Modulation of Calmodulin Gene Expression as a Novel Mechanism for Growth Hormone Feedback Control by Insulin-like Growth Factor in Grass Car
http://www.100md.com
《内分泌学杂志》
Department of Zoology, University of Hong Kong, Hong Kong, Peoples Republic of China
Abstract
Calmodulin (CaM), the Ca2+ sensor in living cells, is essential for biological functions mediated by Ca2+-dependent mechanisms. However, modulation of CaM gene expression at the pituitary level as a means to regulate pituitary hormone synthesis has not been characterized. In this study we examined the functional role of CaM in the feedback control of GH by IGF using grass carp pituitary cells as a cell model. To establish the structural identity of CaM expressed in the grass carp, a CaM cDNA, CaM-L, was isolated from the carp pituitary using 3'/5' rapid amplification of cDNA ends. The open reading frame of this cDNA encodes a 149-amino acid protein sharing the same primary structure with CaMs reported in mammals, birds, and amphibians. This CaM cDNA is phylogenetically related to the CaM I gene family, and its transcripts are ubiquitously expressed in the grass carp. In carp pituitary cells, IGF-I and IGF-II induced CaM mRNA expression with a concurrent drop in GH transcript levels. These stimulatory effects on CaM mRNA levels were not mimicked by insulin and appeared to be a direct consequence of IGF activation of CaM gene transcription without altering CaM transcript stability. CaM antagonism and inactivation of calcineurin blocked the inhibitory effects of IGF-I and IGF-II on GH gene expression, and CaM overexpression also suppressed the 5' promoter activity of the grass carp GH gene. These results, as a whole, provide evidence for the first time that IGF feedback on GH gene expression is mediated by activation of CaM gene expression at the pituitary level.
Introduction
CALMODULIN (CaM) IS a Ca2+-binding protein serving as an intracellular Ca2+ sensor in living cells. Upon Ca2+ binding, conformational changes occur in the CaM molecule with the exposure of hydrophobic domains for subsequent association with CaM target proteins (1). The interactions of CaM with its target proteins activate a variety of cellular processes, including gene expression (2), protein synthesis (3), hormone secretion (4, 5), cell motility (6), apoptosis (7, 8), and cell proliferation (9, 10). In mammals, a high level of CaM expression was noted in the brain-pituitary axis (11, 12). This finding is consistent with the reports that CaM is involved in the synthesis and/or secretion of hypothalamic factors [e.g. somatostatin (SRIF) (13) and GHRH (14)] as well as pituitary hormones [e.g. prolactin (PRL) (4, 15) and LH (16)]. In GH3 cells, modulation of GH release by CaM has been implicated (5). In this case, CaM antagonists, by altering phosphodiesterase activity, can induce a biphasic effect on GH release, being inhibitory at low doses and stimulatory at high doses. Although CaM can activate PRL gene expression both in vivo (17) and in vitro (18) by elevating PRL promoter activity through a proximal enhancer element (15), CaM antagonism has little or no effect on GH gene expression in these reports. Given that functional studies of CaM production have not been fully characterized, direct evidence for modulation of CaM gene expression at the pituitary level as a means to regulate pituitary functions is still lacking.
IGF, a polypeptide with structural similarity to proinsulin, is produced mainly in the liver under the stimulatory influence of GH. In general, IGF serves as a mediator for GH actions in regulating somatic growth, body metabolism, and cell proliferation and differentiation (19, 20). IGF can also exert a long-loop feedback to inhibit GH synthesis and secretion (21), and this negative feedback is well documented in humans (22) and nonprimate mammals (23, 24). The site of action for IGF long-loop feedback appears to be species specific. Direct actions of IGF at the pituitary level without a hypothalamic component have been reported in sheep (25), whereas central actions to modify GHRH and SRIF secretion from the hypothalamus have been clearly demonstrated in the rat (26). In somatotroph cell lines, e.g. MtT/S cells, IGF inhibits GH gene expression by reducing GH promoter activity through phosphotidylinositol 3-kinase-mediated signaling events (24). Although no information is available to date in lower vertebrates on IGF regulation of GH-releasing factors at the hypothalamic level, the long-loop feedback by IGF appears to be conserved in fish. This idea is supported by the findings that 1) the molecular structure of IGFs (including IGF-I and IGF-II) is highly conserved from fish to mammals (27); 2) GH stimulates IGF production in the liver of fish species (28, 29); and 3) IGF inhibits GH secretion (30, 31) and synthesis (30, 31) in fish pituitary cells in vitro. Recently, IGF has been shown to induce LH and PRL release in Coho salmon (32) and striped bass (33), respectively. These findings suggest that IGF may also act on pituitary cells other than somatotrophs in lower vertebrates.
In fish models, e.g. goldfish, the availability of extracellular Ca2+ and its entry via voltage-sensitive Ca2+ channels are essential for both basal and stimulated GH release (34). In goldfish pituitary cells, GH secretion induced by GH-releasing factors, e.g. GnRH and dopamine, can be blocked by CaM antagonists and CaM kinase II inhibitors (35, 36). These results indicate that CaM and its downstream signaling pathways are involved in the GH-releasing actions of these GH regulators. At present, however, it is still unclear whether GH gene expression can be regulated by CaM at the pituitary level, and to our knowledge, the functional role of CaM in GH feedback by IGF has not been previously examined. In this study, using grass carp as a model, we tested the hypothesis that IGF inhibits GH gene expression directly at the pituitary level via modulation of CaM gene expression. As a first step, the structural identity of grass carp CaM was established by molecular cloning using 5'/3' rapid amplification of cDNA ends (RACE). The expression profile of CaM mRNA was examined in various tissues of the grass carp by Northern blot. Based on the CaM sequence obtained, a slot-blot assay was set up for CaM mRNA measurement. Using this assay system, the effects of IGF-I and IGF-II on CaM mRNA expression were tested in grass carp pituitary cells and correlated with the corresponding changes in GH mRNA levels. In parallel experiments, the effects of IGF-I and IGF-II on CaM transcript stability, CaM primary transcript expression, and 5' promoter activity of the CaM gene were also examined. The functional role of CaM in GH gene expression was also evaluated by testing the effects of CaM antagonism on GH mRNA levels and the effects of CaM overexpression on GH promoter activity.
Materials and Methods
Animals
One-year old (1+) Chinese grass carp (Ctenopharyngodon idellus) were purchased from the local markets in Hong Kong and housed in 200-liter aquaria at 18 ± 2 C for at least 3 d before tissue collection and/or pituitary cell preparation. On the day of the experiments, the fish were killed by anesthesia in MS222 (tricaine methanesulfonate; Sigma, St. Louis, MO), followed by spinosectomy according to the regulations for animal use at University of Hong Kong.
Reagents and test substances
Human IGF-I and IGF-II were purchased from Sigma-Aldrich Corp. (St. Louis, MO). TRIzol, medium 199, Opti-MEM, Ham’s F-10, and fetal bovine serum (FBS) were obtained from Invitrogen Life Technologies, Inc. (Grand Island, NY). Calmidazolium and human insulin were acquired from Sigma-Aldrich Corp., and actinomycin D was obtained from Calbiochem (La Jolla, CA). Insulin was first dissolved in double-distilled deionized water to prepare a working stock with an activity level of 109 μU/ml. The stock solution was then aliquoted into small volume and stored frozen at –80 C until used. IGF-I and IGF-II were prepared in a similar manner, except that these peptide hormones were dissolved in 10 nM HCl to give a stock concentration at 100 μM. Given that calmidazolium has a low solubility in aqueous solution, the stock solution at 10 mM was prepared in dimethylsulfoxide. On the day of experiments, test substances were diluted to appropriate concentrations with culture medium 20 min before drug treatment. In these studies, the final dilutions of dimethylsulfoxide and HCl were always 0.1% (vol/vol) or less of the original levels and did not affect basal expression of CaM and GH mRNA in grass carp pituitary cells.
Molecular cloning of grass carp CaM cDNA
Total RNA (5 μg) was extracted from grass carp pituitaries using TRIzol (Invitrogen Life Technologies, Inc.), and first strand cDNA was synthesized using a SuperScript II first strand synthesis kit (Invitrogen Life Technologies, Inc.). A pair of primers designed based on the conserved regions of mammalian CaM cDNAs were used to isolate a 463-bp fragment of grass carp CaM cDNA by PCR using the first strand cDNA as the template. Based on the sequence obtained, new primers were designed for 5'/3' RACE using a GeneRacer kit (Invitrogen Life Technologies, Inc., Carlsbad, CA) with pituitary total RNA (5 μg) as the template. PCRs were conducted according to the instructions of the manufacturer. PCR products were gel-purified and subcloned into the pGEM-T Easy vector (Promega Corp., Madison, WI) for DNA sequencing using a BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA). The full-length cDNA of CaM (GenBank no. AY627883) was reconstructed from its 5' and 3' sequences using MacVector 6.5 (Oxford Molecular, Madison, WI) and MacDNASIS PRO 3.5 programs (Hitachi, San Bruno, CA). After that, a digoxigenin (DIG)-labeled cDNA probe for grass carp CaM mRNA was prepared for subsequent Northern blot and slot-blot assays using a PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany).
Northern blot of CaM transcripts
Total RNA (25 μg) was extracted with TRIzol from selected tissues of the grass carp, including brain, pituitary, muscle, liver, kidney, spleen, heart, intestine, and gills. After that, mRNA was purified using a PolyATract mRNA Isolation System (Promega Corp.). These mRNA samples were denatured, size fractionated in 1% agarose gel, and transblotted onto a positively charged nylon membrane (Roche) using a VacuGene Vacuum Blotting System (Pharmacia Biotech, Piscataway, NJ). The membrane was then UV cross-linked, prehybridized for 3 h, and incubated with the DIG-labeled CaM cDNA probe for 15 h at 42 C. On the following day, the membrane was washed twice at 68 C in 0.5x standard saline citrate with 0.1% sodium dodecyl sulfate, and hybridization signals were detected using a DIG Lumin Detection Kit (Roche). In these studies, a Northern blot of -actin was used as an internal control.
Measurement of steady-state CaM and GH mRNA expression
Primary cultures of pituitary cells were prepared from 1-yr-old grass carp by a trypsin/deoxyribonuclease digestion method as described previously (37). The average cell yield was approximately 8 x 106 cells/pituitary, and cell viability was always 95% or greater. After cell dispersion, pituitary cells were cultured overnight (>15 h) with 5% FBS in 24-well culture plates precoated with poly-D-lysine (Sigma-Aldrich Corp.) at a seeding density of about 3 x 106 cells/ml/well. After that, test substances at appropriate concentrations were prepared in carp MEM (38) and gently overlaid onto pituitary cells after removal of old culture medium. For dose-response studies, the duration of drug treatment was routinely fixed at 48 h. For the evaluation of CaM transcript stability, cells were pretreated with actinomycin D (8 μM) and incubated with or without IGF treatment for the duration as indicated in individual experiments. After drug treatment, total RNA (2.5 μg) was extracted with TRIzol, heat-denatured at 70 C for 15 min, and vacuum-blotted onto nylon membranes in duplicate using a Bio-Dot microfiltration unit (Bio-Rad Laboratories, Hercules, CA). The membranes were UV cross-linked, prehybridized for 3 h, and hybridized overnight with DIG-labeled CaM probe and grass carp GH cDNA probe (38), respectively. After hybridization, the membranes were washed, and signal development was carried out as described for the Northern blot. Measurement of hybridization signals was conducted in a Kodak 440 Image Station (Kodak Digital Science, Rochester, NY). In this study, slot blots of 18S RNA were also conducted to serve as an internal control.
Real-time PCR of mature CaM mRNA and CaM primary transcript
Based on the sequence of the newly cloned CaM cDNA, intron trapping was performed using genomic DNA as a template to pull out introns 1–5 of the grass carp CaM gene (GenBank no. AY656698). Primers flanking the junction between intron 4 and exon 5 from positions 10,408–10,789 were used for real-time PCR of CaM primary transcripts. To serve as a parallel control, real-time PCR of mature CaM mRNA was also performed using primers flanking intron 4 from position 10,209 (in exon 4) to 10,789 (in exon 5). In this study, pituitary cells were incubated for 48 h with increasing levels of IGF-I and IGF-II. Total RNA (5 μg) was isolated and digested with deoxyribonuclease I to remove genomic DNA contamination. After that, RT was carried out, and RT samples were subjected to PCR using a LightCycler-DNA Master SYBR Green I Kit (Roche) in a RotorGene 2000 Real-Time PCR System (Corbett Research, Eight Mile Plains, Australia). PCRs for CaM mRNA and primary transcripts were conducted for 35 cycles at 94 C for 30 sec for denaturing, 69 C for 30 sec for annealing, and 72 C for 30 sec for primer extension. Based on our validation, the primers for CaM primary transcripts were ineffective in PCR amplification of mature CaM mRNA. Furthermore, PCR of CaM primary transcripts in the nuclear fraction of grass carp pituitary cells consistently produced a 381-bp PCR product with a melting temperature (Tm) at 92.1 C, whereas the mature CaM mRNA in the cytosolic fraction would generate a 181-bp PCR product with Tm at 89.8 C. In these experiments, serial dilutions of plasmid DNA with CaM cDNA or the amplicon covering the junction between intron 4 and exon 5 of the CaM gene were used as the standards for real-time PCR of mature CaM mRNA and CaM primary transcripts, respectively.
