INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CA²⁺/calmodulin-dependent protein kinase ii (CaM kinase II) mediates many cellular responses to elevated Ca²⁺ in a wide variety of cells and tissues (33). CaM kinase II is involved in the regulation of ion channels, cytoskeletal dynamics, gene transcription, neurotransmitter synthesis, insulin secretion, and cell division (5a). In vascular smooth muscle tissues and cultured cells, CaM kinase II is implicated in the modulation of myosin light chain kinase sensitivity to Ca²⁺ and in regulation of cell migration, and it is involved in the control of Ca²⁺ channels, sarcoplasmic reticulum Ca²⁺-ATPase activity, and MAP kinase activation (2, 3, 10, 24, 37). CaM kinase II is activated by physiological Ca²⁺ elevations in vascular smooth muscle cells, and its inhibition decreases contraction and force maintenance by inhibiting MAP kinase activation and myosin light chain (LC20) phosphorylation (2, 18, 26). In cardiac muscle phosphorylation of phospholamban Thr¹⁷ by CaM kinase II in response to -adrenergic stimulation activates the sarcoplasmic reticulum Ca²⁺-ATPase, resulting in faster relaxation (31). In addition, the membrane potential and excitability of proximal colon smooth muscle cells are modulated by CaM kinase II regulation of delayed rectifier K⁺ currents and Ca²⁺-activated K⁺ channels (19, 20). These findings indicate that a number of proteins are substrates for CaM kinase II in smooth muscle and suggest that CaM kinase II can function at several points to enhance or attenuate the contractile response.

CaM kinase II holoenzymes are multimers composed of 6-12 kinase subunits arranged as a stacked pair of hexameric rings (33). The central core of the holoenzyme contains the COOH-terminal association domains of each subunit, with the variable domain and NH₂-terminal regulatory and catalytic domains extending outward (33). Four genes encode the kinase subunit isoforms (, , , ), and alternative splicing within the variable domain generates additional diversity (39). The - and -isoforms have narrow distributions, being restricted to neuronal and endocrine tissues, whereas the - and -isoforms are ubiquitously expressed within neuronal and nonneuronal tissues (8). The functional significance of many of the variable domains to the enzymatic activity of CaM kinase II is not clear. Variable domain region I regulates the cytoplasmic distribution of the -isoform and increases its affinity for Ca²⁺/calmodulin (CaM) but has no effect on the affinity of the _A isoform for Ca²⁺/CaM (21, 40). Variable domain region III is responsible for the nuclear localization of _B CaM kinase II (25, 40).

Ca²⁺/CaM binding to the holoenzyme kinase subunits results in the rapid phosphorylation of Thr²⁸⁶ (numbering based on the -isoform) on adjacent activated subunits (5a). Two consequences of Thr²⁸⁶ autophosphorylation are that 1) the rate of Ca²⁺/CaM dissociation in response to Ca²⁺ removal is decreased by several orders of magnitude, and 2) the kinase maintains activity toward its substrates in the absence of Ca²⁺/CaM (independent or autonomous activity) (5a). Thus the phosphorylation of target substrates by CaM kinase II in response to transient increases in cytosolic Ca²⁺ levels is sustained after the Ca²⁺ levels decrease. The Ca²⁺-independent activity of CaM kinase II plays an important role in the physiological response of cells to transient Ca²⁺ increases (33). Additional autophosphorylation in the CaM-binding domain at Thr^305/306 following Thr²⁸⁶ autophosphorylation prevents subsequent CaM binding and lowers total CaM kinase II activity levels (13, 14). Thr^305/306 autophosphorylation also promotes CaM kinase II disassociation from postsynaptic sites, providing a mechanism for regulating the subcellular distribution of autonomous CaM kinase II (30). The combined effects of these dual autophosphorylations result in complex activation of the holoenzyme in response to Ca²⁺ oscillations (7, 12, 13). The level of autonomous activity of CaM kinase II in a tissue lysate is an indication of prior activation of the enzyme because Thr²⁸⁶ autophosphorylation can only occur after Ca²⁺/CaM binding to the holoenzyme. However, because Thr^305/306 autophosphorylation regulates the number of kinase subunits that can bind Ca²⁺/CaM, the extent of Thr²⁸⁶ autophosphorylation is also regulated by Thr^305/306 autophosphorylation (12-14). Thus identical CaM kinase II holoenzymes with different levels of Thr²⁸⁶ and Thr^305/306 autophosphorylation will have different kinetics of CaM association and dissociation and of activation (12-14). However, the subunit composition of the CaM kinase II holoenzyme also regulates the kinetics of activation, autophosphorylation, and CaM dissociation in response to Ca²⁺ oscillations (5a, 7, 26). The holoenzyme structure of CaM kinase II and the effects of Thr²⁸⁶ autophosphorylation on enzyme activation led to the hypothesis that CaM kinase II can be activated to different levels in response to different Ca²⁺ oscillation frequencies (12). The in vitro study of immobilized CaM kinase II by De Koninck and Schulman provides strong evidence supporting this hypothesis (7).

