| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Cardiac Surgery, Department of Surgery; and Departments of 2Anatomy and Neurobiology and 3Physiology and Biophysics, Dalhousie University, Halifax, Nova Scotia, Canada
Submitted 11 June 2007 ; accepted in final form 25 September 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
heat shock protein; confocal microscopy; mitogen-activated protein kinase inhibitors
Insulin also stimulates expression of the highly inducible 70-kDa heat shock protein (Hsp), Hsp70, in hep3B/T2 cells (66). We recently reported that Hsp70 is induced by insulin at a relatively low level in rat heart at 6 h after insulin treatment. Insulin increased the expression of Hsp70 in rat heart via the activation of heat shock factor-1 (HSF-1) and appeared to modulate the localization of Hsp70 to cardiomyocyte membranes in rat heart (42, 43).
Hsp70 is the highly inducible member of the 70-kDa family of Hsps. Whereas many Hsps are constitutive and have functions in normal cell homeostasis, Hsp70 is at very low levels in most cells and tissues under normal conditions. Many nonlethal but noxious stimuli induce a high expression of Hsp70 and other Hsps. Subsequent to the induction and expression of various Hsps, cells and tissues have a remarkable resistance to further metabolic injury (4, 15, 24, 31). For example, 24 h after heat shock treatment, hearts have high levels of Hsp70 and enhanced recovery of myocardial contractility after ischemic injury (13, 29). Transgenic overexpression of rat and human Hsp70 provides strong evidence for a direct role in protection of the mouse myocardium from ischemic injury (47, 60).
In general, Hsps, acting as molecular chaperones, regulate folding of nascent proteins; participate in refolding or the renaturation of misfolded, damaged, or denatured proteins; stabilize structural proteins; and facilitate the translocation of proteins across membranes among cellular compartments (30, 46). Several Hsps prevent aggregation of protein and target unstable or damaged protein for degradation. Hsp70 also suppresses intracellular apoptotic signaling pathways (53).
Hsp27 is constitutively expressed in muscle, brain stem, and spinal cord but not in the cerebral cortex where it is highly inducible after strokelike injury (12). Hsp27 is constitutive in the myocardium and is associated with contractile elements, and after heat shock treatment, its abundance is minimally changed (40). However, engineered overexpression of Hsp27 in cardiomyocytes provides protection against simulated ischemia (48, 68). Similarly, overexpression of Hsp27 in transgenic mice (19) protects the myocardium by reducing infarct size after ischemic injury. Hsp27 has more than one function. Specifically, Hsp27 acts as a molecular chaperone (17), inhibits actin and intermediate filament polymerization (72), reduces oxidative stress related to tumor necrosis factor-
-mediated cell death (49, 50), and suppresses signaling events leading to apoptosis (2, 6, 10, 11, 57). In mitochondrial-mediated apoptosis, Hsp27 blocks cytochrome c release and activation of caspase-3 and -9 (5, 7, 57) by sequestering cytochrome c and procaspase-3 (11, 57, 58). The function of Hsp27 appears to be regulated by its phosphorylation state and oligomeric size up to 800 kDa (62, 70). MAPKAPK-2 phosphorylates human Hsp27 on Ser15, Ser78, and Ser82. Interestingly, ischemia-reperfusion injury in rabbit heart causes two phosphorylation patterns of Hsp27: specifically, Ser15, Ser78, and Ser82; or Ser15 and Ser82 (71). In rat, Hsp27 is phosphorylatable at Ser13, Ser15, Ser27, and Ser86 (27), where rat Hsp27 Ser15 and Ser86 appear to correspond to human Hsp27 Ser15 and Ser82. For human Hsp27, Ser82 appears to be the major site of phosphorylation (37).
Inducible cytoprotection is regulated by various kinases in several cell signaling pathways. These survival kinases, and particularly PI3K-Akt and the MEK1/2-ERK1/2 pathways, are activated in ischemic preconditioning and ischemic postconditioning (26), and they regulate phosphorylation of Hsp27 (18). Interestingly, insulin activates MAPK (41, 54, 55), and MAPK activation leads to Hsp27 phosphorylation (1, 63). However, it is unclear whether insulin induces phosphorylation of Hsp27 in rat heart.
In our previous study at 6 h after insulin treatment, hearts had increased expression of Hsp70 and were resistant to ischemia-reperfusion injury (42). We hypothesized that at 1 h after insulin treatment, insufficient time would have limited any increase in the abundance of newly synthesized Hsps, and if myocardial protection where evident, other mechanisms such as MAPK activation and Hsp27 phosphorylation would be involved.
In this study we show that when hearts are isolated 1 h after a single physiological dose of insulin, they have significantly increased levels of phosphorylated Hsp27 during ischemia-reperfusion. Coincidental with the phosphorylation of Hsp27 is a significant improvement in functional myocardial recovery from ischemic injury. Inhibition of p38 MAPK blocked Hsp27 phosphorylation and insulin-induced myocardial protection.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Male Sprague-Dawley rats (250–300 g, Charles River, St. Constant, QC, Canada) were used in these experiments. All animal care, handling, and experimental procedures on animals were in accordance with the Guide to Care and Use of Experimental Animalsof the Canadian Council on Animal Care and were approved by the Dalhousie University Committee on Laboratory Animals.