Measurement of CaM promoter activity
Based on the sequence of grass carp CaM cDNA, nested primers covering the 5'-untranslated region (5'UTR) were designed to pull out the 5' promoter of CaM gene (GenBank no. AY656698) using a Universal GenomeWalker Kit (BD Clontech, Palo Alto, CA). Genomic DNA prepared from whole blood was used as the template, and PCRs were conducted according to the instructions of the manufacturer. The 5' promoter obtained (–1369 to +49) was subcloned into the NheI and XhoI sites of the pGL3.Basic vector (Promega Corp.) to generate the reporter construct pCaM(–1369).luciferase (Luc). This luciferase-expressing construct was used to examine IGF action on CaM promoter activity using the mouse pituitary cell line T3-1 as the host cells. The cells were maintained in DMEM with 44 mM NaHCO3, 10% FBS, and 1% (vol/vol) antibiotics-antimycotics (pH 7.4). Transfection was conducted for 6 h in 0.5 ml Opti-MEM with 1.16 μl Lipofectamine (Invitrogen Life Technologies, Inc.), 0.05 μg pCaM(–1369).Luc, 0.01 μg pEGFP-N1 (EGFP, enhanced green fluorescence protein), and 0.44 μg pBluescript (Stratagene) as carrier DNA. After transfection, T3-1 cells were cultured in DMEM with 10% FBS for 6 h, followed by serum starvation for 6 h before the initiation of IGF treatment. Based on our validation, the optimal duration of drug treatment was fixed at 12 h. After IGF treatment, T3-1 cells were dissolved in lysis buffer (Promega Corp.), and the lysate was cleared by centrifugation at 20,000 x g at 4 C for 15 min. For measurement of Luc activity, a 50-μl volume of cleared lysate was mixed with 100 μl luciferase assay reagent (Promega Corp.), and luminescence signal was quantified using a Lumat LB9507 luminometer (EG&G, Gaithersburg, MD). In these experiments, cotransfection with pEGFP-N1 was used as an internal control, and GFP expression was monitored by measuring the fluorescence signal in 100 μl lysate using a Cytofluor series 4000 Multiple-Well Plate Reader (Applied Biosystems, Foster City, CA).
CaM overexpression on GH promoter activity
To examine the functional role of CaM in GH gene transcription, the coding sequence of grass carp CaM cDNA was PCR isolated and subcloned into the expression vector pcDNA3.1 to produce pcDNA.CaM. The effects of CaM overexpression were examined with cotransfection of pGH.Luc constructs (39) carrying decreasing lengths (–986 to –115) of the grass carp GH gene promoter (GenBank no. X30419) using the rat pituitary cell line GH4C1 as the host cells. The cells were maintained in Ham’s F-10 with 14 mM NaHCO3, 10% FBS, and 1% (vol/vol) antibiotics-antimycotics (pH 7.4). Transfection was conducted for 6 h in 0.5 ml Opti-MEM with 1.16 μl Lipofectamine, 0.18 μg pGH.Luc, 0.01 μg pEGFP-N1, and 0.05–0.25 μg pcDNA.CaM (with pBluescript as carrier DNA). After transfection, GH4C1 cells were cultured in Ham’s F-10 for 18 h, dissolved in lysis buffer, and assayed for luciferase activity and GFP expression as described in the preceding section. To confirm CaM overexpression at the protein level, Western blot was also performed in GH4C1 cells after pcDNA.CaM transfection. Cell lysate was prepared by three cycles of repeated freezing and thawing in phenylmethylsulfonylfluoride (150 μg/ml) with EDTA (1.5 mM), cleared by centrifugation, and resolved in a 15% gel by SDS-PAGE for 1 h at 200 V. After that, protein samples were transferred onto a polyvinylidene difluoride membrane by low current electroblotting at 65 V for 2.5 h. The membrane was then subjected to blocking with 3% BSA and incubation with sheep antibovine CaM antibody (1:1000; Chemicon International, Temecula, CA) for 2 h. After washing, rabbit antisheep IgG-horseradish peroxidase (1:5000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was added, and signal development was carried out in a Kodak 440 Image Station with SuperSignal WestPico (Pierce Chemical Co., Rockford, IL) as the substrate. Using a similar approach, Western blot was also performed in carp pituitary cells to examine the effects of IGF treatment on CaM expression at the protein level. In this case, parallel probing of -actin using an Actin Ab-1 Kit (Calbiochem) was also conducted to serve as an internal control.
Data transformation and statistical analysis
For quantitation of CaM gene expression, CaM mRNA levels were measured in terms of arbitrary density unit and normalized against 18S rRNA expressed in the same sample. Given that no significant changes in 18S rRNA levels were noted, these data were simply transformed into a percentage of the mean value in the control group for statistical analysis (as percentage of control). The transformation was conducted to allow for data pooling from separate experiments without increasing the overall variability of the results. The half-life (t1/2) of CaM mRNA, defined as the time required for CaM mRNA to decrease to half its original level, was deduced by the one-phase exponential decay model using PRISM 3.02 (GraphPad, Inc., San Diego, CA). For real-time PCR, a standard curve with a dynamic range of 105 or greater and a correlation coefficient of 0.95 or more were used for data calibration. Mature CaM mRNA and primary transcripts were measured in terms of femtomoles per million cells and transformed as a percentage of the control for statistical analysis. For luciferase activity measurement, the raw data in arbitrary light units were normalized against the level of GFP expression in the same sample expressed in arbitrary fluorescence units. These data were expressed as the ratio of arbitrary light units/arbitrary fluorescence units or transformed as a percentage of the control as described in the preceding section. The transformed data for CaM mRNA and luciferase activity were analyzed by Student’s t test or ANOVA (two-way), followed by Fisher’s least significance difference (LSD) test. Differences were considered significant at P < 0.05.
Results
Molecular cloning of grass carp CaM cDNA
Using RT-PCR coupled to 5'/3' RACE, a full-length CaM cDNA, CaM-L, with a size of 1.55 kb was isolated from the grass carp pituitary (Fig. 1). In parallel studies, a shorter version of CaM cDNA, CaM-S, with a size of 0.73 kb was also obtained with CaM-L using RT samples prepared from the brain, gills, heart, and kidney. Nucleotide sequence analysis revealed that CaM-S is a truncated form of CaM-L with a shorter 3'UTR probably produced by differential use of polyadenylation signals. The 447-bp ORF of these newly cloned CaM cDNAs encodes a 149-amino acid (a.a.) protein with primary sequence identical with the CaMs reported in mammals (e.g. human, bovine, mouse, and rat), birds (e.g. chicken and duck), and amphibian (e.g. Xenopus) as well as in the representative species of modern bony fish (e.g. medaka and perch). Compared with CaMs reported in early evolved bony fish (e.g. eel), cyclostome (e.g. hagfish), urochordate (e.g. ciona), cephalochordate (e.g. brachiostoma), mollusca (e.g. aplysia), nematode (e.g. C. elegans), porifera (e.g. sponge), and protozoa (e.g. trypanosome), only a low level of a.a. substitutions were noted, and most of them were conserved mutations (Fig. 2). Sequence alignment at the a.a. level also revealed that the four helix-loop-helix EF hands (EF-I to EF-IV), the functional motifs of CaM for Ca2+ binding, are structurally conserved in vertebrates from modern bony fish to mammals. Although the structure of CaM is highly conserved, multiple copies of CaM genes, namely CaM I, CaM II, and CaM III, have been reported in birds and mammals (40, 41). Using unrooted analysis by PHYLIP (version 3.6), grass carp CaM can be clustered in the same clade with the CaM I genes reported in the rat, mouse, human, and chicken and is distally related to CaM II and CaM III gene subfamilies (Bootstrap values, 998–1000).
Tissue distribution of CaM transcripts
A DIG-labeled cDNA probe covering the ORF of CaM-L and -S was used to examine the tissue distribution of CaM expression by Northern blot (Fig. 3). mRNA samples were prepared from various tissues of the grass carp, including spleen, pituitary, muscle, liver, gut, kidney, heart, gills, and brain. The highest levels of CaM transcripts were noted in brain and gut; to a lesser extent in spleen, muscle, and pituitary; and to a lower level in heart, kidney, liver, and gills. In this study, three CaM transcripts (1.3, 1.7, and 3.0 kb, respectively) were identified, and all of them were expressed in brain, spleen, kidney, heart, intestine, and muscle. In the gills, only the 1.3- and 1.7-kb, not the 3.0-kb, transcripts were detected. Interestingly, the 3.0- and 1.7-kb, but not the 1.3-kb, transcripts were expressed in the liver. In the case of the pituitary, however, only the 1.7-kb transcript could be recognized. These results, as a whole, indicate that different forms of CaM mRNA are expressed in the grass carp in a tissue-specific manner.
Regulation of CaM mRNA expression in grass carp pituitary cells
Because Northern blot using purified mRNA is not ideal for quantitative studies with a large number of samples, a slot-blot assay using total RNA was established to study steady-state CaM mRNA expression in grass carp pituitary cells. To examine IGF effects on CaM gene expression at the pituitary level, pituitary cells were incubated for 48 h with increasing doses (0.1–100 nM) of IGF-I (Fig. 4A) and IGF-II (Fig. 4B). In these experiments, IGF-I and IGF-II were both effective in elevating CaM mRNA levels in a dose-dependent manner, with a concurrent drop in GH mRNA expression. The differential effects of IGFs on CaM and GH mRNA expression, however, were not mimicked by increasing doses of insulin (102–106 μU/ml). Unlike IGFs, insulin inhibited CaM mRNA levels (Fig. 5A), with a concurrent rise in GH mRNA expression (Fig. 5B). In these experiments, a negative correlation between CaM and GH mRNA levels was noted after IGF and insulin treatment (Fig. 5B), suggesting that modulation of CaM expression at the pituitary level may be involved in GH gene regulation. Because multiple transcripts of CaM are expressed in the grass carp, the possibility of IGF induction of CaM gene expression via differential expression of various forms of CaM mRNA was examined by Northern blot (Fig. 6A). In this case, IGF-I (100 nM) and IGF-II (100 nM) significantly increased the expression level of the 1.7-kb CaM mRNA in grass carp pituitary cells (Fig. 6B). The 1.3- and 3.0-kb CaM transcripts, however, remained undetectable in this study. Using Western blot, a time-dependent increase in CaM protein content was noted from 12–48 h in carp pituitary cells after the initiation of IGF-I treatment (100 nM; Fig. 6C). The protein levels of -actin, the internal control of these experiments, remained unchanged during the period of drug treatment. Similar results were also obtained using IGF-II (100 nM) as the stimulant for CaM expression (data not shown).
To evaluate the role of CaM in IGF inhibition of GH gene expression, spatial and temporal correlations between CaM and GH gene expression after IGF treatment were examined to test whether 1) the opposite changes in CaM and GH mRNA levels indeed occurred in the same cell type (i.e. in carp somatotrophs), and 2) the response in CaM mRNA expression was initiated before and/or during the corresponding changes in GH mRNA levels. For the study of spatial correlation, enriched somatotrophs were prepared from mixed populations of pituitary cells using Percoll gradient centrifugation (Fig. 7A). After a 48-h incubation with IGF-I (100 nM) and IGF-II (100 nM), CaM mRNA levels were increased in this somatotroph preparation, with a concurrent drop in GH mRNA expression (Fig. 7, B and C). In parallel time-course studies, basal levels of CaM mRNA were significantly increased (P < 0.05) in carp pituitary cells after 6-h incubation with IGF-I (100 nM) and IGF-II (100 nM), respectively, and remained elevated from 12–48 h (Fig. 8A). The corresponding drop in GH mRNA levels (P < 0.05), however, was noted only after 24-h incubation with IGF-I and IGF-II. After that, GH mRNA levels were reduced gradually from 24 to 48 h in a time-dependent manner (Fig. 8B). To establish the functional role of CaM in GH feedback by IGF, the effects of calmidazolium, a CaM antagonist, on both basal and IGF-inhibited GH mRNA expression were tested at the pituitary level. Calmidazolium is known to bind CaM tightly with nanomolar affinity to perturb a wide range of Ca2+-dependent mechanisms by immobilizing the methionine residues in the Ca2+-binding sites of helix-loop-helix EF motifs (42). In this case, GH mRNA levels were elevated in a dose-related fashion by increasing doses of calmidazolium (0.1 nM to 1 μM; Fig. 9A). In the presence of calmidazolium (5 nM), the inhibitory actions of IGF-I (100 nM) and IGF-II (100 nM) on GH mRNA expression were totally abolished (Fig. 9B). To elucidate the downstream signaling events that occurred after CaM activation, the functional roles of CaM kinase II and calcineurin in IGF inhibition of GH mRNA expression were also examined (Fig. 9C). Similar to calmidazolium, a 48-h incubation with the calcineurin inhibitor cyclosporin A (100 nM) increased basal levels of GH mRNA and blocked the inhibitory effect of IGF-I (100 nM) on GH transcript expression. The CaM kinase II inhibitor KN62 (5 μM), in contrast, reduced GH mRNA levels in grass carp pituitary cells and did not modify the inhibitory action of IGF-I (100 nM) on GH gene expression. Similar results were obtained by substituting IGF-II for IGF-I in these experiments (data not shown).