The proximal colon and gastric fundus are representative phasic and tonic gastrointestinal smooth muscles, respectively, that are characterized by different Ca²⁺ signaling patterns and sensitivities to excitation-contraction coupling (5, 34). Because CaM kinase II activity levels are modulated by Ca²⁺ oscillations, we are investigating the characteristics of CaM kinase II expressed in fundus and proximal colon smooth muscle tissues to test the hypothesis that tissues with different Ca²⁺ signaling patterns express CaM kinase II holoenzymes with different activation properties. Because the subunit composition influences activation, we investigated the CaM kinase II isoforms expressed in fundus and proximal colon smooth muscle tissues by Western blot and real-time PCR analysis. We investigated the regulation of CaM kinase II activation by Thr²⁸⁶ and Thr^305/306 autophosphorylation by measuring the total and autonomous CaM kinase II activity levels and comparing the kinetics of generation of CaM kinase II autonomous activity in each smooth muscle tissue. Generation of autonomous CaM kinase II activity in fundus lysates required prior alkaline phosphatase treatment. Total CaM kinase II activity levels in fundus are also increased following alkaline phosphatase treatment. Our findings indicate that gastric fundus and proximal colon are characterized by CaM kinase II holoenzymes having distinct enzymatic characteristics and may be differentially regulated by Thr²⁸⁶ and Thr^305/306 autophosphorylation. Incubation of fundus and proximal colon smooth muscle tissues with acetylcholine (ACh) increased autonomous CaM kinase II activities. The CaM kinase II inhibitor KN-93 enhanced the generation of tone in fundus and inhibited phasic contractions in proximal colon smooth muscle tissues in response to ACh stimulation. Together, these results indicate that CaM kinase II is activated by contractile stimuli and modulates contractile force in fundus and proximal colon smooth muscle tissues in vivo.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Materials. Alkaline phosphatase and EDTA-free protease inhibitor tablets were purchased from Roche (Indianapolis, IN). CaM was obtained from Sigma (St. Louis, MO). [-³²P]ATP (6,000 Ci/mmol) was purchased from ICN (Costa Mesa, CA). Autocamtide-2 was obtained through United Biomedical Research (Seattle, WA). Taq DNA polymerase, dNTPs, and SuperScript II reverse transcriptase were purchased from Life Technologies (Gaithersburg, MD). Primers were synthesized by Life Technologies. Goat anti-CaM kinase II , , , and  antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-PO₄-Thr²⁸⁶ antibody (anti-active CaM kinase II) was purchased from Promega (Madison WI). This antibody recognizes the sequence surrounding the autophosphorylated Thr²⁸⁶ or Thr²⁸⁷ in CaM kinase II - and -subunits [MHRQET(PO₄)VDCLKKFN]. Alkaline phosphatase-conjugated rabbit anti-goat IgG antibodies were obtained from Chemicon (Temecula, CA). L-Phenylalanine and ACh were from Sigma. KN-93 and KN-92 were purchased from Biomol (Plymouth Meeting, PA) and Calbiochem (San Diego, Ca), respectively. CD-1 mice were purchased from Charles River (Cambridge, MA).

CaM kinase II activity assays. Fundus and proximal colon smooth muscle tissues were prepared from the stomachs and colons removed from adult CD-1 mice. Briefly, mice were anesthetized with CHCl₃, followed by cervical dislocation and removal of tissues as approved by the Institutional Animal Care and Use Committee. The fundus and proximal colon tissues were pinned out in a Sylgard-lined dish and washed with Ca²⁺-free Hanks' buffer (125 mM NaCl, 5.36 mM KCl, 15.5 mM NaOH, 0.336 mM Na₂HPO₄, 0.44 mM KH₂PO₄, 10 mM glucose, 2.9 mM sucrose, and 11 mM HEPES, pH 7.2). The mucosa and submucosa layers were removed with fine-tipped forceps. Fundus and proximal colon smooth muscle tissues were obtained as described above and pinned out in small dishes containing Krebs buffer (120 mM NaCl, 6 mM KCl, 15 mM NaHCO₃, 12 mM glucose, 3 mM MgCl₂, 1.5 mM NaH₂PO₄, and 3.5 mM CaCl₂, pH 7.2). For determining the effect of ACh on CaM kinase II activity, the tissues were equilibrated in Krebs buffer for 1 h at 37°C and then incubated at 37°C for various times in the absence or presence of 10 µM ACh or for 30 min with various ACh concentrations. KN-93 was added to the equilibrated tissues 20 min before the addition of ACh. After treatment, the tissues were collected, frozen in liquid nitrogen, and stored at 80°C until used for the CaM kinase II assays. When needed for assays, frozen tissues were homogenized at 4°C with a glass tissue grinder in lysis buffer (50 mM MOPS, 2% Nonidet P-40, 100 mM Na₄P₂O₇, 100 mM NaF, 250 mM NaCl, 3 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and a protease inhibitor tablet). Homogenates were centrifuged at 16,000 g for 15 min at 4°C. The supernatant was aliquoted and stored at 80°C until the kinase assay was performed. Kinase assays were done at least three times in triplicate from each tissue from three animals. Protein concentrations were determined by using the Bradford assay with bovine gamma globulin as standard.

Dephosphorylation of fundus and proximal colon smooth muscle tissue lysates. Smooth muscle tissues were obtained and homogenized as described in CaM kinase II activity assays, with the exception that phosphatase inhibitors (Na₄P₂O₇ and NaF) were omitted from the homogenization buffer. Lysates were incubated with calf intestinal alkaline phosphatase at 37°C for 40 min by using a range of 0.2-1.0 unit of enzyme per microgram of lysate. Phenylalanine was added to the samples (5 mM final concentration) to inhibit alkaline phosphatase (17). Assays for total and autonomous CaM kinase II activity were performed on control and dephosphorylated lysates as described above.

