers.30 33 34 The second construct encodes a chimeric protein that targets the inhibitory domain of the Engrailed repressor to promoters of c-Myb–regulated genes.9 Both approaches demonstrate that an ≈50% reduction in the functional c-Myb activity of VSMCs results in an ≈40% decrease in serum-stimulated entry into S phase, an ≈35% suppression of serum-stimulated cell proliferation. Similar decreases in resting [Ca2+]cyt and releasable Ca2+ pools were observed at both G0 and the G1/S interface. The finding that c-Myb–dependent effects on intracellular Ca2+ are not compartmentalized indirectly supports the premise that the regulated Ca2+ flux mechanism(s) is plasmalemmal. However, further experiments specifically examining Ca2+ efflux and influx rates across the plasma membrane will be required to test this hypothesis.
The alterations in cell growth and Ca2+ homeostasis generated by the dominant-interfering c-Myb constructs are likely secondary to suppression of c-Myb activity rather than to inhibition of other intracellular components. The identical results obtained by inhibiting c-Myb at the protein level via formation of inactive heterodimers or by suppressing c-Myb–dependent genes at the transcriptional level via targeting of a strong repressor argues for the specificity of the observed cellular effects. More recently, expression of the Δ5-Myb dominant-negative construct in embryonic fibroblasts derived from homozygous c-myb knockout embryos, compared with embryonic fibroblasts from wt embryos, has revealed that alterations in Ca2+ homeostasis require the presence of endogenous c-Myb (K. Bein, M. Husain, M. Mucenski, R.D. Rosenberg, M. Simons, unpublished data, 1996). This observation strongly supports the specificity of the Δ5-Myb construct.
The data indicate that c-Myb exerts a direct effect on Ca2+ homeostasis rather than an indirect action via modulation of cell cycle progression. On the one hand, the reduction of c-Myb activity during growth arrest alters [Ca2+]cyt and pharmacologically elicited Ca2+ responses to the same extent as observed at the G1/S interface. On the other hand, acute inhibition of c-Myb–dependent gene expression in late G1 generates the same magnitude of changes in these Ca2+ parameters as does suppression of proto-oncogene function early in the cell cycle. Potential experimental artifacts have been minimized by using different promoters to express the two dominant-interfering constructs and by using different stable clonal cell lines as well as transiently transfected heterogeneous cell populations with differing insertion sites.
It remains conceivable that c-Myb independently regulates the cell cycle machinery and Ca2+ homeostatic mechanisms. However, four lines of evidence support the notion that the c-Myb–dependent regulation of free and stored intracellular Ca2+ is critically involved in VSMC cell cycle progression. First, the removal of extracellular Ca2+ significantly decreases the numbers of proliferating VSMCs, which can be reversed by replacement of extracellular Ca2+.44 Second, the late G1 block induced by antisense c-myb oligonucleotides in VSMCs is overcome by elevation of [Ca2+]cyt through the use of calcium ionophore.21 Third, growing evidence suggests that the sizes of releasable Ca2+ pools are likely to regulate cell cycle progression in VSMCs.17 18 Indeed, depletion of releasable Ca2+ stores by Tg exposure, without corresponding changes in [Ca2+]cyt, leads to growth arrest of VSMCs until refilling of Ca2+ stores has occurred.18 Fourth, the present investigation further supports the relationship between Ca2+ homeostasis and the cell cycle machinery by demonstrating that certain releasable Ca2+ pools, like resting [Ca2+]cyt, increase during normal G0 to G1/S cell cycle progression and that reductions in c-Myb activity abolish both of these changes and reduce entry into S phase.
The results presented above were mainly generated in a VSMC line (SVE) immortalized with SV40 large-T antigen.36 Because this reagent is known to disturb normal cell cycle inhibition, its presence likely diminished the probability of observing a growth-retarding phenotype. It should also be noted that unlike freshly dispersed contractile VSMCs, in which caffeine-evoked Ca2+ responses usually exceed those obtained with Tg, our cell populations exhibit the reverse. Such phenotypic adaptations to tissue culture may limit the extrapolation of our data to native cells. However, previous observations on c-Myb–dependent alterations in cell growth and/or Ca2+ homeostasis in primary rat VSMCs5 10 11 21 and the replication of key findings of the present study in the A10 SMC line suggest broad applicability of our findings. Indeed, remarkably similar changes in c-Myb–dependent cell growth and Ca2+ homeostasis have also been documented in fibroblasts transfected with the same dominant-interfering constructs (K. Bein, M. Husain, M. Mucenski, R.D. Rosenberg, M. Simons, unpublished data, 1996).