Molecular mechanisms for IGF induction of CaM gene expression
To shed light on the molecular mechanisms for IGF induction of CaM gene expression, the effects of IGF on CaM gene transcription and CaM transcript stability were examined. As a first step, real-time PCR of CaM primary transcripts was performed in grass carp pituitary cells after a 48-h incubation with increasing doses (0.1–100 nM) of IGF-I and IGF-II. As a parallel control, mature CaM mRNA levels were also monitored in these experiments. Calibration of primary transcript levels was conducted by linear regression of critical thresholds of the standard curve (Fig. 10A). In the same experiment, a separate standard curve was used for calibration for mature CaM mRNA (data not shown). In this study, IGF-I and IGF-II consistently induced a dose-dependent increase in CaM primary transcripts and mature mRNA (Fig. 10B). The specificity of PCRs was also confirmed by melting curve analysis and ethidium bromide staining of PCR products. In this case, a single 381-bp (Tm,92.1 C) and a single 181-bp PCR product (Tm, 89.8 C) were detected for CaM primary transcripts and mature CaM mRNA, respectively.
To provide additional evidence that IGF increases CaM mRNA levels via activation of CaM gene transcription, a reporter gene approach was used to examine the effects of IGF treatment on CaM promoter activity expressed in T3-1 cells. Compared with the time-matched controls, a significant increase in luciferase activity (P < 0.05) was noted after exposure to IGF-I (100 nM) for 6 h (Fig. 10C). The maximal difference in luciferase activity expressed in the control groups vs. the treatment groups was observed 12 h after the initiation of IGF-I treatment. Therefore, the duration of drug treatment was fixed at 12 h for our dose-response studies with IGF-I and IGF-II, respectively. In this case, increasing levels of IGF-I and IGF-II (0.01–100 nM) were both effective in stimulating luciferase activity expression in a dose-dependent manner (Fig. 10D). Although the maximal stimulatory effects on luciferase activity were similar, IGF-I appeared to be more potent than IGF-II in inducing luciferase expression, especially in subnanomolar doses.
Using grass carp pituitary cells pretreated with the transcription inhibitor actinomycin D (8 μM), the clearance curves of CaM mRNA were constructed in the presence or absence of IGF-I (100 nM; Fig. 10E). In these experiments, the two curves were found to be overlapping, and the t1/2 values deduced for CaM mRNA (73.4 h for the control vs. 73.2 h for IGF treatment) were also similar (P > 0.05). Comparable results were obtained in parallel experiments with IGF-II (data not shown), confirming that IGF-induced CaM gene expression in carp pituitary cells does not involve modification of CaM transcript stability.
CaM overexpression on GH gene promoter activity
The rat somatotroph cell line GH4C1 was used for the promoter studies of grass carp GH gene, because it has endogenous expression of Pit-1, a pituitary-specific transcription factor, to support basal GH promoter activity. Overexpression of CaM at the protein level was tested by Western blot in GH4C1 cells. Compared with the control transfected with the blank vector pcDNA 3.1, a significant increase in CaM immunoreactivity (16.8 kDa) was detected in the cell lysate prepared from the treatment group transfected with pcDNA.CaM (Fig. 11A). In GH4C1 cells cotransfected with pGH(–986).Luc carrying the –986 to +13 promoter of the grass carp GH gene, CaM overexpression was effective in reducing luciferase activity in a time-dependent manner (Fig. 11B). Apparently, this inhibitory effect was specific to the GH promoter, and the expression levels of GFP, the internal control, were not affected by CaM overexpression. In parallel experiments, transfection with increasing levels of pcDNA.CaM (0.05–0.25 μg) also suppressed luciferase activity in GH4C1 cells in a dose-dependent manner (Fig. 11C). In these studies, transfection of pGL3.Control and pGL3.Basic were used as the positive and negative controls, respectively. A high level of luciferase activity was consistently detected in the group transfected with pGL3.Control (with a simian virus 40 promoter) compared with that in the group transfected with the promoterless pGL3.Basic, confirming that the reporter system was appropriate for quantitative analysis of promoter activity. In our validation studies, the blank vector pGL3.Basic was also found to be nonresponsive to CaM overexpression (data not shown). To map out the coarse location of the promoter region responsible for CaM inhibition of GH promoter activity, CaM overexpression was conducted in GH4C1 cells transfected with pGH.Luc constructs carrying decreasing lengths of GH promoter from –986 to –115. In all pGH.Luc constructs examined, basal expression of luciferase activity was significantly suppressed by CaM overexpression (Fig. 11D), suggesting that the CaM-responsive element(s) is located within the proximal promoter region before position –115.
Discussion
In the present study, two CaM cDNAs, CaM-L and CaM-S, have been isolated in the grass carp. CaM-S is a truncated form of CaM-L, and both of them share a common ORF encoding a 149-a.a. protein with primary sequence identical with CaMs in mammals, birds, and amphibians. In mammals, nonallelic CaM genes, namely CaM I, CaM II, and CaM III, encoding the same CaM protein have been reported (40, 41). In general, these isoforms of CaM genes are believed to be the result of gene duplication that occurred 1400 million yr ago (43). Using unrooted analysis, it has been shown that the newly cloned CaM cDNAs are phylogenetically related to the CaM I gene subfamily. In higher vertebrates, the genetic redundancy of CaM genes has been attributed to the important function of CaM as the intracellular Ca2+ sensor in living cells (40). This idea is also in agreement with our findings that CaM transcripts are ubiquitously expressed in the grass carp. In this case, the highest levels of CaM transcripts can be found in the brain and intestine. In the rat, a high level of CaM (up to 0.5% of total protein) can be detected in the brain (44), which may be related to the functions of CaM in modulating neuronal excitability (45, 46) and exocytosis of neurotransmitters (47, 48). In this study, CaM transcripts of 1.3, 1.7, and 3.0 kb have been identified, and the expression pattern of these isoforms appears to be tissue specific. In the case of the pituitary, only the 1.7-kb transcript could be recognized. The presence of a single transcript for CaM in the pituitary is consistent with the results of our cloning studies, in which only CaM-L (corresponding to the 1.7-kb transcript), but not CaM-S, could be isolated from the RT samples prepared from the pituitary. Because both CaM-L and -S were pulled out from the RT samples of the gills, in which only the 1.3- and 1.7-kb transcripts could be detected, it would be logical to assume that the 1.3-kb transcript is the mRNA template responsible for the cloning of CaM-S. Given that our 3'/5' RACE using RT samples prepared from brain, heart, and kidney did not yield additional cDNAs for CaM, the sequence identity of the 3.0-kb transcript is still unknown. Although nucleotide sequence alignment of CaM-L and -S indicates that these two transcripts are generated probably by differential use of polyadenylation signals, we do not exclude the possibility that the variants of CaM mRNA may be caused by alterations in promoter usage or differential expression of multiple CaM genes. At present, the physiological relevance of tissue-specific expression of different isoforms of CaM transcripts is unclear.
Using static incubation of grass carp pituitary cells as a model, IGF-I and -II have been demonstrated for the first time to stimulate CaM mRNA expression at the pituitary level. These stimulatory effects were specific to the 1.7-kb CaM transcript and did not involve the expression of 1.3- and 3.0-kb CaM transcripts. This increase in CaM transcripts also paralleled a rise in CaM content in pituitary cells, indicating that IGF can induce CaM synthesis by up-regulation of CaM gene expression. Based on spatial and temporal correlation studies, these CaM mRNA responses occurred in grass carp somatotrophs preceding the inhibitory actions of IGF-I and IGF-II on GH mRNA expression. Apparently, IGF-induced CaM gene expression could not be due to a cross-reactivity with insulin receptors, because insulin did not mimic the effects of IGF-I and IGF-II in the present study. Unlike IGF, insulin treatment increased GH mRNA expression, but reduced CaM mRNA levels, in grass carp pituitary cells. In mammals, it is well documented that IGF can exert a long-loop feedback at the pituitary level to suppress GH release and GH gene expression (49). The present finding of IGF inhibition on GH mRNA expression in the grass carp confirms that the role of IGF as a negative regulator of GH synthesis is conserved during vertebrate evolution. Similar findings have been reported in other fish species, including the tilapia (31), striped bass (33), and turbot (50). It is worth mentioning that insulin can induce CaM gene transcription (51) by activation of specificity protein-1 (Sp1)/Sp3 binding to the Sp1 sites in the CaM promoter (52). In rat pituitary cells (53) or pituitary cell lines transfected with GH promoter (54), insulin treatment can also suppress GH release and GH gene transcription. In fish models, the actions of insulin on GH release tend to be more variable. Insulin has no effect on GH secretion in trout pituitary cells (55) but can suppress GH release in pituitary cells prepared from striped bass (33) and turbot (50). In this study, insulin, by acting directly at the pituitary level, was shown to activate GH gene expression in a carp species. These findings as a whole suggest that the functional role of insulin on GH regulation may be species specific in lower vertebrates. Given that 1) both mammalian and fish insulin can induce bioactivity with similar potency in fish hepatocytes (56); and 2) mammalian insulin does not bind IGF receptors in fish models (57), the possibility of differential actions of insulin on CaM and GH transcript expression caused by the use of heterologous insulin (i.e. human insulin) is rather unlikely.
Apparently, IGF induction of CaM mRNA expression is not mediated through modulation of CaM transcript stability, because the t1/2 of CaM mRNA in carp pituitary cells was not affected by IGF treatment. In contrast, the expression levels of CaM primary transcripts were elevated in a dose-dependent manner after IGF-I and IGF-II stimulation, indicating that CaM gene transcription might have been activated. This idea is consistent with the results of transfection studies in T3-1 cells, in which IGF-I and IGF-II could up-regulate grass carp CaM promoter activity. Although cis-acting elements, including activator protein-1, Pit-1, insulin receptor substrate, and Sp1 binding sites, can be identified in the grass carp CaM promoter (Huo, L., and A. O. L. Wong, unpublished observations), their functional role in IGF induction of CaM gene expression remains to be determined. Given that 1) a negative correlation between CaM and GH mRNA expression was noted; and 2) the rise in CaM mRNA occurred before the drop in GH gene expression induced by IGF treatment, the involvement of CaM in GH long-loop feedback by IGF is suspected. Using a pharmacological approach, suppressing the functionality of endogenous CaM by CaM antagonism was found to be effective in raising basal GH mRNA levels as well as blocking the inhibitory effects of IGF on GH gene expression. Besides, lowering of GH mRNA levels by IGF treatment could be abolished by inactivation of calcineurin, but not CaM kinase II, suggesting that the CaM/calcineurin cascades may be acting downstream of IGF receptors. In GH4C1 cells, GH promoter activity was inhibited by CaM overexpression, and the responsive element(s) could be mapped to the proximal promoter of GH gene before position –115. Our sequence analysis based on TESS site search has revealed that a consensus sequence of nuclear factor-AT (NF-AT) binding site can be located in this region (Wong, A. O. L., unpublished observations), which raises the possibility that CaM by acting through calcineurin-induced dephosphorylation, and nuclear translocation of (NF-AT) can modulate GH gene expression at the transcriptional level. Our results as a whole have two implications regarding the function and regulation of CaM in the carp pituitary: 1) the basal level of GH gene transcription may be under the negative regulation of endogenous CaM; and 2) IGF-I and IGF-II may up-regulate CaM gene expression to inhibit GH synthesis at the pituitary level via CaM/calcineurin-dependent mechanisms. Although our results are in agreement with the general observations that IGF inhibition of GH release and GH gene expression tends to have a slow onset (12 h) and requires de novo protein synthesis (24), the situation in the fish model appears to be different from that in mammalian counterparts. In mammals, especially in the rat, CaM is involved in GH secretion (5), but not GH gene expression (18). In somatotroph cell lines, e.g. MtT/S cells, IGF inhibits GH promoter activity via activation of the phosphotidylinositol 3-kinase, but not MAPK or p70S6K cascades (24). To our knowledge, the involvement of CaM gene expression in the postreceptor signaling mechanisms for IGF receptors has not been previously demonstrated.
In summary, we have established the structural identity of grass carp CaM by molecular cloning and confirmed that multiple transcripts of CaM are ubiquitously expressed in the grass carp. The newly cloned CaM cDNAs are phylogenetically related to the CaM I gene subfamily. Based on the sequence obtained, a slot-blot system was set up to quantify the expression of CaM mRNA in grass carp pituitary cells. In this case, we have demonstrated for the first time that IGF-I and IGF-II induced CaM mRNA expression at the pituitary level, and this stimulatory effect could be correlated both spatially and temporally with the drop in GH mRNA levels. The stimulatory action of IGFs on CaM transcript expression was not mediated by modulation of CaM mRNA stability, but was caused by a direct activation of CaM gene transcription. In parallel experiments, we have shown that CaM antagonism and calcineurin inactivation were effective in blocking IGF-inhibited GH gene expression, whereas CaM overexpression significantly suppressed grass carp GH promoter activity. These results, taken together, indicate that IGF inhibits GH synthesis in carp pituitary cells through up-regulation of CaM gene expression. The present study provides direct evidence that modulation of CaM gene expression at the pituitary level can serve as the site of regulation for GH synthesis. Our studies in fish also provide a comparative model for mammalian research to revisit the signal transduction of IGF long-loop feedback at the pituitary level. Given that IGF is involved in the transformation and metastasis of tumor cells (53), the identification of CaM gene expression as a novel signaling component for IGF action may have clinical implications in cancer research.
Acknowledgments
Special thanks are given to Dr. P. C. Leung for the supply of CaM antiserum and bovine CaM. We are also indebted to Drs. W. K. K. Ho and W. W. M. Lee for their support in setting up the assay systems for GH mRNA and CaM primary transcripts, respectively.