SDS-PAGE and Western blot analysis of CaM kinase II from proximal colon and fundus smooth muscle tissues. Fundus and proximal colon smooth muscle tissue lysates were obtained from CD-1 mice, and protein concentrations were determined as described in CaM kinase II activity assays. Tissue lysate proteins were separated by SDS-PAGE (7.5%) and transferred to nitrocellulose by Western blotting. The amount of lysate protein per lane is indicated. The blots were incubated with primary and secondary antibodies, washed, and processed for image detection by using the Western-Light chemiluminescence system from Tropix (Bedford, MA). The CaM kinase II antibodies were used at 1:200 dilutions, and the alkaline phosphatase-conjugated rabbit anti-goat IgG antibody was used at a 1:5,000 dilution. Protein bands were visualized with a charge-coupled device camera-based detection system (Epi Chem II; UVP Laboratory Products). The collected images were opened in Adobe Photoshop and inverted for analysis. Densitometry was carried out by using Un-Scan-It software from Silk Scientific. 316 CaM kinase II was baculovirus-expressed and purified by CaM-Sepharose chromatography as described previously (23). Autophosphorylation and alkaline phosphatase treatment of purified 316 CaM kinase II was carried out as described above.

RT-PCR of , , , and  CaM kinase II isoforms from proximal colon and fundus smooth muscle tissues. Fundus and proximal colon smooth muscle tissues were obtained from CD-1 mice, and smooth muscle cells from each tissue were enzymatically dispersed and collected as described (9). Total RNA was purified by using the Epicenter MasterPure RNA purification kit. First-strand cDNA was synthesized from 4.0 µg of total RNA by using random primers and SuperScript II Moloney murine leukemia virus RNase H reverse transcriptase (Life Technologies). PCR was performed on 2 µl of cDNA with a final concentration of 1.5 mM MgCl₂, 0.2 mM dNTPs, 0.5 µM primers, and 0.025 unit of Taq DNA polymerase (Life Technologies). First-strand cDNA preparations were monitored for the absence of genomic DNA by using intron-flanking -actin primers (9). To identify splice variants, we used primers designed to flank the variable region of each CaM kinase II isoform. The  and  forward primers were designed at nucleotides encoding for amino acids WDTVTPE (5'-GGGACACAGTGACACCTGAA-3' and 5'-GGACACAGTCACTCCTGAA-3', respectively), the  reverse primer was complementary to bases encoding for amino acids EALGNLV (5'-CACTAAGTTGCCCAATGCTTC-3'), and the  reverse primer was complimentary to bases encoding for amino acids LGNLVE (5'-CTCCACGAGGTTACCAACC-3'). The forward  primer corresponds to the nucleotides encoding amino acids LLASKL (5'-GTTGCTGGCTTCGAAGCTC-3'), and the  reverse primer is complementary to the nucleotide sequence encoding amino acids GLDFHR (5'-ATCGATGAAAGTCCAGCCC-3'). The forward  primer corresponds to the nucleotides encoding amino acids EVLRKE (5'-CGAGGTCCTTCGGAAGGAGG-3'), and the  reverse primer is complementary to the nucleotide sequence encoding amino acids EALGNL (5'-GACCAGGTTGCCCAGAGCTT-3'). The PCR profile for the  primers was as follows: 30 s of denaturation at 94°C, 30 s of annealing at 63°C, and 1 min of extension at 72°C for 40 cycles, with a final 5-min extension at 72°C. The PCR profile for the , , and  primers was the same as that for the  primers, except for annealing at 55°C, 61°C, and 63°C, respectively. The PCR products were separated in 4.0% agarose gels and visualized with GelStar stain (BioWhittaker Molecular Applications; Rockland, CA.). The fragments were excised from the gel, isolated with Micropure Separators (Amicon), and cloned with the Topo II TA cloning kit (Invitrogen). Putative positive clones were identified by diagnostic PCR. The plasmids were subsequently purified by using the Quantum Prep plasmid miniprep kit (Bio-Rad), and the cloned PCR fragments were sequenced using an ABI dye terminator cycle sequencer (Applied Biosystems, Foster City, CA.).

Real-time PCR analysis of , , , and  CaM kinase II isoforms from proximal colon and fundus smooth muscle tissues. Total RNA was purified as described above, and the pan , , , or  primers were used in the real-time PCR reactions. Relative quantitation of the , , , or  CaM kinase II isoforms was performed by using SYBR green detection real-time PCR. cDNA from CD-1 mouse brain cerebellum was initially diluted 1:20 and then serially diluted 1:2 for six dilutions, with 2 µl subsequently used as template for the standard curves of the CaM kinase II isoforms and the internal standard, GAPDH. Proximal colon and fundus cDNA samples were diluted 1:20 and 1:40, with 2 µl used as template in triplicate reactions for the CaM kinase II isoforms and GAPDH. The SYBR green master mix (Applied Biosystems) and the gene-specific pan primers at final concentrations of 0.3 µM were used for amplification and detection of product on an ABI Prism 7700 sequence detection system (Applied Biosystems). The relative quantitation was then determined by the standard curve method (ABI, User Bulletin no. 2).