In conclusion, this investigation has specifically examined the role of c-Myb in regulating cell cycle progression and Ca2+ homeostasis of cultured VSMCs. The results demonstrate that functional levels of c-Myb regulate (1) the G1/S transition, (2) the resting [Ca2+]cyt, and (3) the magnitude of Tg- and caffeine-sensitive Ca2+ stores. By using the above experimental system, future studies should allow for the identification of which of the known regulators of [Ca2+]cyt and releasable Ca2+ stores are mediating c-Myb–dependent effects. Further elucidation of the c-Myb Ca2+ cell cycle pathway may be critical for the development of novel molecular treatments for a variety of vascular diseases.
Selected Abbreviations and Acronyms
|AS-c-myb||=||antisense c-myb oligonucleotide(s)|
|AS-IGF1R||=||antisense IGF-1 receptor oligonucleotide(s)|
|[Ca2+]cyt||=||cytoplasmic Ca2+ concentration|
|GRE||=||glucocorticoid response element|
|IGF||=||insulin-like growth factor|
|Rmax, Rmin||=||maximum and minimum fluorescence ratio (R)|
|SERCA||=||sarcoplasmic/endoplasmic reticulum Ca2+-ATPase|
|SV40||=||simian virus 40|
|VSMC||=||vascular smooth muscle cell|
This study was supported in part by National Institutes of Health grants DK-43495, DK-51059, and DK-34854 to Dr Alper and HL-53793 to Dr Simons, American Heart Association Grant-in-Aid 95007560 to Dr Simons, and a grant from the National Foundation for Cancer Research to Dr Rosenberg. Dr Alper is an Established Investigator of the American Heart Association. Dr Husain is the recipient of a Medical Research Council of Canada Clinician Scientist Award. We thank Hamid Akbarali for helpful discussions during this project.
- Received September 30, 1996.
- Accepted February 5, 1997.
- © 1997 American Heart Association, Inc.
Reilly CF, Kindy MS, Brown KE, Rosenberg RD, Sonenshein GE. Heparin prevents vascular smooth muscle cell progression through the G1 phase of the cell cycle. J Biol Chem. 1989;264:6990-6995.
Brown KE, Kindy MS, Sonenshein GE. Expression of the c-myb proto-oncogene in bovine vascular smooth muscle cells. J Biol Chem. 1992;267:4625-4630.
Gerwitz AM, Calabretta B. A c-myb antisense oligodeoxynucleotide inhibits normal human heamtopoiesis in vitro. Science. 1988;242:1303-1306.
Anfossi G, Gerwitz M, Calabretta B. An oligomer complementary to c-myb-encoded mRNA inhibits proliferation of human myeloid leukemia cell lines. Proc Natl Acad Sci U S A. 1989;86:3379-3383.
Badiani P, Corbella P, Kioussis D, Marvel J, Weston K. Dominant interfering alleles define a role for c-Myb in T-cell development. Genes Dev. 1994;8:770-782.
Simons M, Rosenberg RD. Antisense nonmuscle myosin heavy chain and c-myb oligonucleotides suppress smooth muscle cell proliferation in vitro. Circ Res. 1992;70:835-843.
Cogswell JP, Cogswell PC, Kuehl WM, Cuddihy AM, Bender TM, Engelke U, Marcu KB, Ting JP. Mechanism of c-myc regulation by c-Myb in different cell lineages. Mol Cell Biol. 1993;13:2858-2869.
Ku DH, Wen SC, Engelhard A, Nicolaides NC, Lipson KE, Marino TA, Calabretta B. c-myb transactivates cdc2 expression via Myb binding sites in the 5′-flanking region of the human cdc2 gene [published erratum appears in J Biol Chem. 1993;268:13010]. J Biol Chem. 1993;268:2255-2259.
Venturelli D, Travali S, Calabretta B. Inhibition of T-cell proliferation by a MYB antisense oligomer is accompanied by selective down-regulation of DNA polymerase alpha expression. Proc Natl Acad Sci U S A. 1990;87:5963-5967.
Ghosh TK, Bian JH, Short AD, Rybak SL, Gill DL. Persistent intracellular calcium pool depletion by thapsigargin and its influence on cell growth. J Biol Chem. 1991;266:24690-24697.
Short AD, Bian J, Ghosh TK, Waldron RT, Rybak SL, Gill DL. Intracellular Ca2+ pool content is linked to control of cell growth. Proc Natl Acad Sci U S A. 1993;90:4986-4990.
Waldron RT, Short AD, Meadows JJ, Ghosh TK, Gill DL. Endoplasmic reticulum calcium pump expression and control of cell growth. J Biol Chem. 1994;269:11927-11933.