Footnotes
This work was supported by RGC (H.K.) and CRCG grants (H.K.U.) to A.O.L.W. Financial support from the Department of Zoology (H.K.U.) to L.H., G.F., and X.W. in the form of postgraduate studentships is also acknowledged.
Abbreviations: a.a., Amino acid; Ca2+/CaM kinase II, calmodulin-dependent protein kinase II; CaM, calmodulin; DIG, digoxigenin; EGFP, enhanced green fluorescence protein; FBS, fetal bovine serum; LSD, least significance difference; Luc, luciferase; ORF, open reading frame; PRL, prolactin; RACE, rapid amplification of cDNA ends; Sp1, specificity protein-1; SRIF, somatostatin; Tm, melting temperature; 3'UTR, 3'-untranslated region.
References
Weinstein H, Mehler EL 1994 Ca2+-binding and structural dynamics in the functions of calmodulin. Annu Rev Physiol 56:213–236
Means AR 2000 Calcium and calmodulin-mediated regulatory mechanisms. In: Conn PM, Means AR, eds. Principles of molecular regulation. Totowa, NJ: Humana Press; 207–231
Kumar RV, Panniers R, Wolfman A, Henshaw EC 1991 Inhibition of protein synthesis by antagonists of calmodulin in Ehrlich ascites tumor cells. Eur J Biochem 195:313–319
Schettini G, Florio T, Meucci O, Landolfi E, Cronin MJ, MacLeod RM 1987 Calmodulin modulates prolactin secretion in vitro: studies with calmodulin containing liposomes. Life Sci 41:2437–2444
Sletholt K, Haug E, Gordeladze J, Sand O, Gautvik KM 1987 Effects of calmodulin antagonists on hormone release and cyclic AMP levels in GH3 pituitary cells. Acta Physiol Scand 130:333–343
Bonafe NM, Sellers, JR 1998 Calmodulin-binding protein of the cytoskeleton. In: Van Eldik L, Watterson DM, eds. Calmodulin and signal transduction. San Diego: Academic Press; 348–396
Schmalzigaug R, Ye Q, Berchtold MW 2001 Calmodulin protects cells from death under normal growth conditions and mitogenic starvation but plays a mediating role in cell death upon B-cell receptor stimulation. Immunology 103:332–342
Takano H, Fukushi H, Morishima Y, Shirasaki Y 2003 Calmodulin and calmodulin-dependent kinase II mediate neuronal cell death induced by depolarization. Brain Res 962:41–47
Prostko CR, Zhang C, Hait WN 1997 The effects of altered cellular calmodulin expression on the growth and viability of C6 glioblastoma cells. Oncol Res 9:13–17
Luk SC, Ngai SM, Tsui SK, Fung KP, Lee CY, Waye MM 1999 In vivo and in vitro association of 14–3-3 epsilon isoform with calmodulin: implication for signal transduction and cell proliferation. J Cell Biochem 73:31–35
Walker SW, Wark JD, MacNeil S, Mellersh H, Brown BL, Tomlinson S 1984 Isolation, purification and cell-free synthesis of calmodulin from the pig anterior pituitary gland. Biochem J 217:827–832
Kakiuchi S, Yasuda S, Yamazaki R, Teshima Y, Kanda K, Kakiuchi R, Sobue K 1982 Quantitative determinations of calmodulin in the supernatant and particulate fractions of mammalian tissues. J Biochem 92:1041–1048
Aguila MC, McCann SM 1988 Calmodulin dependence of somatostatin release stimulated by growth hormone-releasing factor. Endocrinology 123:305–309
Honegger J, D’Urso R, Besser GM, Grossman AB 1991 Calcium and calmodulin mediation of growth hormone-releasing hormone release from the rat hypothalamus in vitro. Endocrinology 129:11–16
Davis JR, Hoggard N, Wilson EM, Vidal ME, Sheppard MC 1991 Calcium/calmodulin regulation of the rat prolactin gene is conferred by the proximal enhancer region. Mol Endocrinol 5:8–12
Izumi S, Iwashita M, Makino T, Saito S, Sakamoto S, Takeda Y, Nozawa S 1991 Phorbol ester-induced LH release in pituitary gonadotrophs: effects of antagonists of calmodulin and GnRH. Endocrinol Jpn 38:195–204
Martinez-Campos A, Hernandez RP, Forsbach G, Barrera-Saldana HA 1991 The stimulatory effect of estradiol 17- on prolactin mRNA is inhibited by anti-calmodulin drugs. Life Sci 48:2475–2485
White BA 1985 Evidence for a role of calmodulin in the regulation of prolactin gene expression. J Biol Chem 260:1213–1217
Le Roith D 2003 The insulin-like growth factor system. Exp Diabetes Res 4:205–212
Svensson J, Lonn L, Johannsson G, Bengtsson BA 2003 Effects of GH and insulin-like growth factor-I on body composition. J Endocrinol Invest 26:823–831
Dupont J, Khan J, Qu BH, Metzler P, Helman L, LeRoith D 2001 Insulin and IGF-I induce different patterns of gene expression in fibroblast NIH-3T3 cells: identification by cDNA microarray analysis. Endocrinology 142:4969–4975
Jaffe CA, Pan W, Brown MB, DeMott-Friberg R, Barkan AL 2001 Regulation of GH secretion in acromegaly: reproducibility of daily GH profiles and attenuated negative feedback by IGF-I. J Clin Endocrinol Metab 86:4364–4370
Weber MM, Melmed S, Rosenbloom J, Yamasaki H, Prager D 1992 Rat somatotroph insulin-like growth factor-II (IGF-II) signaling: role of the IGF-I receptor. Endocrinology 131:2147–2153
Niiori-Onishi A, Iwasaki Y, Mutsuga N, Oiso Y, Inoue K, Saito H 1999 Molecular mechanisms of the negative effect of insulin-like growth factor-I on growth hormone gene expression in MtT/S somatotroph cells. Endocrinology 140:344–349
Fletcher TP, Thomas GB, Dunshea FR, Moore LG, Clarke IJ 1995 IGF feedback effects on growth hormone secretion in ewes: evidence for action at the pituitary but not the hypothalamic level. J Endocrinol 144:323–331
Becker K, Stegenga S, Conway S 1995 Role of insulin-like growth factor I in regulating growth hormone release and feedback in the male rat. Neuroendocrinology 61:573–583
Moriyama S, Ayson FG, Kawauchi H 2000 Growth regulation by insulin-like growth factor-I in fish. Biosci Biotechnol Biochem 64:1553–1562
Tse MC, Vong QP, Cheng CH, Chan KM 2002 PCR-cloning and gene expression studies in common carp insulin-like growth factor-II. Biochim Biophys Acta 1575:63–74
Vong QP, Chan KM, Cheng CH 2003 Quantification of common carp IGF-I and IGF-II mRNA by real-time PCR: differential regulation of expression by GH. J Endocrinol 178:513–521
Fruchtman S, Gift B, Howes B, Borski R 2001 Insulin-like growth factor-I augments prolactin and inhibits growth hormone release through distinct as well as overlapping cellular signaling pathways. Comp Biochem Physiol B 129:237–242
Kajimura S, Uchida K, Yada T, Hirano T, Aida K, Gordon Grau E 2002 Effects of insulin-like growth factors (IGF-I and -II) on growth hormone and prolactin release and gene expression in euryhaline tilapia, Oreochromis mossambicus. Gen Comp Endocrinol 127:223–231
Baker DM, Davies B, Dickhoff WW, Swanson P 2000 Insulin-like growth factor I increases follicle-stimulating hormone (FSH) content and gonadotropin-releasing hormone-stimulated FSH release from Coho salmon pituitary cells in vitro. Biol Reprod 63:865–871
Fruchtman S, Jackson L, Borski R 2000 Insulin-like growth factor I disparately regulates prolactin and growth hormone synthesis and secretion: studies using the teleost pituitary model. Endocrinology 141:2886–2894
Chang JP, Johnson JD, Van Goor F, Wong CJ, Yunker WK, Uretsky AD, Taylor D, Jobin RM, Wong AOL, Goldberg JI 2000 Signal transduction mechanisms mediating secretion in goldfish gonadotropes and somatotropes. Biochem Cell Biol 78:139–153
Chang JP, Abele JT, Van Goor F, Wong AOL, Neumann CM 1996 Role of arachidonic acid and calmodulin in mediating dopamine D1- and GnRH-stimulated growth hormone release in goldfish pituitary cells. Gen Comp Endocrinol 102:88–101
Jobin RM, Neumann CM, Chang JP 1996 Roles of calcium and calmodulin in the mediation of acute and sustained GnRH-stimulated gonadotropin secretion from dispersed goldfish pituitary cells. Gen Comp Endocrinol 101:91–106
Wong AOL, Ng S, Lee EK, Leung RC, Ho WK 1998 Somatostatin inhibits (D-Arg6, Pro9-NEt) salmon gonadotropin-releasing hormone- and dopamine D1-stimulated growth hormone release from perifused pituitary cells of Chinese grass carp, Ctenopharyngodon idellus. Gen Comp Endocrinol 110:29–45
Zhou H, Wang X, Ko WK, Wong AOL 2004 Evidence for a novel intrapituitary autocrine/paracrine feedback loop regulating growth hormone synthesis and secretion in carp pituitary cells by functional interactions between gonadotrophs and somatotrophs. Endocrinology 145:5548–5559
Zhou H, Jiang Y, Ko WK, Li W, Wong AOL 2005 Paracrine regulation of growth hormone gene expression by gonadotrophin release in grass carp pituitary cells: functional implications, molecular mechanisms and signal transduction. J Mol Endocrinol 34:415–432
Toutenhoofd SL, Strehler EE 2000 The calmodulin multigene family as a unique case of genetic redundancy: multiple levels of regulation to provide spatial and temporal control of calmodulin pools Cell Calcium 28:83–96
Berchtold MW, Egli R, Rhyner JA, Hameister H, Strehler EE 1993 Localization of the human bona fide calmodulin genes CALM1, CALM2, and CALM3 to chromosomes 14q24–q31, 2p21.1-p21.3, and 19q13.2-q13.3. Genomics 16:461–465
Reid DG, MacLachlan LK, Gajjar K, Voyle M, King RJ, England PJ 1990 A proton nuclear magnetic resonance and molecular modeling study of calmidazolium (R24571) binding to calmodulin and skeletal muscle troponin C. J Biol Chem 265:9744–9753
Nojima H 1987 Molecular evolution of the calmodulin gene. FEBS Lett 217:187–190
Saimi Y, Kung C 2002 Calmodulin as an ion channel subunit. Annu Rev Physiol 64:289–311
Sola C, Barron S, Tusell JM, Serratosa J 1999 The Ca2+/calmodulin signaling system in the neural response to excitability. Involvement of neuronal and glial cells. Prog Neurobiol 58:207–232
Sola C, Barron S, Tusell JM, Serratosa J 2001 The Ca2+/calmodulin system in neuronal hyperexcitability. Int J Biochem Cell Biol 33:439–455
Gill KD, Gupta V, Sandhir R 2003 Ca2+/calmodulin-mediated neurotransmitter release and neurobehavioural deficits following lead exposure. Cell Biochem Funct 21:345–353
Zhang J, Suneja SK, Potashner SJ, Rameau GA, Chiu LY, Ziff EB 2004 Protein kinase A and calcium/calmodulin-dependent protein kinase II regulate glycine and GABA release in auditory brain stem nuclei: bidirectional regulation of neuronal nitric-oxide synthase phosphorylation at serine 847 by the N-methyl-D-aspartate receptor. J Neurosci Res 75:361–370
Le Roith D, Bondy C, Yakar S, Liu JL, Butler A 2001 The somatomedin hypothesis: 2001. Endocr Rev 22:53–74
Duval H, Rousseau K, Elies G, Le Bail PY, Dufour S, Boeuf G, Boujard D 2002 Cloning, characterization, and comparative activity of turbot IGF-I and IGF-II. Gen Comp Endocrinol 126:269–278
Solomon SS, Palazzolo MR, Smoake JA, Raghow RS 1995 Insulin-stimulated calmodulin gene expression in rat H-411E cells can be selectively blocked by antisense oligonucleotides. Biochem Biophys Res Commun 210:921–930
Solomon SS, Palazzolo MR, Takahashi T, Raghow R 1997 Insulin stimulates rat calmodulin I gene transcription through activation of Sp1. Proc Assoc Am Physicians 109:470–477
Yamashita S, Melmed S 1986 Effects of insulin on rat anterior pituitary cells. Inhibition of growth hormone secretion and mRNA levels. Diabetes 35:440–447
Prager D, Melmed S 1988 Insulin regulates expression of the human growth hormone gene in transfected cells. J Biol Chem 263:16580–16585
Weil C, Carre F, Blaise O, Breton B, Le Bail PY 1999 Differential effect of insulin-like growth factor I on in vitro gonadotropin and growth hormone secretions in rainbow trout (Oncorhynchus mykiss) at different stages of the reproductive cycle. Endocrinology 140:2054–2062
Pierce AL, Fukada H, Dickhoff WW 2005 Metabolic hormones modulate the effect of growth hormone (GH) on insulin-like growth factor-I (IGF-I) mRNA level in primary culture of salmon hepatocytes. J Endocrinol 184:341–349
Castillo J, Le Bail PY, Paboeuf G, Navarro I, Weil C, Fauconneau B, Gutierrez J 2002 IGF-I binding in primary culture of muscle cells of rainbow trout: changes during in vitro development. Am J Physiol 283:R647–R652(Longfei Huo, Guodong Fu, )
Abstract
Calmodulin (CaM), the Ca2+ sensor in living cells, is essential for biological functions mediated by Ca2+-dependent mechanisms. However, modulation of CaM gene expression at the pituitary level as a means to regulate pituitary hormone synthesis has not been characterized. In this study we examined the functional role of CaM in the feedback control of GH by IGF using grass carp pituitary cells as a cell model. To establish the structural identity of CaM expressed in the grass carp, a CaM cDNA, CaM-L, was isolated from the carp pituitary using 3'/5' rapid amplification of cDNA ends. The open reading frame of this cDNA encodes a 149-amino acid protein sharing the same primary structure with CaMs reported in mammals, birds, and amphibians. This CaM cDNA is phylogenetically related to the CaM I gene family, and its transcripts are ubiquitously expressed in the grass carp. In carp pituitary cells, IGF-I and IGF-II induced CaM mRNA expression with a concurrent drop in GH transcript levels. These stimulatory effects on CaM mRNA levels were not mimicked by insulin and appeared to be a direct consequence of IGF activation of CaM gene transcription without altering CaM transcript stability. CaM antagonism and inactivation of calcineurin blocked the inhibitory effects of IGF-I and IGF-II on GH gene expression, and CaM overexpression also suppressed the 5' promoter activity of the grass carp GH gene. These results, as a whole, provide evidence for the first time that IGF feedback on GH gene expression is mediated by activation of CaM gene expression at the pituitary level.