Mechanical responses of fundus and proximal colon smooth muscle tissues. A standard organ bath technique was employed to measure changes in isometric force provided by fundus muscle strips. The mucosa was removed from the gastric fundus by sharp dissection, and strips of muscle (~6 × 3 mm) were isolated. One end of the tissue was attached to a fixed mount, and the opposite end to a Fort 10 isometric strain gauge (WPI, Sarasota, FL). The muscles were immersed in organ baths maintained at 37 ± 0.5°C with oxygenated Krebs solution (KRB). A resting force of 400 mg was applied to set the muscles at optimum length (data not shown). This was followed by an equilibration period of 1 h, during which time the bath was continuously perfused with oxygenated KRB. Mechanical responses were recorded on a personal computer running Acqknowledge 3.2.6 (BIOPAC Systems, Santa Barbara, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Identification of CaM kinase II isoforms in fundus and proximal colon smooth muscle tissues. Isoform-specific primer pairs that anneal to sequences common to each reported isoform splice variant were used in RT-PCR analysis to first determine which mRNAs are present for , , , and  CaM kinase II in murine proximal colon and fundus smooth muscle tissues. As shown in Fig. 1A, PCR products were obtained from the , , , and  CaM kinase II primer pairs. Sequencing confirmed that the PCR products obtained with each CaM kinase II isoform-specific primer pair corresponded to , , , and  CaM kinase II, indicating that mRNAs for all four CaM kinase II isoforms are present in murine proximal colon and fundus smooth muscle tissues. It is well established that the expression of  and  CaM kinase II is restricted to neuronal and neuroendocrine cells and that  and  CaM kinase II are expressed in cardiac, skeletal, and vascular smooth muscle cells (5a, 29, 32, 38). To determine which CaM kinase II isoforms are expressed in gastrointestinal smooth muscle cells, cDNA from isolated fundus and proximal colon smooth muscle cells was analyzed by RT-PCR. With the use of cDNA from collected fundus smooth muscle cells, no PCR products were obtained with the  and  CaM kinase II primers (Fig. 1C). In contrast, PCR products were obtained with the  and  CaM kinase II primers. These findings indicate that murine fundus and proximal colon smooth muscle cells express  and  CaM kinase II. These findings also suggest that the  and  CaM kinase II detected in the fundus and proximal colon smooth muscle tissue samples are most likely expressed in enteric neurons.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   CaM kinase II isoforms detected by RT-PCR in murine fundus and proximal colon smooth muscle cells and tissues. PCR products were obtained from fundus and proximal colon smooth muscle tissues (A and B) or freshly dispersed and collected fundus smooth muscle cells (C) by using primers specific for , , , or  Ca2+/calmodulin-dependent protein kinase II (CaM kinase II). Aliquots of the PCR reactions were separated by agarose gel (4%) electrophoresis and visualized by SYBR green staining.

Relative expression of CaM kinase II isoforms in fundus and proximal colon smooth muscle tissues. Real-time PCR analysis was used to quantitate the relative message levels of the , , , and  CaM kinase II expressed in murine fundus and proximal colon smooth muscle tissues. The PCR products were amplified by the appropriate isoform-specific pan primers, and amounts produced were measured relative to GAPDH. As shown in Fig. 2, in fundus and proximal colon smooth muscle tissues,  CaM kinase II message levels are more abundant than  CaM kinase II message levels, relative to GAPDH message levels. The : ratio in fundus is ~1.66:1, whereas the : ratio in proximal colon smooth muscle tissues is ~1.3:1. Relative to GAPDH and to  and  CaM kinase II, the message levels of the - and -isoforms are present at extremely low levels in fundus and proximal colon smooth muscle tissues. These findings are consistent with the RT-PCR results (Fig. 1C) showing that message for the - and -isoforms was not amplified from isolated smooth muscle cells.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Real-time PCR analysis of mRNA expression levels of , , , and  CaM kinase II relative to GAPDH in murine fundus and proximal colon smooth muscle tissues. Results represent means ± SD of 6 real-time PCR reactions for each CaM kinase II isoform-specific primer pair. Three fundus and proximal colon smooth muscle tissues obtained from 3 different animals were used to prepare cDNA.

Detection of CaM kinase II isoforms in murine fundus and proximal colon smooth muscle tissues. The results of the SDS-PAGE and Western blot analysis of gastric fundus and proximal colon smooth muscle tissue lysates using CaM kinase II isoform-specific antibodies are shown in Fig. 3. Murine brain lysates were used as a positive control for  and  CaM kinase II expression (5a, 26). In agreement with previous reports, a strong signal at ~50 kDa was generated from brain lysate by using the anti- CaM kinase II antibody (5a, 26). Similarly, a protein band at ~60 kDa was observed from brain lysates by using the anti- CaM kinase II antibody. The  CaM kinase II staining is less intense than the  CaM kinase II staining from brain lysates. These results are consistent with previous findings that  CaM kinase II is the predominant isoform expressed in neuronal tissues (5a, 26). The anti- CaM kinase II antibody generated a modest signal at 50 kDa from proximal colon lysates and an almost undetectable signal from fundus lysates. Protein staining of fundus and proximal colon smooth muscle tissue lysates using the anti- CaM kinase II antibody was not detected. These results indicate that proximal colon smooth muscle tissue is characterized by a higher level of  CaM kinase II expression than fundus and suggest that  CaM kinase II is not expressed or is expressed at levels below the limits of detection in both smooth muscle tissues (see below). It should be noted here that the amount of protein from fundus and proximal colon lysates in each lane is twice the amount of the protein in the lanes with brain lysates. These results indicate that the  and  CaM kinase II expression levels in fundus and proximal colon smooth muscle tissues are lower than the expression levels found in brain.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   CaM kinase II isoforms detected by SDS-PAGE and Western blotting in murine fundus and proximal colon smooth muscle tissues. Specificity of the isoform-specific anti-CaM kinase II antibodies used are indicated (brain, 30 µg of murine brain lysate; aorta, 60 µg of rat aorta lysate; heart, 60 µg of rat heart lysate; fundus, 60 µg of fundus smooth muscle lysate; and proximal colon, 60 µg of proximal colon smooth muscle lysate). Results are representative of 3 Western blotting experiments from fundus and proximal colon smooth muscle tissue lysates obtained from several animals.