Waldron RT, Short AD, Gill DL. Thapsigargin-resistant intracellular calcium pumps: role in calcium pool function and growth of thapsigargin-resistant cells [published erratum appears in J Biol Chem. 1995;270:18710]. J Biol Chem. 1995;270:11955-11961.
Simons M, Morgan KG, Parker C, Collins E, Rosenberg RD. The proto-oncogene c-myb mediates an intracellular calcium rise during the late G1 phase of the cell cycle [published erratum appears in J Biol Chem. 1993;268:16082]. J Biol Chem. 1993;268:627-632.
Simons M, Hideao A, Salzman E, Rosenberg R. c-Myb affects intracellular calcium handling in vascular smooth muscle cells. Am J Physiol. 1995;37:C856-C868.
Burgess T, Fisher E, Ross S, Bready J, Qian Y-X, Bayewitch L, Cohen A, Herrera C, Hu S-F, Kramer T, Lott F, Martin F, Pierce G, Simonet L, Farrell C. The antiproliferative activity of c-myb and c-myc antisense oligonucleotides in smooth muscle cells is caused by a nonantisense mechanism. Proc Natl Acad Sci U S A. 1995;92:4051-4055.
Guvakova M, Yakubov L, Vlodavsky I, Tonkinson J, Stein C. Phosphothioate oligodeoxynucleotides bind to basic fibroblast growth factor, inhibit its binding to cell surface receptors, and remove it from low affinity binding sites on extracellular matrix. J Biol Chem. 1995;270:2620-2627.
Villa AE, Guzman LA, Poptic EJ, Labhasetwar V, D’Souza S, Farrell CL, Plow EF, Levy RJ, DiCorleto PE, Topol EJ. Effects of antisense c-myb oligonucleotides on vascular smooth muscle cell proliferation and response to vessel wall injury. Circ Res. 1995;76:505-513.
Simons M, Rosenberg RD, Edelman ER. … and surreal antisense? Nat Med. 1996;2:131-132. Letter.
Nakagoshi H, Nagase T, Kanei-Ishii C, Ueno Y, Ishii S. Binding of the c-myb proto-oncogene product to the simian virus 40 enhancer stimulates transcription. J Biol Chem. 1990;265:3479-3483.
Nomura T, Sakai N, Sarai A, Sudo T, Kanei-Ishii C, Ramsay RG, Favier D, Gonda TJ, Ishii S. Negative autoregulation of c-Myb activity by homodimer formation through the leucine zipper. J Biol Chem. 1993;268:21914-21923.
Ogata K, Hojo H, Aimoto S, Nakai T, Nakamura H, Sarai A, Ishii S, Nishimura Y. Solution structure of a DNA-binding unit of Myb: a helix-turn-helix-related motif with conserved tryptophans forming a hydrophobic core. Proc Natl Acad Sci U S A. 1992;89:6428-6432.
Kanei-Ishii C, Macmillan EM, Nomura T, Sarai A, Ramsay RG, Aimoto S, Ishii S, Gonda TJ. Transactivation and transformation by Myb are negatively regulated by a leucine-zipper structure. Proc Natl Acad Sci U S A. 1992;89:3088-3092.
Kanei-Ishii C, Yasukawa T, Morimoto RI, Ishii S. c-Myb-induced trans-activation mediated by heat shock elements without sequence-specific DNA binding of c-Myb. J Biol Chem. 1994;269:15768-15775.
Weiler-Guettler H, Aird WC, Husain M, Rayburn H, Rosenberg RD. Targeting of transgene expression to the vascular endothelium of mice by homologous recombination at the thrombomodulin locus. Circ Res. 1996;78:180-187.
Jiang L, Stuart Tilley A, Parkash J, Alper SA. pHi and serum regulate AE2-mediated Cl−/HCO3− exchange in CHOP cells of defined transient transfection status. Am J Physiol. 1994;36:C845-C856.
Grynkiewicz G, Poenie M, Tsien RV. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem. 1985;260:3440-3450.
Kojima I, Matsunaga H, Kurokawa K, Ogata E, Nishimoto I. Calcium influx: an intracellular message of the mitogenic action of insulin-like growth factor-I. J Biol Chem. 1988;263:16561-16567.
Pietrzkowski Z, Sell C, Lammers R, Ullrich A, Baserga R. Roles of insulinlike growth factor 1 (IGF-1) and the IGF-1 receptor in epidermal growth factor-stimulated growth of 3T3 cells. Mol Cell Biol. 1992;12:3883-3889.
Kojima I, Mogami H, Shibata H, Ogata E. Role of calcium entry and protein kinase C in the progression activity of insulin-like growth factor-I in Balb/c 3T3 cells. J Biol Chem. 1993;268:10003-10006.