Introduction
CALMODULIN (CaM) IS a Ca2+-binding protein serving as an intracellular Ca2+ sensor in living cells. Upon Ca2+ binding, conformational changes occur in the CaM molecule with the exposure of hydrophobic domains for subsequent association with CaM target proteins (1). The interactions of CaM with its target proteins activate a variety of cellular processes, including gene expression (2), protein synthesis (3), hormone secretion (4, 5), cell motility (6), apoptosis (7, 8), and cell proliferation (9, 10). In mammals, a high level of CaM expression was noted in the brain-pituitary axis (11, 12). This finding is consistent with the reports that CaM is involved in the synthesis and/or secretion of hypothalamic factors [e.g. somatostatin (SRIF) (13) and GHRH (14)] as well as pituitary hormones [e.g. prolactin (PRL) (4, 15) and LH (16)]. In GH3 cells, modulation of GH release by CaM has been implicated (5). In this case, CaM antagonists, by altering phosphodiesterase activity, can induce a biphasic effect on GH release, being inhibitory at low doses and stimulatory at high doses. Although CaM can activate PRL gene expression both in vivo (17) and in vitro (18) by elevating PRL promoter activity through a proximal enhancer element (15), CaM antagonism has little or no effect on GH gene expression in these reports. Given that functional studies of CaM production have not been fully characterized, direct evidence for modulation of CaM gene expression at the pituitary level as a means to regulate pituitary functions is still lacking.
IGF, a polypeptide with structural similarity to proinsulin, is produced mainly in the liver under the stimulatory influence of GH. In general, IGF serves as a mediator for GH actions in regulating somatic growth, body metabolism, and cell proliferation and differentiation (19, 20). IGF can also exert a long-loop feedback to inhibit GH synthesis and secretion (21), and this negative feedback is well documented in humans (22) and nonprimate mammals (23, 24). The site of action for IGF long-loop feedback appears to be species specific. Direct actions of IGF at the pituitary level without a hypothalamic component have been reported in sheep (25), whereas central actions to modify GHRH and SRIF secretion from the hypothalamus have been clearly demonstrated in the rat (26). In somatotroph cell lines, e.g. MtT/S cells, IGF inhibits GH gene expression by reducing GH promoter activity through phosphotidylinositol 3-kinase-mediated signaling events (24). Although no information is available to date in lower vertebrates on IGF regulation of GH-releasing factors at the hypothalamic level, the long-loop feedback by IGF appears to be conserved in fish. This idea is supported by the findings that 1) the molecular structure of IGFs (including IGF-I and IGF-II) is highly conserved from fish to mammals (27); 2) GH stimulates IGF production in the liver of fish species (28, 29); and 3) IGF inhibits GH secretion (30, 31) and synthesis (30, 31) in fish pituitary cells in vitro. Recently, IGF has been shown to induce LH and PRL release in Coho salmon (32) and striped bass (33), respectively. These findings suggest that IGF may also act on pituitary cells other than somatotrophs in lower vertebrates.
In fish models, e.g. goldfish, the availability of extracellular Ca2+ and its entry via voltage-sensitive Ca2+ channels are essential for both basal and stimulated GH release (34). In goldfish pituitary cells, GH secretion induced by GH-releasing factors, e.g. GnRH and dopamine, can be blocked by CaM antagonists and CaM kinase II inhibitors (35, 36). These results indicate that CaM and its downstream signaling pathways are involved in the GH-releasing actions of these GH regulators. At present, however, it is still unclear whether GH gene expression can be regulated by CaM at the pituitary level, and to our knowledge, the functional role of CaM in GH feedback by IGF has not been previously examined. In this study, using grass carp as a model, we tested the hypothesis that IGF inhibits GH gene expression directly at the pituitary level via modulation of CaM gene expression. As a first step, the structural identity of grass carp CaM was established by molecular cloning using 5'/3' rapid amplification of cDNA ends (RACE). The expression profile of CaM mRNA was examined in various tissues of the grass carp by Northern blot. Based on the CaM sequence obtained, a slot-blot assay was set up for CaM mRNA measurement. Using this assay system, the effects of IGF-I and IGF-II on CaM mRNA expression were tested in grass carp pituitary cells and correlated with the corresponding changes in GH mRNA levels. In parallel experiments, the effects of IGF-I and IGF-II on CaM transcript stability, CaM primary transcript expression, and 5' promoter activity of the CaM gene were also examined. The functional role of CaM in GH gene expression was also evaluated by testing the effects of CaM antagonism on GH mRNA levels and the effects of CaM overexpression on GH promoter activity.
Materials and Methods
Animals
One-year old (1+) Chinese grass carp (Ctenopharyngodon idellus) were purchased from the local markets in Hong Kong and housed in 200-liter aquaria at 18 ± 2 C for at least 3 d before tissue collection and/or pituitary cell preparation. On the day of the experiments, the fish were killed by anesthesia in MS222 (tricaine methanesulfonate; Sigma, St. Louis, MO), followed by spinosectomy according to the regulations for animal use at University of Hong Kong.
Reagents and test substances
Human IGF-I and IGF-II were purchased from Sigma-Aldrich Corp. (St. Louis, MO). TRIzol, medium 199, Opti-MEM, Ham’s F-10, and fetal bovine serum (FBS) were obtained from Invitrogen Life Technologies, Inc. (Grand Island, NY). Calmidazolium and human insulin were acquired from Sigma-Aldrich Corp., and actinomycin D was obtained from Calbiochem (La Jolla, CA). Insulin was first dissolved in double-distilled deionized water to prepare a working stock with an activity level of 109 μU/ml. The stock solution was then aliquoted into small volume and stored frozen at –80 C until used. IGF-I and IGF-II were prepared in a similar manner, except that these peptide hormones were dissolved in 10 nM HCl to give a stock concentration at 100 μM. Given that calmidazolium has a low solubility in aqueous solution, the stock solution at 10 mM was prepared in dimethylsulfoxide. On the day of experiments, test substances were diluted to appropriate concentrations with culture medium 20 min before drug treatment. In these studies, the final dilutions of dimethylsulfoxide and HCl were always 0.1% (vol/vol) or less of the original levels and did not affect basal expression of CaM and GH mRNA in grass carp pituitary cells.
Molecular cloning of grass carp CaM cDNA
Total RNA (5 μg) was extracted from grass carp pituitaries using TRIzol (Invitrogen Life Technologies, Inc.), and first strand cDNA was synthesized using a SuperScript II first strand synthesis kit (Invitrogen Life Technologies, Inc.). A pair of primers designed based on the conserved regions of mammalian CaM cDNAs were used to isolate a 463-bp fragment of grass carp CaM cDNA by PCR using the first strand cDNA as the template. Based on the sequence obtained, new primers were designed for 5'/3' RACE using a GeneRacer kit (Invitrogen Life Technologies, Inc., Carlsbad, CA) with pituitary total RNA (5 μg) as the template. PCRs were conducted according to the instructions of the manufacturer. PCR products were gel-purified and subcloned into the pGEM-T Easy vector (Promega Corp., Madison, WI) for DNA sequencing using a BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA). The full-length cDNA of CaM (GenBank no. AY627883) was reconstructed from its 5' and 3' sequences using MacVector 6.5 (Oxford Molecular, Madison, WI) and MacDNASIS PRO 3.5 programs (Hitachi, San Bruno, CA). After that, a digoxigenin (DIG)-labeled cDNA probe for grass carp CaM mRNA was prepared for subsequent Northern blot and slot-blot assays using a PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany).
Northern blot of CaM transcripts
Total RNA (25 μg) was extracted with TRIzol from selected tissues of the grass carp, including brain, pituitary, muscle, liver, kidney, spleen, heart, intestine, and gills. After that, mRNA was purified using a PolyATract mRNA Isolation System (Promega Corp.). These mRNA samples were denatured, size fractionated in 1% agarose gel, and transblotted onto a positively charged nylon membrane (Roche) using a VacuGene Vacuum Blotting System (Pharmacia Biotech, Piscataway, NJ). The membrane was then UV cross-linked, prehybridized for 3 h, and incubated with the DIG-labeled CaM cDNA probe for 15 h at 42 C. On the following day, the membrane was washed twice at 68 C in 0.5x standard saline citrate with 0.1% sodium dodecyl sulfate, and hybridization signals were detected using a DIG Lumin Detection Kit (Roche). In these studies, a Northern blot of -actin was used as an internal control.
Measurement of steady-state CaM and GH mRNA expression
Primary cultures of pituitary cells were prepared from 1-yr-old grass carp by a trypsin/deoxyribonuclease digestion method as described previously (37). The average cell yield was approximately 8 x 106 cells/pituitary, and cell viability was always 95% or greater. After cell dispersion, pituitary cells were cultured overnight (>15 h) with 5% FBS in 24-well culture plates precoated with poly-D-lysine (Sigma-Aldrich Corp.) at a seeding density of about 3 x 106 cells/ml/well. After that, test substances at appropriate concentrations were prepared in carp MEM (38) and gently overlaid onto pituitary cells after removal of old culture medium. For dose-response studies, the duration of drug treatment was routinely fixed at 48 h. For the evaluation of CaM transcript stability, cells were pretreated with actinomycin D (8 μM) and incubated with or without IGF treatment for the duration as indicated in individual experiments. After drug treatment, total RNA (2.5 μg) was extracted with TRIzol, heat-denatured at 70 C for 15 min, and vacuum-blotted onto nylon membranes in duplicate using a Bio-Dot microfiltration unit (Bio-Rad Laboratories, Hercules, CA). The membranes were UV cross-linked, prehybridized for 3 h, and hybridized overnight with DIG-labeled CaM probe and grass carp GH cDNA probe (38), respectively. After hybridization, the membranes were washed, and signal development was carried out as described for the Northern blot. Measurement of hybridization signals was conducted in a Kodak 440 Image Station (Kodak Digital Science, Rochester, NY). In this study, slot blots of 18S RNA were also conducted to serve as an internal control.
Real-time PCR of mature CaM mRNA and CaM primary transcript
Based on the sequence of the newly cloned CaM cDNA, intron trapping was performed using genomic DNA as a template to pull out introns 1–5 of the grass carp CaM gene (GenBank no. AY656698). Primers flanking the junction between intron 4 and exon 5 from positions 10,408–10,789 were used for real-time PCR of CaM primary transcripts. To serve as a parallel control, real-time PCR of mature CaM mRNA was also performed using primers flanking intron 4 from position 10,209 (in exon 4) to 10,789 (in exon 5). In this study, pituitary cells were incubated for 48 h with increasing levels of IGF-I and IGF-II. Total RNA (5 μg) was isolated and digested with deoxyribonuclease I to remove genomic DNA contamination. After that, RT was carried out, and RT samples were subjected to PCR using a LightCycler-DNA Master SYBR Green I Kit (Roche) in a RotorGene 2000 Real-Time PCR System (Corbett Research, Eight Mile Plains, Australia). PCRs for CaM mRNA and primary transcripts were conducted for 35 cycles at 94 C for 30 sec for denaturing, 69 C for 30 sec for annealing, and 72 C for 30 sec for primer extension. Based on our validation, the primers for CaM primary transcripts were ineffective in PCR amplification of mature CaM mRNA. Furthermore, PCR of CaM primary transcripts in the nuclear fraction of grass carp pituitary cells consistently produced a 381-bp PCR product with a melting temperature (Tm) at 92.1 C, whereas the mature CaM mRNA in the cytosolic fraction would generate a 181-bp PCR product with Tm at 89.8 C. In these experiments, serial dilutions of plasmid DNA with CaM cDNA or the amplicon covering the junction between intron 4 and exon 5 of the CaM gene were used as the standards for real-time PCR of mature CaM mRNA and CaM primary transcripts, respectively.