Identification of CaM kinase II isoform splice variants in fundus and proximal colon smooth muscle tissues. RT-PCR analysis was carried out by using isoform-specific primers flanking the variable region of CaM kinase II to identify the , , , and  splice variants expressed in murine proximal colon and fundus smooth muscle tissues. After agarose gel electrophoresis, the PCR products were excised from the gels and cloned by using Topo cloning vectors. As shown in Fig. 4A, four similarly sized PCR products ranging from ~550 to 750 bp and 450 to 700 bp were obtained from fundus and proximal colon smooth muscle tissues with the - and -specific primers, respectively. Sequencing the PCR products obtained with the -isoform-specific primers identified the 550-, 650-, 700-, and 750-bp PCR products as _C, _J, _B, and _I, respectively. In contrast, although four PCR products are visible in the gel, following cloning and sequencing, all five  CaM kinase II splice variants were identified. Sequencing identified the 450-, 550-, 600-, and 700-bp PCR products from the -isoform-specific primers as _C, _B/_E, _D, and _A, respectively. The predicted sizes of the _B and _E amplicons differ from each other by only nine base pairs; thus these two amplicons are unlikely to be separated by the agarose gel electrophoresis. In addition, by using the variable domain-flanking primers specific for  CaM kinase II, we amplified two PCR products of ~285 and 318 bp from fundus and proximal colon smooth muscle tissues and determined them to be  and 33 CaM kinase, respectively, by sequence analysis. Five PCR products were obtained by using the variable domain-flanking primers specific for  CaM kinase II. After cloning and sequencing, on the basis of the predicted sizes of the splice variant amplicons, the 377-, 404-, 449-, and 635-bp products correspond to '_e, _e, ', and  CaM kinase II, respectively. Sequence analysis indicates that the 318-bp amplicon corresponds to 33, indicating that the -isoform-specific primers also annealed to the complementary sequences of 33 CaM kinase II. The nomenclature of the , , , and  splice variants is from Tombes and Krystal (39). _I and _J were previously identified in rabbit liver as novel splice variants and were named _H and _I, respectively (36). Our results are the first demonstration of the expression of these two  splice variants outside the liver, indicating that they have a wider tissue distribution. Figure 4B shows a chart of the variable region domain structure of these  and  splice variants and their corresponding predicted amino acid sequences (39).

View larger version (38K): [in this window] [in a new window]   Fig. 4.   CaM kinase II isoform splice-variants detected by RT-PCR in murine fundus and proximal colon smooth muscle tissues. A: PCR products obtained using primers flanking the variable regions of CaM kinase II , , , and . Aliquots of the PCR reactions were separated by agarose gel (4%) electrophoresis and visualized by SYBER green staining: lane 1, fundus; lane 2, proximal colon. Variable domain structure and amino acid sequences of CaM kinase II  and  (B) and  and  (C) isoform splice variants detected in murine fundus and proximal colon smooth muscle tissues are shown.

Total and autonomous CaM kinase II activity in fundus and proximal colon smooth muscle lysates. Because the subunit composition and levels of autophosphorylation influence the kinase activity of the CaM kinase II holoenzyme, we compared total and autonomous CaM kinase II activity from murine fundus and proximal colon smooth muscle tissue lysates using the peptide substrate autocamtide-2. Addition of the specific CaM kinase II inhibitor KN-93 (1 µM) to the tissue lysates inhibited 90% of the total kinase activity toward autocamtide-2 (data not shown), indicating that, similar to previous reports, the autocamtide-2 kinase activity measured in the tissue lysates is due to CaM kinase II (26, 35). Total CaM kinase II activity in fundus smooth muscle lysates is threefold lower than the activity in proximal colon smooth muscle tissue lysates (1.09 ± 0.19 vs. 3.65 ± 0.74 pmol · min¹ · µg¹, respectively). In contrast, autonomous CaM kinase II activity in fundus is 3.5-fold higher than the autonomous activity in proximal colon smooth muscle tissue (0.32 ± 0.07 vs. 0.09 ± 0.03 pmol · min¹ · µg¹, respectively). Thus, as a percentage of total CaM kinase II activity, autonomous CaM kinase II activity is high in fundus smooth muscle tissues (28.9 ± 3.35%) but low in proximal colon smooth muscle tissues (3.6 ± 0.6%). These findings indicate that fundus smooth muscle tissue contains higher levels of autonomous CaM kinase II activity than proximal colon smooth muscle tissue.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Western blots of autophosphorylated CaM kinase II in murine fundus (A) and proximal colon tissues (B). Autophosphorylated CaM kinase II was visualized from tissue lysates by using an antibody that recognizes PO4-Thr286 of CaM kinase II. Lanes 1, 2, 3, and 4 contain 10, 20, 40, and 60 µg of lysate protein, respectively. Total CaM kinase II was detected by using an antibody that recognizes all four CaM kinase II isoforms. Results are representative of 3 Western blotting experiments from fundus and proximal colon smooth muscle tissue lysates obtained from several animals.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Generation of autonomous CaM kinase II activity in murine fundus and proximal colon smooth muscle lysates. Smooth muscle tissues were obtained, and autonomous CaM kinase II activity was measured as described in MATERIALS AND METHODS. Results are averages ± SD from 3 assays in triplicate from each tissue from 3 animals (, fundus; , proximal colon).