Measurement of CaM promoter activity
Based on the sequence of grass carp CaM cDNA, nested primers covering the 5'-untranslated region (5'UTR) were designed to pull out the 5' promoter of CaM gene (GenBank no. AY656698) using a Universal GenomeWalker Kit (BD Clontech, Palo Alto, CA). Genomic DNA prepared from whole blood was used as the template, and PCRs were conducted according to the instructions of the manufacturer. The 5' promoter obtained (–1369 to +49) was subcloned into the NheI and XhoI sites of the pGL3.Basic vector (Promega Corp.) to generate the reporter construct pCaM(–1369).luciferase (Luc). This luciferase-expressing construct was used to examine IGF action on CaM promoter activity using the mouse pituitary cell line T3-1 as the host cells. The cells were maintained in DMEM with 44 mM NaHCO3, 10% FBS, and 1% (vol/vol) antibiotics-antimycotics (pH 7.4). Transfection was conducted for 6 h in 0.5 ml Opti-MEM with 1.16 μl Lipofectamine (Invitrogen Life Technologies, Inc.), 0.05 μg pCaM(–1369).Luc, 0.01 μg pEGFP-N1 (EGFP, enhanced green fluorescence protein), and 0.44 μg pBluescript (Stratagene) as carrier DNA. After transfection, T3-1 cells were cultured in DMEM with 10% FBS for 6 h, followed by serum starvation for 6 h before the initiation of IGF treatment. Based on our validation, the optimal duration of drug treatment was fixed at 12 h. After IGF treatment, T3-1 cells were dissolved in lysis buffer (Promega Corp.), and the lysate was cleared by centrifugation at 20,000 x g at 4 C for 15 min. For measurement of Luc activity, a 50-μl volume of cleared lysate was mixed with 100 μl luciferase assay reagent (Promega Corp.), and luminescence signal was quantified using a Lumat LB9507 luminometer (EG&G, Gaithersburg, MD). In these experiments, cotransfection with pEGFP-N1 was used as an internal control, and GFP expression was monitored by measuring the fluorescence signal in 100 μl lysate using a Cytofluor series 4000 Multiple-Well Plate Reader (Applied Biosystems, Foster City, CA).
CaM overexpression on GH promoter activity
To examine the functional role of CaM in GH gene transcription, the coding sequence of grass carp CaM cDNA was PCR isolated and subcloned into the expression vector pcDNA3.1 to produce pcDNA.CaM. The effects of CaM overexpression were examined with cotransfection of pGH.Luc constructs (39) carrying decreasing lengths (–986 to –115) of the grass carp GH gene promoter (GenBank no. X30419) using the rat pituitary cell line GH4C1 as the host cells. The cells were maintained in Ham’s F-10 with 14 mM NaHCO3, 10% FBS, and 1% (vol/vol) antibiotics-antimycotics (pH 7.4). Transfection was conducted for 6 h in 0.5 ml Opti-MEM with 1.16 μl Lipofectamine, 0.18 μg pGH.Luc, 0.01 μg pEGFP-N1, and 0.05–0.25 μg pcDNA.CaM (with pBluescript as carrier DNA). After transfection, GH4C1 cells were cultured in Ham’s F-10 for 18 h, dissolved in lysis buffer, and assayed for luciferase activity and GFP expression as described in the preceding section. To confirm CaM overexpression at the protein level, Western blot was also performed in GH4C1 cells after pcDNA.CaM transfection. Cell lysate was prepared by three cycles of repeated freezing and thawing in phenylmethylsulfonylfluoride (150 μg/ml) with EDTA (1.5 mM), cleared by centrifugation, and resolved in a 15% gel by SDS-PAGE for 1 h at 200 V. After that, protein samples were transferred onto a polyvinylidene difluoride membrane by low current electroblotting at 65 V for 2.5 h. The membrane was then subjected to blocking with 3% BSA and incubation with sheep antibovine CaM antibody (1:1000; Chemicon International, Temecula, CA) for 2 h. After washing, rabbit antisheep IgG-horseradish peroxidase (1:5000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was added, and signal development was carried out in a Kodak 440 Image Station with SuperSignal WestPico (Pierce Chemical Co., Rockford, IL) as the substrate. Using a similar approach, Western blot was also performed in carp pituitary cells to examine the effects of IGF treatment on CaM expression at the protein level. In this case, parallel probing of -actin using an Actin Ab-1 Kit (Calbiochem) was also conducted to serve as an internal control.
Data transformation and statistical analysis
For quantitation of CaM gene expression, CaM mRNA levels were measured in terms of arbitrary density unit and normalized against 18S rRNA expressed in the same sample. Given that no significant changes in 18S rRNA levels were noted, these data were simply transformed into a percentage of the mean value in the control group for statistical analysis (as percentage of control). The transformation was conducted to allow for data pooling from separate experiments without increasing the overall variability of the results. The half-life (t1/2) of CaM mRNA, defined as the time required for CaM mRNA to decrease to half its original level, was deduced by the one-phase exponential decay model using PRISM 3.02 (GraphPad, Inc., San Diego, CA). For real-time PCR, a standard curve with a dynamic range of 105 or greater and a correlation coefficient of 0.95 or more were used for data calibration. Mature CaM mRNA and primary transcripts were measured in terms of femtomoles per million cells and transformed as a percentage of the control for statistical analysis. For luciferase activity measurement, the raw data in arbitrary light units were normalized against the level of GFP expression in the same sample expressed in arbitrary fluorescence units. These data were expressed as the ratio of arbitrary light units/arbitrary fluorescence units or transformed as a percentage of the control as described in the preceding section. The transformed data for CaM mRNA and luciferase activity were analyzed by Student’s t test or ANOVA (two-way), followed by Fisher’s least significance difference (LSD) test. Differences were considered significant at P < 0.05.
Results
Molecular cloning of grass carp CaM cDNA
Using RT-PCR coupled to 5'/3' RACE, a full-length CaM cDNA, CaM-L, with a size of 1.55 kb was isolated from the grass carp pituitary (Fig. 1). In parallel studies, a shorter version of CaM cDNA, CaM-S, with a size of 0.73 kb was also obtained with CaM-L using RT samples prepared from the brain, gills, heart, and kidney. Nucleotide sequence analysis revealed that CaM-S is a truncated form of CaM-L with a shorter 3'UTR probably produced by differential use of polyadenylation signals. The 447-bp ORF of these newly cloned CaM cDNAs encodes a 149-amino acid (a.a.) protein with primary sequence identical with the CaMs reported in mammals (e.g. human, bovine, mouse, and rat), birds (e.g. chicken and duck), and amphibian (e.g. Xenopus) as well as in the representative species of modern bony fish (e.g. medaka and perch). Compared with CaMs reported in early evolved bony fish (e.g. eel), cyclostome (e.g. hagfish), urochordate (e.g. ciona), cephalochordate (e.g. brachiostoma), mollusca (e.g. aplysia), nematode (e.g. C. elegans), porifera (e.g. sponge), and protozoa (e.g. trypanosome), only a low level of a.a. substitutions were noted, and most of them were conserved mutations (Fig. 2). Sequence alignment at the a.a. level also revealed that the four helix-loop-helix EF hands (EF-I to EF-IV), the functional motifs of CaM for Ca2+ binding, are structurally conserved in vertebrates from modern bony fish to mammals. Although the structure of CaM is highly conserved, multiple copies of CaM genes, namely CaM I, CaM II, and CaM III, have been reported in birds and mammals (40, 41). Using unrooted analysis by PHYLIP (version 3.6), grass carp CaM can be clustered in the same clade with the CaM I genes reported in the rat, mouse, human, and chicken and is distally related to CaM II and CaM III gene subfamilies (Bootstrap values, 998–1000).
Tissue distribution of CaM transcripts
A DIG-labeled cDNA probe covering the ORF of CaM-L and -S was used to examine the tissue distribution of CaM expression by Northern blot (Fig. 3). mRNA samples were prepared from various tissues of the grass carp, including spleen, pituitary, muscle, liver, gut, kidney, heart, gills, and brain. The highest levels of CaM transcripts were noted in brain and gut; to a lesser extent in spleen, muscle, and pituitary; and to a lower level in heart, kidney, liver, and gills. In this study, three CaM transcripts (1.3, 1.7, and 3.0 kb, respectively) were identified, and all of them were expressed in brain, spleen, kidney, heart, intestine, and muscle. In the gills, only the 1.3- and 1.7-kb, not the 3.0-kb, transcripts were detected. Interestingly, the 3.0- and 1.7-kb, but not the 1.3-kb, transcripts were expressed in the liver. In the case of the pituitary, however, only the 1.7-kb transcript could be recognized. These results, as a whole, indicate that different forms of CaM mRNA are expressed in the grass carp in a tissue-specific manner.
Regulation of CaM mRNA expression in grass carp pituitary cells
Because Northern blot using purified mRNA is not ideal for quantitative studies with a large number of samples, a slot-blot assay using total RNA was established to study steady-state CaM mRNA expression in grass carp pituitary cells. To examine IGF effects on CaM gene expression at the pituitary level, pituitary cells were incubated for 48 h with increasing doses (0.1–100 nM) of IGF-I (Fig. 4A) and IGF-II (Fig. 4B). In these experiments, IGF-I and IGF-II were both effective in elevating CaM mRNA levels in a dose-dependent manner, with a concurrent drop in GH mRNA expression. The differential effects of IGFs on CaM and GH mRNA expression, however, were not mimicked by increasing doses of insulin (102–106 μU/ml). Unlike IGFs, insulin inhibited CaM mRNA levels (Fig. 5A), with a concurrent rise in GH mRNA expression (Fig. 5B). In these experiments, a negative correlation between CaM and GH mRNA levels was noted after IGF and insulin treatment (Fig. 5B), suggesting that modulation of CaM expression at the pituitary level may be involved in GH gene regulation. Because multiple transcripts of CaM are expressed in the grass carp, the possibility of IGF induction of CaM gene expression via differential expression of various forms of CaM mRNA was examined by Northern blot (Fig. 6A). In this case, IGF-I (100 nM) and IGF-II (100 nM) significantly increased the expression level of the 1.7-kb CaM mRNA in grass carp pituitary cells (Fig. 6B). The 1.3- and 3.0-kb CaM transcripts, however, remained undetectable in this study. Using Western blot, a time-dependent increase in CaM protein content was noted from 12–48 h in carp pituitary cells after the initiation of IGF-I treatment (100 nM; Fig. 6C). The protein levels of -actin, the internal control of these experiments, remained unchanged during the period of drug treatment. Similar results were also obtained using IGF-II (100 nM) as the stimulant for CaM expression (data not shown).
To evaluate the role of CaM in IGF inhibition of GH gene expression, spatial and temporal correlations between CaM and GH gene expression after IGF treatment were examined to test whether 1) the opposite changes in CaM and GH mRNA levels indeed occurred in the same cell type (i.e. in carp somatotrophs), and 2) the response in CaM mRNA expression was initiated before and/or during the corresponding changes in GH mRNA levels. For the study of spatial correlation, enriched somatotrophs were prepared from mixed populations of pituitary cells using Percoll gradient centrifugation (Fig. 7A). After a 48-h incubation with IGF-I (100 nM) and IGF-II (100 nM), CaM mRNA levels were increased in this somatotroph preparation, with a concurrent drop in GH mRNA expression (Fig. 7, B and C). In parallel time-course studies, basal levels of CaM mRNA were significantly increased (P < 0.05) in carp pituitary cells after 6-h incubation with IGF-I (100 nM) and IGF-II (100 nM), respectively, and remained elevated from 12–48 h (Fig. 8A). The corresponding drop in GH mRNA levels (P < 0.05), however, was noted only after 24-h incubation with IGF-I and IGF-II. After that, GH mRNA levels were reduced gradually from 24 to 48 h in a time-dependent manner (Fig. 8B). To establish the functional role of CaM in GH feedback by IGF, the effects of calmidazolium, a CaM antagonist, on both basal and IGF-inhibited GH mRNA expression were tested at the pituitary level. Calmidazolium is known to bind CaM tightly with nanomolar affinity to perturb a wide range of Ca2+-dependent mechanisms by immobilizing the methionine residues in the Ca2+-binding sites of helix-loop-helix EF motifs (42). In this case, GH mRNA levels were elevated in a dose-related fashion by increasing doses of calmidazolium (0.1 nM to 1 μM; Fig. 9A). In the presence of calmidazolium (5 nM), the inhibitory actions of IGF-I (100 nM) and IGF-II (100 nM) on GH mRNA expression were totally abolished (Fig. 9B). To elucidate the downstream signaling events that occurred after CaM activation, the functional roles of CaM kinase II and calcineurin in IGF inhibition of GH mRNA expression were also examined (Fig. 9C). Similar to calmidazolium, a 48-h incubation with the calcineurin inhibitor cyclosporin A (100 nM) increased basal levels of GH mRNA and blocked the inhibitory effect of IGF-I (100 nM) on GH transcript expression. The CaM kinase II inhibitor KN62 (5 μM), in contrast, reduced GH mRNA levels in grass carp pituitary cells and did not modify the inhibitory action of IGF-I (100 nM) on GH gene expression. Similar results were obtained by substituting IGF-II for IGF-I in these experiments (data not shown).
Molecular mechanisms for IGF induction of CaM gene expression
To shed light on the molecular mechanisms for IGF induction of CaM gene expression, the effects of IGF on CaM gene transcription and CaM transcript stability were examined. As a first step, real-time PCR of CaM primary transcripts was performed in grass carp pituitary cells after a 48-h incubation with increasing doses (0.1–100 nM) of IGF-I and IGF-II. As a parallel control, mature CaM mRNA levels were also monitored in these experiments. Calibration of primary transcript levels was conducted by linear regression of critical thresholds of the standard curve (Fig. 10A). In the same experiment, a separate standard curve was used for calibration for mature CaM mRNA (data not shown). In this study, IGF-I and IGF-II consistently induced a dose-dependent increase in CaM primary transcripts and mature mRNA (Fig. 10B). The specificity of PCRs was also confirmed by melting curve analysis and ethidium bromide staining of PCR products. In this case, a single 381-bp (Tm,92.1 C) and a single 181-bp PCR product (Tm, 89.8 C) were detected for CaM primary transcripts and mature CaM mRNA, respectively.
To provide additional evidence that IGF increases CaM mRNA levels via activation of CaM gene transcription, a reporter gene approach was used to examine the effects of IGF treatment on CaM promoter activity expressed in T3-1 cells. Compared with the time-matched controls, a significant increase in luciferase activity (P < 0.05) was noted after exposure to IGF-I (100 nM) for 6 h (Fig. 10C). The maximal difference in luciferase activity expressed in the control groups vs. the treatment groups was observed 12 h after the initiation of IGF-I treatment. Therefore, the duration of drug treatment was fixed at 12 h for our dose-response studies with IGF-I and IGF-II, respectively. In this case, increasing levels of IGF-I and IGF-II (0.01–100 nM) were both effective in stimulating luciferase activity expression in a dose-dependent manner (Fig. 10D). Although the maximal stimulatory effects on luciferase activity were similar, IGF-I appeared to be more potent than IGF-II in inducing luciferase expression, especially in subnanomolar doses.
Using grass carp pituitary cells pretreated with the transcription inhibitor actinomycin D (8 μM), the clearance curves of CaM mRNA were constructed in the presence or absence of IGF-I (100 nM; Fig. 10E). In these experiments, the two curves were found to be overlapping, and the t1/2 values deduced for CaM mRNA (73.4 h for the control vs. 73.2 h for IGF treatment) were also similar (P > 0.05). Comparable results were obtained in parallel experiments with IGF-II (data not shown), confirming that IGF-induced CaM gene expression in carp pituitary cells does not involve modification of CaM transcript stability.
CaM overexpression on GH gene promoter activity
The rat somatotroph cell line GH4C1 was used for the promoter studies of grass carp GH gene, because it has endogenous expression of Pit-1, a pituitary-specific transcription factor, to support basal GH promoter activity. Overexpression of CaM at the protein level was tested by Western blot in GH4C1 cells. Compared with the control transfected with the blank vector pcDNA 3.1, a significant increase in CaM immunoreactivity (16.8 kDa) was detected in the cell lysate prepared from the treatment group transfected with pcDNA.CaM (Fig. 11A). In GH4C1 cells cotransfected with pGH(–986).Luc carrying the –986 to +13 promoter of the grass carp GH gene, CaM overexpression was effective in reducing luciferase activity in a time-dependent manner (Fig. 11B). Apparently, this inhibitory effect was specific to the GH promoter, and the expression levels of GFP, the internal control, were not affected by CaM overexpression. In parallel experiments, transfection with increasing levels of pcDNA.CaM (0.05–0.25 μg) also suppressed luciferase activity in GH4C1 cells in a dose-dependent manner (Fig. 11C). In these studies, transfection of pGL3.Control and pGL3.Basic were used as the positive and negative controls, respectively. A high level of luciferase activity was consistently detected in the group transfected with pGL3.Control (with a simian virus 40 promoter) compared with that in the group transfected with the promoterless pGL3.Basic, confirming that the reporter system was appropriate for quantitative analysis of promoter activity. In our validation studies, the blank vector pGL3.Basic was also found to be nonresponsive to CaM overexpression (data not shown). To map out the coarse location of the promoter region responsible for CaM inhibition of GH promoter activity, CaM overexpression was conducted in GH4C1 cells transfected with pGH.Luc constructs carrying decreasing lengths of GH promoter from –986 to –115. In all pGH.Luc constructs examined, basal expression of luciferase activity was significantly suppressed by CaM overexpression (Fig. 11D), suggesting that the CaM-responsive element(s) is located within the proximal promoter region before position –115.
Discussion
In the present study, two CaM cDNAs, CaM-L and CaM-S, have been isolated in the grass carp. CaM-S is a truncated form of CaM-L, and both of them share a common ORF encoding a 149-a.a. protein with primary sequence identical with CaMs in mammals, birds, and amphibians. In mammals, nonallelic CaM genes, namely CaM I, CaM II, and CaM III, encoding the same CaM protein have been reported (40, 41). In general, these isoforms of CaM genes are believed to be the result of gene duplication that occurred 1400 million yr ago (43). Using unrooted analysis, it has been shown that the newly cloned CaM cDNAs are phylogenetically related to the CaM I gene subfamily. In higher vertebrates, the genetic redundancy of CaM genes has been attributed to the important function of CaM as the intracellular Ca2+ sensor in living cells (40). This idea is also in agreement with our findings that CaM transcripts are ubiquitously expressed in the grass carp. In this case, the highest levels of CaM transcripts can be found in the brain and intestine. In the rat, a high level of CaM (up to 0.5% of total protein) can be detected in the brain (44), which may be related to the functions of CaM in modulating neuronal excitability (45, 46) and exocytosis of neurotransmitters (47, 48). In this study, CaM transcripts of 1.3, 1.7, and 3.0 kb have been identified, and the expression pattern of these isoforms appears to be tissue specific. In the case of the pituitary, only the 1.7-kb transcript could be recognized. The presence of a single transcript for CaM in the pituitary is consistent with the results of our cloning studies, in which only CaM-L (corresponding to the 1.7-kb transcript), but not CaM-S, could be isolated from the RT samples prepared from the pituitary. Because both CaM-L and -S were pulled out from the RT samples of the gills, in which only the 1.3- and 1.7-kb transcripts could be detected, it would be logical to assume that the 1.3-kb transcript is the mRNA template responsible for the cloning of CaM-S. Given that our 3'/5' RACE using RT samples prepared from brain, heart, and kidney did not yield additional cDNAs for CaM, the sequence identity of the 3.0-kb transcript is still unknown. Although nucleotide sequence alignment of CaM-L and -S indicates that these two transcripts are generated probably by differential use of polyadenylation signals, we do not exclude the possibility that the variants of CaM mRNA may be caused by alterations in promoter usage or differential expression of multiple CaM genes. At present, the physiological relevance of tissue-specific expression of different isoforms of CaM transcripts is unclear.
Using static incubation of grass carp pituitary cells as a model, IGF-I and -II have been demonstrated for the first time to stimulate CaM mRNA expression at the pituitary level. These stimulatory effects were specific to the 1.7-kb CaM transcript and did not involve the expression of 1.3- and 3.0-kb CaM transcripts. This increase in CaM transcripts also paralleled a rise in CaM content in pituitary cells, indicating that IGF can induce CaM synthesis by up-regulation of CaM gene expression. Based on spatial and temporal correlation studies, these CaM mRNA responses occurred in grass carp somatotrophs preceding the inhibitory actions of IGF-I and IGF-II on GH mRNA expression. Apparently, IGF-induced CaM gene expression could not be due to a cross-reactivity with insulin receptors, because insulin did not mimic the effects of IGF-I and IGF-II in the present study. Unlike IGF, insulin treatment increased GH mRNA expression, but reduced CaM mRNA levels, in grass carp pituitary cells. In mammals, it is well documented that IGF can exert a long-loop feedback at the pituitary level to suppress GH release and GH gene expression (49). The present finding of IGF inhibition on GH mRNA expression in the grass carp confirms that the role of IGF as a negative regulator of GH synthesis is conserved during vertebrate evolution. Similar findings have been reported in other fish species, including the tilapia (31), striped bass (33), and turbot (50). It is worth mentioning that insulin can induce CaM gene transcription (51) by activation of specificity protein-1 (Sp1)/Sp3 binding to the Sp1 sites in the CaM promoter (52). In rat pituitary cells (53) or pituitary cell lines transfected with GH promoter (54), insulin treatment can also suppress GH release and GH gene transcription. In fish models, the actions of insulin on GH release tend to be more variable. Insulin has no effect on GH secretion in trout pituitary cells (55) but can suppress GH release in pituitary cells prepared from striped bass (33) and turbot (50). In this study, insulin, by acting directly at the pituitary level, was shown to activate GH gene expression in a carp species. These findings as a whole suggest that the functional role of insulin on GH regulation may be species specific in lower vertebrates. Given that 1) both mammalian and fish insulin can induce bioactivity with similar potency in fish hepatocytes (56); and 2) mammalian insulin does not bind IGF receptors in fish models (57), the possibility of differential actions of insulin on CaM and GH transcript expression caused by the use of heterologous insulin (i.e. human insulin) is rather unlikely.
Apparently, IGF induction of CaM mRNA expression is not mediated through modulation of CaM transcript stability, because the t1/2 of CaM mRNA in carp pituitary cells was not affected by IGF treatment. In contrast, the expression levels of CaM primary transcripts were elevated in a dose-dependent manner after IGF-I and IGF-II stimulation, indicating that CaM gene transcription might have been activated. This idea is consistent with the results of transfection studies in T3-1 cells, in which IGF-I and IGF-II could up-regulate grass carp CaM promoter activity. Although cis-acting elements, including activator protein-1, Pit-1, insulin receptor substrate, and Sp1 binding sites, can be identified in the grass carp CaM promoter (Huo, L., and A. O. L. Wong, unpublished observations), their functional role in IGF induction of CaM gene expression remains to be determined. Given that 1) a negative correlation between CaM and GH mRNA expression was noted; and 2) the rise in CaM mRNA occurred before the drop in GH gene expression induced by IGF treatment, the involvement of CaM in GH long-loop feedback by IGF is suspected. Using a pharmacological approach, suppressing the functionality of endogenous CaM by CaM antagonism was found to be effective in raising basal GH mRNA levels as well as blocking the inhibitory effects of IGF on GH gene expression. Besides, lowering of GH mRNA levels by IGF treatment could be abolished by inactivation of calcineurin, but not CaM kinase II, suggesting that the CaM/calcineurin cascades may be acting downstream of IGF receptors. In GH4C1 cells, GH promoter activity was inhibited by CaM overexpression, and the responsive element(s) could be mapped to the proximal promoter of GH gene before position –115. Our sequence analysis based on TESS site search has revealed that a consensus sequence of nuclear factor-AT (NF-AT) binding site can be located in this region (Wong, A. O. L., unpublished observations), which raises the possibility that CaM by acting through calcineurin-induced dephosphorylation, and nuclear translocation of (NF-AT) can modulate GH gene expression at the transcriptional level. Our results as a whole have two implications regarding the function and regulation of CaM in the carp pituitary: 1) the basal level of GH gene transcription may be under the negative regulation of endogenous CaM; and 2) IGF-I and IGF-II may up-regulate CaM gene expression to inhibit GH synthesis at the pituitary level via CaM/calcineurin-dependent mechanisms. Although our results are in agreement with the general observations that IGF inhibition of GH release and GH gene expression tends to have a slow onset (12 h) and requires de novo protein synthesis (24), the situation in the fish model appears to be different from that in mammalian counterparts. In mammals, especially in the rat, CaM is involved in GH secretion (5), but not GH gene expression (18). In somatotroph cell lines, e.g. MtT/S cells, IGF inhibits GH promoter activity via activation of the phosphotidylinositol 3-kinase, but not MAPK or p70S6K cascades (24). To our knowledge, the involvement of CaM gene expression in the postreceptor signaling mechanisms for IGF receptors has not been previously demonstrated.
In summary, we have established the structural identity of grass carp CaM by molecular cloning and confirmed that multiple transcripts of CaM are ubiquitously expressed in the grass carp. The newly cloned CaM cDNAs are phylogenetically related to the CaM I gene subfamily. Based on the sequence obtained, a slot-blot system was set up to quantify the expression of CaM mRNA in grass carp pituitary cells. In this case, we have demonstrated for the first time that IGF-I and IGF-II induced CaM mRNA expression at the pituitary level, and this stimulatory effect could be correlated both spatially and temporally with the drop in GH mRNA levels. The stimulatory action of IGFs on CaM transcript expression was not mediated by modulation of CaM mRNA stability, but was caused by a direct activation of CaM gene transcription. In parallel experiments, we have shown that CaM antagonism and calcineurin inactivation were effective in blocking IGF-inhibited GH gene expression, whereas CaM overexpression significantly suppressed grass carp GH promoter activity. These results, taken together, indicate that IGF inhibits GH synthesis in carp pituitary cells through up-regulation of CaM gene expression. The present study provides direct evidence that modulation of CaM gene expression at the pituitary level can serve as the site of regulation for GH synthesis. Our studies in fish also provide a comparative model for mammalian research to revisit the signal transduction of IGF long-loop feedback at the pituitary level. Given that IGF is involved in the transformation and metastasis of tumor cells (53), the identification of CaM gene expression as a novel signaling component for IGF action may have clinical implications in cancer research.
Acknowledgments
Special thanks are given to Dr. P. C. Leung for the supply of CaM antiserum and bovine CaM. We are also indebted to Drs. W. K. K. Ho and W. W. M. Lee for their support in setting up the assay systems for GH mRNA and CaM primary transcripts, respectively.