Activation of fundus smooth muscle CaM kinase II by alkaline phosphatase. A simplified diagram of the current model of CaM kinase II regulation by autophosphorylation is diagrammed in Fig. 7. Ca²⁺/CaM can cause little or no additional Thr²⁸⁶ autophosphorylation and autonomous activity as the number of Thr²⁸⁶-autophosphorylated kinase subunits of a holoenzyme increases (Fig. 7, II and III). This model and our results showing that autonomous CaM kinase II activity in fundus smooth muscle lysates is 30% of the total CaM kinase II activity and is largely resistant to generation of additional autonomous activity suggest that fundus smooth muscle CaM kinase II may be almost fully Thr²⁸⁶ phosphorylated (Fig. 7, III). Evidence in support of this conclusion is provided by the results of the experiments comparing autonomous CaM kinase II activity before and after alkaline phosphatase treatment. We first confirmed that alkaline phosphatase dephosphorylates Thr²⁸⁶-autophosphorylated CaM kinase II using autophosphorylated 316 CaM kinase II. As shown in Fig. 8, lane 1, the anti-PO₄-Thr²⁸⁶ antibody generated a strong signal toward autophosphorylated 316 CaM kinase II. However, no signal was detected following incubation of autophosphorylated 316 CaM kinase II with alkaline phosphatase (Fig. 8, lane 2). These results demonstrate that alkaline phosphatase dephosphorylates Thr²⁸⁶. Next, we examined the ability of alkaline phosphatase to dephosphorylate CaM kinase II in fundus and proximal colon smooth muscle lysates and the ability of CaM kinase II to undergo Thr²⁸⁶ autophosphorylation following alkaline phosphatase treatment of fundus and proximal colon smooth muscle lysates. Lysates were incubated with alkaline phosphatase, followed by phenylalanine (5 mM final concentration) to inhibit alkaline phosphatase, as previously reported (17). The lysates were then incubated CaM, CaCl₂, and ATP to allow Thr²⁸⁶ autophosphorylation to occur. No signal was detected from alkaline phosphatase-treated fundus and proximal colon smooth muscle lysates with the anti-PO₄-Thr²⁸⁶ antibody (Fig. 8B, lanes 1 and 4). However, strong signals were generated from the lysates that were incubated alkaline phosphatase, followed by phenylalanine, and CaM, CaCl₂, and ATP (Fig. 8, lanes 2 and 3). These results demonstrate that alkaline phosphatase dephosphorylates Thr²⁸⁶ in fundus and proximal colon smooth muscle tissue lysates. These results also demonstrate the ability of CaM kinase II in fundus and proximal colon smooth muscle tissue lysates to undergo Thr²⁸⁶ autophosphorylation following inhibition of alkaline phosphatase by 5 mM phenylalanine.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Simplified model of CaM kinase II regulation by autophosphorylation. The 6 lines connected at the center represent a hexameric CaM kinase II holoenzyme. State I represents the inactive, nonphosphorylated enzyme with the autoinhibitory domain bound to the catalytic site in the absence of Ca2+/calmodulin (CaM). In the presence of excess Ca2+/CaM, each kinase subunit binds Ca2+/CaM and becomes autophosphorylated on Thr286. Maximum Ca2+/CaM-stimulated activity (total activity) of the holoenzyme is achieved because all 6 kinase subunits are activated (state II). Thr286 autophosphorylation and kinase activity are maintained after Ca2+/CaM is removed from the holoenzyme (state III). Maximum Ca2+-independent activity (autonomous activity) is achieved because all 6 kinase subunits are autophosphorylated on Thr286 (state III). This activity typically ranges from 20 to 80% of the activity of the enzyme in state II (5a). In the absence of Ca2+/CaM, the state III holoenzyme can eventually revert back to state I because of phosphatase (ppases) activity or can undergo the second burst of autophosphorylation on Thr305/306 (state IV). In state IV, autonomous activity is unaffected but Ca2+/CaM binding is blocked. Because the state V holoenzyme is insensitive to Ca2+/CaM, its activity in the presence of Ca2+/CaM is equal to the autonomous activity (states III and IV).

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Effect of alkaline phosphatase on autophosphorylated CaM kinase II from proximal colon and fundus smooth muscle lysates. A: purified 316 CaM kinase II was autophosphorylated and aliquots were incubated with alkaline phosphatase (1 U) as described in MATERIALS AND METHODS. Untreated (lane 1) and alkaline phosphatase-treated (lane 2) samples (1 µg/lane) were separated by SDS-PAGE (10%) and electroblotted onto nitrocellulose. Western blots were probed with rabbit anti-PO4-Thr286 antibody, and specific staining was detected by enhanced chemiluminescence as described in MATERIALS AND METHODS. B: alkaline phosphatase-treated fundus (lane 1) and proximal colon (lane 4) tissue lysates. After alkaline phosphatase treatment, fundus (lane 2) and proximal colon (lane 3) tissue lysates were incubated with 5 mM phenylalanine, followed by CaCl2, CaM, and ATP to allow autophosphorylation (40 µg protein/lane).