Footnotes
This work was supported by RGC (H.K.) and CRCG grants (H.K.U.) to A.O.L.W. Financial support from the Department of Zoology (H.K.U.) to L.H., G.F., and X.W. in the form of postgraduate studentships is also acknowledged.
Abbreviations: a.a., Amino acid; Ca2+/CaM kinase II, calmodulin-dependent protein kinase II; CaM, calmodulin; DIG, digoxigenin; EGFP, enhanced green fluorescence protein; FBS, fetal bovine serum; LSD, least significance difference; Luc, luciferase; ORF, open reading frame; PRL, prolactin; RACE, rapid amplification of cDNA ends; Sp1, specificity protein-1; SRIF, somatostatin; Tm, melting temperature; 3'UTR, 3'-untranslated region.
References
Weinstein H, Mehler EL 1994 Ca2+-binding and structural dynamics in the functions of calmodulin. Annu Rev Physiol 56:213–236
Means AR 2000 Calcium and calmodulin-mediated regulatory mechanisms. In: Conn PM, Means AR, eds. Principles of molecular regulation. Totowa, NJ: Humana Press; 207–231
Kumar RV, Panniers R, Wolfman A, Henshaw EC 1991 Inhibition of protein synthesis by antagonists of calmodulin in Ehrlich ascites tumor cells. Eur J Biochem 195:313–319
Schettini G, Florio T, Meucci O, Landolfi E, Cronin MJ, MacLeod RM 1987 Calmodulin modulates prolactin secretion in vitro: studies with calmodulin containing liposomes. Life Sci 41:2437–2444
Sletholt K, Haug E, Gordeladze J, Sand O, Gautvik KM 1987 Effects of calmodulin antagonists on hormone release and cyclic AMP levels in GH3 pituitary cells. Acta Physiol Scand 130:333–343
Bonafe NM, Sellers, JR 1998 Calmodulin-binding protein of the cytoskeleton. In: Van Eldik L, Watterson DM, eds. Calmodulin and signal transduction. San Diego: Academic Press; 348–396
Schmalzigaug R, Ye Q, Berchtold MW 2001 Calmodulin protects cells from death under normal growth conditions and mitogenic starvation but plays a mediating role in cell death upon B-cell receptor stimulation. Immunology 103:332–342
Takano H, Fukushi H, Morishima Y, Shirasaki Y 2003 Calmodulin and calmodulin-dependent kinase II mediate neuronal cell death induced by depolarization. Brain Res 962:41–47
Prostko CR, Zhang C, Hait WN 1997 The effects of altered cellular calmodulin expression on the growth and viability of C6 glioblastoma cells. Oncol Res 9:13–17
Luk SC, Ngai SM, Tsui SK, Fung KP, Lee CY, Waye MM 1999 In vivo and in vitro association of 14–3-3 epsilon isoform with calmodulin: implication for signal transduction and cell proliferation. J Cell Biochem 73:31–35
Walker SW, Wark JD, MacNeil S, Mellersh H, Brown BL, Tomlinson S 1984 Isolation, purification and cell-free synthesis of calmodulin from the pig anterior pituitary gland. Biochem J 217:827–832
Kakiuchi S, Yasuda S, Yamazaki R, Teshima Y, Kanda K, Kakiuchi R, Sobue K 1982 Quantitative determinations of calmodulin in the supernatant and particulate fractions of mammalian tissues. J Biochem 92:1041–1048
Aguila MC, McCann SM 1988 Calmodulin dependence of somatostatin release stimulated by growth hormone-releasing factor. Endocrinology 123:305–309
Honegger J, D’Urso R, Besser GM, Grossman AB 1991 Calcium and calmodulin mediation of growth hormone-releasing hormone release from the rat hypothalamus in vitro. Endocrinology 129:11–16
Davis JR, Hoggard N, Wilson EM, Vidal ME, Sheppard MC 1991 Calcium/calmodulin regulation of the rat prolactin gene is conferred by the proximal enhancer region. Mol Endocrinol 5:8–12
Izumi S, Iwashita M, Makino T, Saito S, Sakamoto S, Takeda Y, Nozawa S 1991 Phorbol ester-induced LH release in pituitary gonadotrophs: effects of antagonists of calmodulin and GnRH. Endocrinol Jpn 38:195–204
Martinez-Campos A, Hernandez RP, Forsbach G, Barrera-Saldana HA 1991 The stimulatory effect of estradiol 17- on prolactin mRNA is inhibited by anti-calmodulin drugs. Life Sci 48:2475–2485
White BA 1985 Evidence for a role of calmodulin in the regulation of prolactin gene expression. J Biol Chem 260:1213–1217
Le Roith D 2003 The insulin-like growth factor system. Exp Diabetes Res 4:205–212
Svensson J, Lonn L, Johannsson G, Bengtsson BA 2003 Effects of GH and insulin-like growth factor-I on body composition. J Endocrinol Invest 26:823–831
Dupont J, Khan J, Qu BH, Metzler P, Helman L, LeRoith D 2001 Insulin and IGF-I induce different patterns of gene expression in fibroblast NIH-3T3 cells: identification by cDNA microarray analysis. Endocrinology 142:4969–4975
Jaffe CA, Pan W, Brown MB, DeMott-Friberg R, Barkan AL 2001 Regulation of GH secretion in acromegaly: reproducibility of daily GH profiles and attenuated negative feedback by IGF-I. J Clin Endocrinol Metab 86:4364–4370
Weber MM, Melmed S, Rosenbloom J, Yamasaki H, Prager D 1992 Rat somatotroph insulin-like growth factor-II (IGF-II) signaling: role of the IGF-I receptor. Endocrinology 131:2147–2153
Niiori-Onishi A, Iwasaki Y, Mutsuga N, Oiso Y, Inoue K, Saito H 1999 Molecular mechanisms of the negative effect of insulin-like growth factor-I on growth hormone gene expression in MtT/S somatotroph cells. Endocrinology 140:344–349
Fletcher TP, Thomas GB, Dunshea FR, Moore LG, Clarke IJ 1995 IGF feedback effects on growth hormone secretion in ewes: evidence for action at the pituitary but not the hypothalamic level. J Endocrinol 144:323–331
Becker K, Stegenga S, Conway S 1995 Role of insulin-like growth factor I in regulating growth hormone release and feedback in the male rat. Neuroendocrinology 61:573–583
Moriyama S, Ayson FG, Kawauchi H 2000 Growth regulation by insulin-like growth factor-I in fish. Biosci Biotechnol Biochem 64:1553–1562
Tse MC, Vong QP, Cheng CH, Chan KM 2002 PCR-cloning and gene expression studies in common carp insulin-like growth factor-II. Biochim Biophys Acta 1575:63–74
Vong QP, Chan KM, Cheng CH 2003 Quantification of common carp IGF-I and IGF-II mRNA by real-time PCR: differential regulation of expression by GH. J Endocrinol 178:513–521
Fruchtman S, Gift B, Howes B, Borski R 2001 Insulin-like growth factor-I augments prolactin and inhibits growth hormone release through distinct as well as overlapping cellular signaling pathways. Comp Biochem Physiol B 129:237–242
Kajimura S, Uchida K, Yada T, Hirano T, Aida K, Gordon Grau E 2002 Effects of insulin-like growth factors (IGF-I and -II) on growth hormone and prolactin release and gene expression in euryhaline tilapia, Oreochromis mossambicus. Gen Comp Endocrinol 127:223–231
Baker DM, Davies B, Dickhoff WW, Swanson P 2000 Insulin-like growth factor I increases follicle-stimulating hormone (FSH) content and gonadotropin-releasing hormone-stimulated FSH release from Coho salmon pituitary cells in vitro. Biol Reprod 63:865–871
Fruchtman S, Jackson L, Borski R 2000 Insulin-like growth factor I disparately regulates prolactin and growth hormone synthesis and secretion: studies using the teleost pituitary model. Endocrinology 141:2886–2894
Chang JP, Johnson JD, Van Goor F, Wong CJ, Yunker WK, Uretsky AD, Taylor D, Jobin RM, Wong AOL, Goldberg JI 2000 Signal transduction mechanisms mediating secretion in goldfish gonadotropes and somatotropes. Biochem Cell Biol 78:139–153
Chang JP, Abele JT, Van Goor F, Wong AOL, Neumann CM 1996 Role of arachidonic acid and calmodulin in mediating dopamine D1- and GnRH-stimulated growth hormone release in goldfish pituitary cells. Gen Comp Endocrinol 102:88–101
Jobin RM, Neumann CM, Chang JP 1996 Roles of calcium and calmodulin in the mediation of acute and sustained GnRH-stimulated gonadotropin secretion from dispersed goldfish pituitary cells. Gen Comp Endocrinol 101:91–106
Wong AOL, Ng S, Lee EK, Leung RC, Ho WK 1998 Somatostatin inhibits (D-Arg6, Pro9-NEt) salmon gonadotropin-releasing hormone- and dopamine D1-stimulated growth hormone release from perifused pituitary cells of Chinese grass carp, Ctenopharyngodon idellus. Gen Comp Endocrinol 110:29–45
Zhou H, Wang X, Ko WK, Wong AOL 2004 Evidence for a novel intrapituitary autocrine/paracrine feedback loop regulating growth hormone synthesis and secretion in carp pituitary cells by functional interactions between gonadotrophs and somatotrophs. Endocrinology 145:5548–5559
Zhou H, Jiang Y, Ko WK, Li W, Wong AOL 2005 Paracrine regulation of growth hormone gene expression by gonadotrophin release in grass carp pituitary cells: functional implications, molecular mechanisms and signal transduction. J Mol Endocrinol 34:415–432
Toutenhoofd SL, Strehler EE 2000 The calmodulin multigene family as a unique case of genetic redundancy: multiple levels of regulation to provide spatial and temporal control of calmodulin pools Cell Calcium 28:83–96
Berchtold MW, Egli R, Rhyner JA, Hameister H, Strehler EE 1993 Localization of the human bona fide calmodulin genes CALM1, CALM2, and CALM3 to chromosomes 14q24–q31, 2p21.1-p21.3, and 19q13.2-q13.3. Genomics 16:461–465
Reid DG, MacLachlan LK, Gajjar K, Voyle M, King RJ, England PJ 1990 A proton nuclear magnetic resonance and molecular modeling study of calmidazolium (R24571) binding to calmodulin and skeletal muscle troponin C. J Biol Chem 265:9744–9753
Nojima H 1987 Molecular evolution of the calmodulin gene. FEBS Lett 217:187–190
Saimi Y, Kung C 2002 Calmodulin as an ion channel subunit. Annu Rev Physiol 64:289–311
Sola C, Barron S, Tusell JM, Serratosa J 1999 The Ca2+/calmodulin signaling system in the neural response to excitability. Involvement of neuronal and glial cells. Prog Neurobiol 58:207–232
Sola C, Barron S, Tusell JM, Serratosa J 2001 The Ca2+/calmodulin system in neuronal hyperexcitability. Int J Biochem Cell Biol 33:439–455
Gill KD, Gupta V, Sandhir R 2003 Ca2+/calmodulin-mediated neurotransmitter release and neurobehavioural deficits following lead exposure. Cell Biochem Funct 21:345–353
Zhang J, Suneja SK, Potashner SJ, Rameau GA, Chiu LY, Ziff EB 2004 Protein kinase A and calcium/calmodulin-dependent protein kinase II regulate glycine and GABA release in auditory brain stem nuclei: bidirectional regulation of neuronal nitric-oxide synthase phosphorylation at serine 847 by the N-methyl-D-aspartate receptor. J Neurosci Res 75:361–370
Le Roith D, Bondy C, Yakar S, Liu JL, Butler A 2001 The somatomedin hypothesis: 2001. Endocr Rev 22:53–74
Duval H, Rousseau K, Elies G, Le Bail PY, Dufour S, Boeuf G, Boujard D 2002 Cloning, characterization, and comparative activity of turbot IGF-I and IGF-II. Gen Comp Endocrinol 126:269–278
Solomon SS, Palazzolo MR, Smoake JA, Raghow RS 1995 Insulin-stimulated calmodulin gene expression in rat H-411E cells can be selectively blocked by antisense oligonucleotides. Biochem Biophys Res Commun 210:921–930
Solomon SS, Palazzolo MR, Takahashi T, Raghow R 1997 Insulin stimulates rat calmodulin I gene transcription through activation of Sp1. Proc Assoc Am Physicians 109:470–477
Yamashita S, Melmed S 1986 Effects of insulin on rat anterior pituitary cells. Inhibition of growth hormone secretion and mRNA levels. Diabetes 35:440–447
Prager D, Melmed S 1988 Insulin regulates expression of the human growth hormone gene in transfected cells. J Biol Chem 263:16580–16585
Weil C, Carre F, Blaise O, Breton B, Le Bail PY 1999 Differential effect of insulin-like growth factor I on in vitro gonadotropin and growth hormone secretions in rainbow trout (Oncorhynchus mykiss) at different stages of the reproductive cycle. Endocrinology 140:2054–2062
Pierce AL, Fukada H, Dickhoff WW 2005 Metabolic hormones modulate the effect of growth hormone (GH) on insulin-like growth factor-I (IGF-I) mRNA level in primary culture of salmon hepatocytes. J Endocrinol 184:341–349
Castillo J, Le Bail PY, Paboeuf G, Navarro I, Weil C, Fauconneau B, Gutierrez J 2002 IGF-I binding in primary culture of muscle cells of rainbow trout: changes during in vitro development. Am J Physiol 283:R647–R652(Longfei Huo, Guodong Fu, )