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effect of alkaline phosphatase on autonomous CaM kinase II activity in fundus smooth muscle lysates. Fundus smooth muscle lysates were obtained and incubated with alkaline phosphatase as described in MATERIALS AND METHODS. Alkaline phosphatase was inhibited with phenylalanine (5 mM), and autonomous CaM kinase II activity in untreated and treated lysates was measured as described in MATERIALS AND METHODS. Results are averages ± SD from 3 assays in triplicate from each tissue from 3 animals (, without alkaline phosphatase; , with alkaline phosphatase).

1

1

1

1

1

1

1

1

Activation of CaM kinase II in fundus and proximal colon smooth muscle tissues by ACh. CaM kinase II is involved in the generation of contraction and maintenance of force in vascular smooth muscle (18, 26). In addition, the in situ Ca²⁺ dependence for CaM kinase II activation in cultured vascular smooth muscle cells falls within the range of cytosolic Ca²⁺ concentration increases induced by angiotensin II (1). However, our findings that CaM kinase II from fundus smooth muscle lysates is unresponsive to Ca²⁺/CaM activation compared with the enzyme from proximal colon smooth muscle lysates suggest that CaM kinase II in fundus and proximal colon smooth muscles may show different sensitivities to physiological stimuli that increase cytosolic Ca²⁺ levels in these tissues. Because ACh is the neurotransmitter primarily responsible for Ca²⁺ mobilization and contraction of gastrointestinal smooth muscle tissues, we assessed the ability of ACh to activate CaM kinase II in fundus and proximal colon smooth muscle tissues. Fundus and proximal colon smooth muscle tissues were cultured in the absence or presence of 10 µM ACh, as described in MATERIALS AND METHODS. CaM kinase II activation was determined by measuring autonomous CaM kinase II activity in lysates from control and ACh-treated tissues. As shown in Fig. 10, CaM kinase II autonomous activity levels were elevated in a time- and dose-dependent manner in ACh-treated fundus and proximal colon tissues. After a 45-min incubation of fundus smooth muscle tissues with 10 µM ACh, autonomous CaM kinase II activity increased from 31 to 54% of total CaM kinase II activity. Similarly, incubation of proximal colon smooth muscle tissues with 10 µM ACh increased autonomous CaM kinase II activity from 7 to 19% of total CaM kinase II activity (Fig. 10A). In addition, autonomous CaM kinase II activity levels in both smooth muscle tissues increased with increasing ACh concentrations and plateaued between 5 and 10 µM ACh (Fig. 10B). Preincubation of the smooth muscle tissues with KN-93 (1 µM) prevented the ACh-induced increases in CaM kinase II autonomous activities (data not shown). These findings indicate that contractile stimuli activate CaM kinase II in fundus and proximal colon smooth muscle tissues.

View larger version (13K): [in this window] [in a new window]   Fig. 10.   Activation of CaM kinase II in murine fundus and proximal colon smooth muscles by ACh. A: smooth muscle tissues were obtained and incubated in Krebs buffer at 37°C without or with 10 µM ACh for the indicated times. Autonomous CaM kinase II activity was measured in lysates from untreated and treated tissues (, fundus; , proximal colon) as described in MATERIALS AND METHODS. B: smooth muscle tissues (open bars, fundus; filled bars, proximal colon) were incubated in Krebs buffer at 37°C for 30 min without or with the indicated ACh concentrations. Results are averages ± SD from 3 assays in triplicate from each tissue from 3 animals.

Effects of KN-93 on contractions of gastric fundus and proximal colon smooth muscles. Because ACh activated CaM kinase II, the contribution of CaM kinase II to the contractile activity of the circular muscle layers of gastric fundus and proximal colon was determined. ACh (10 µM) caused the contractile tone of fundus smooth muscle to increase by 4.11 ± 0.8 mN (Fig. 11A). This increase in tone was reversible upon washout of ACh. Perfusion of muscle strips with KN-93 (5 µM) did not produce any resolvable change in basal tension. After a 30-min perfusion period with KN-93, the increase in contractile force produced by ACh increased to 7.9 ± 1.3 mN (Fig. 11B). Thus preincubation with KN-93 caused an ~90% increase in the amplitude of contractile force elicited by ACh. Perfusion of the proximal colon with ACh (10 µM) produced an increase in the tone of the circular muscle layer by 11.1 ± 1.7 mN and also increased the amplitudes of the phasic contractions (Fig. 11C). In contrast to the circular muscle layer of the fundus, incubation of proximal colon circular muscle with 1 µM KN-93 decreased spontaneous mechanical activity and reduced the tone produced by ACh by 2.9 ± 0.4 mN. KN-93 also reduced the steady-state amplitudes of ACh-induced phasic contractions to (Fig. 11D). The inactive KN-93 analog KN-92 (5 µM) had no effect on basal and ACh-induced contractions of fundus and proximal colon smooth muscles (data not shown). These findings indicate differential roles for CaM kinase II in regulating contractile activity of gastric fundus and proximal colon smooth muscle tissues.

View larger version (16K): [in this window] [in a new window]   Fig. 11.   Mechanical responses of fundus or proximal colon smooth muscle to exogenous ACh before and after incubation with KN-93. ACh (10 µM) produced a sustained contraction of fundus muscle that was reversible upon washout (A). After 30-min perfusion with KN-93 (5 µM), the contractile response to ACh was significantly enhanced (B). Perfusion of proximal colon with ACh (10 µM) increased the tone and steady-state amplitudes of contractions, which was reversed upon washout (C). After perfusion with KN-93 (1 µM), tonic contraction and steady-state amplitudes of contractions were significantly decreased (D). These recordings are representative of 6 strips from 2 animals for each tissue.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The holoenzyme structure and subunit composition of CaM kinase II modulate its activation in response to Ca²⁺ oscillations (7). Although each subunit is independently activated by Ca²⁺/CaM, the kinase subunits regulate each other by intersubunit autophosphorylation reactions on Thr²⁸⁶ and Thr^305/306 (28, 33). The number of Thr²⁸⁶-autophosphorylated subunits determines the sensitivity of the CaM kinase II holoenzyme to Ca²⁺/CaM stimulation, and the level and duration of autonomous activity (7). Different frequencies and amplitudes of Ca²⁺ oscillations give rise to different levels of Thr²⁸⁶ autophosphorylation and to CaM kinase II holoenzymes with different levels of autonomous activity and sensitivities to Ca²⁺/CaM stimulation (7). These findings suggest that tissues and cells having different Ca²⁺ oscillation patterns may express CaM kinase II holoenzymes characterized by different subunit compositions and levels of autophosphorylation. To test this hypothesis, we have initiated studies to characterize CaM kinase II in gastric fundus and proximal colon smooth muscle tissues. The fundus and proximal colon are tonic and phasic gastrointestinal smooth muscle tissues, respectively, with distinct Ca²⁺ signaling characteristics (5, 34). Tonic smooth muscles generally have higher resting Ca²⁺ levels compared with phasic smooth muscles (30). These differences in Ca²⁺ signaling underlie the cyclic depolarizations and repolarizations that determine the phasic contractile activity of the colon and also the tonic activity of the gastric fundus due to neural and hormonal regulation (16, 34).

RT-PCR analysis demonstrated the presence of message for , , , and  CaM kinase II in fundus and proximal colon smooth muscle tissues (Fig. 1). In contrast, in isolated smooth muscle cells, the RT-PCR analysis shows only  and  CaM kinase II expression (Fig. 1C). These findings indicate that smooth muscle cells of the digestive tract express  and  CaM kinase II. These findings also suggest that  and  CaM kinase II are expressed in the enteric neurons, because it has been previously established that  and  expression is restricted to neurons (5a). A strong  CaM kinase II signal was obtained from brain lysates, whereas weaker signals were generated from fundus and proximal colon smooth muscle lysates. Although we detected  CaM kinase II in Western blots of murine brain lysates (Fig. 3), no signal was obtained from fundus and proximal colon smooth muscle tissue lysates. It is well established that CaM kinase II message levels correlate with protein levels (4). Thus the lack of a  CaM kinase II signal from fundus and proximal colon smooth muscle lysates is most likely due to a level of expression that is below the limits of detection by Western blot analysis. Indeed, the real-time PCR results indicate that the expression of CaM kinase II  and  expression is >100-fold lower than  and  expression in fundus and proximal colon smooth muscle tissues. In the brain, CaM kinase II mRNA is abundant and protein expression levels approach 2% of total (33). Similarly, immunohistochemical results suggest that  and  CaM kinase II may be expressed at high levels in enteric neurons (19). However, neuronal mRNA represents a small minority of the total mRNA purified from the fundus and proximal colon smooth muscle tissues, with the vast majority of mRNA purified from the smooth muscle cells.

Several subtypes of each CaM kinase II isoform are generated by alternative splicing (29, 34). Using RT-PCR analysis, we found the same two , four , five , and four  CaM kinase II isoform splice variants expressed in murine fundus and proximal colon smooth muscle tissues: _B, _C, _I, _J, _A, _B, _C, _D, _E, , ₃₃, , ', _e, and '_e. These results do not support the hypothesis that tissues having different Ca²⁺ oscillation patterns express different CaM kinase II subunit isoforms. However, the real-time PCR analysis indicates that both fundus and proximal colon have slightly different : ratios. We found : ratios of ~1.66:1 and 1.3:1 in fundus and proximal colon smooth muscle tissues, respectively. Although these differences are less than twofold, they may have significant functional implications for the CaM kinase II holoenzyme subunit compositions. For example, a heptameric CaM kinase II holoenzyme with a 1.3:1 : ratio indicates that four subunits are  and three are . In contrast, a 1.66:1 : ratio suggests an octameric CaM kinase II holoenzyme with five  and three  subunits. We are currently carrying out coimmunoprecipitation experiments to determine the presence of  and  CaM kinase II heteromeric holoenzymes in fundus and proximal colon smooth muscle tissues.

Consistent with the real-time PCR results and previous reports of CaM kinase II expression in cardiac and vascular smooth muscle, Western blot analysis indicates that  and  CaM kinase II protein expression are also prevalent in murine fundus and proximal colon smooth muscle tissues. The apparent mass (~60 kDa) of the protein bands stained with the anti- CaM kinase II antibody suggest that _A is the predominant subtype expressed in murine fundus and proximal colon smooth muscles. The single protein band detected with the anti- CaM kinase II antibody is probably