Experimental Protocols and Groups
Two sets of experiments were done. In the initial experiments, to determine the effects of a single injection of insulin, rats were randomized into five groups: 1) naive control (Con), 2) sham control (Sham), 3) insulin treated (INS), 4) heat shock treated (HS), and 5) heat shock and insulin treated (HSINS). To identify the effects of Hsp27 phosphorylation in myocardial protection induced by insulin, rats were randomized into five groups: 1) Con, 2) Sham, 3) INS, 4) SB-203580 and insulin treated (SBINS), and 5) PD-098059 and insulin treated (PDINS). For all groups, rats were anesthetized with pentobarbital sodium (50 mg/kg ip). Naive control rats received an injection of pentobarbital sodium but were not heated or injected with insulin. In addition, sham-operated control rats received an injection of normal saline intramuscularly, similar to the insulin treatment group. Insulin-treated rats were injected intramuscularly in the thigh with 200 µU/g body wt. Heat shock-treated rats were placed on a temperature-controlled heating pad (48–50°C) and monitored with a rectal thermometer until core body temperature reached 42°C. Core body temperature was maintained between 42°C and 42.5°C for 15 min. Heat shock and insulin-treated rats were subjected to the heat shock treatment first and, at 10 min after the heat shock treatment, were injected with insulin (200 µU/g body wt) intramuscularly in the thigh. SB-203580 and insulin-treated rats were injected with the selective p38 MAPK blocker SB-203580 (1.0 mg/kg ip) and, 30 min later, were injected with insulin (200 µU/g body wt) intramuscularly in the thigh. PD-098059 and insulin-treated rats were injected with the selective ERK1/2 blocker PD-098059 (0.3 mg/kg ip) (22) and, 30 min later, were injected with insulin (200 µU/g body wt) intramuscularly in the thigh. Both blockers were purchased from LC Laboratories (Cat. No. S-3400 and P-4313, Woburn, MA). After treatment, all rats were returned to their cages.
At 1 h after the various treatments, rats were injected with pentobarbital sodium (50 mg/kg) and decapitated. For Langendorff perfusion, isolated hearts were immediately immersed in ice-chilled normal saline and quickly mounted on a cannula for perfusion. For Western blot analysis, hearts were briefly perfused and washed with normal saline at 4°C to remove blood and were then immediately freeze clamped. The cardiac tissue was also collected from Langendorff-perfused hearts after ischemia-reperfusion. For comparison, hearts were also collected from rats at 3 h after the various treatments. The frozen tissues were stored at –70°C before protein concentration determination (45) and Western blot analysis. For confocal microscopy, fresh heart samples were immersed in 2% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS, pH 7.4, 4°C) overnight.
Langendorff Perfusion and Cardiac Function
Isolated hearts were perfused in the nonejecting-modified Langendorff mode as previously described in detail (42). Modified Krebs-Henseleit bicarbonate buffer gassed with 95% O2-5% CO2 at 37°C was perfused into hearts at a constant rate of 10 ml/min with a static roller pump. Hearts were equilibrated to the perfusion system for 30 min. Heart rate and perfusion pressure were continuously recorded with a Grass model 7 polygraph (Grass Instruments, Quincy, MA). A balloon connected to a pressure transducer (P-23D, Gould) was inserted into the left ventricle through the mitral valve to measure left ventricular end-systolic and end-diastolic pressures as well as the maximal first derivative of LV pressure increase and decrease (±dP/dt). The maximal left ventricular developed pressure (LVDP) was calculated. After 30 min of equilibration, all hearts except for those in the Sham group were subjected to 30 min no-flow global ischemia. Sham hearts were continually perfused at 10 ml/min for 180 min. Following ischemia, hearts were reperfused at 10 ml/min for 120 min. Physiological measurements were taken at 10 min before the initiation of ischemia (time point = –40 min in Figs. 1
–9) during preischemic perfusion and at 5, 30, 60, 90, and 120 min of reperfusion. An analysis of left ventricular function and a calculation of the pressure-volume relationship were done according to Burkhoff and Sagawa (8) and as recently described by Li et al. (42).
|
|
|
|
|
|
|
|
|
Collected cardiac tissues were homogenized in 1 ml of homogenization buffer with a sonicator, as previously described (42, 43). The homogenized samples were immediately frozen in liquid nitrogen and then kept at –70°C until analyzed.
Western Blot Analysis
Heart tissue samples containing 20 µg of protein were solubilized in sodium dodecyl sulfate (SDS) sample buffer (36), heated at 95°C for 10 min, and loaded onto a mini-SDS-polyacrylamide gel (2.5% stacking gel, 7.5–12% running gel), according to previously described methods (42, 43). For dephosphorylation of proteins, samples were incubated with 10 units of alkaline phosphatase (Cat. No. M0290S, New England Biolabs, Ipswich, MA) for 15 min at room temperature before adding sample buffer. Proteins were separated by electrophoresis at 75 V for 20 min and 125 V for 
60 min and then electrotransferred onto a PVDF membrane. Membranes were incubated in 5% skim milk in 1x 0.1% Tris-buffered saline with Tween 20 (TBST, 0.1%, pH 7.6) as previously described (42, 43) for 1 h to block nonspecific binding of primary antibody. Membranes were incubated at 4°C overnight with the following primary antibodies: rabbit polyclonal anti-Hsp25 antibody (1:5,000, Cat. No. SPA-801, StressGen, Victoria, Canada), rabbit polyclonal anti-phospho (p)-Hsp27 (Ser15) (pHsp27Ser15; 1:1,000, Cat. No. SPA-525, StressGen, Ann Arbor, MI), rabbit polyclonal anti-pHsp27 (Ser82) (pHsp27Ser82; 1:1,000, Cat. No. SPA-524PU, StressGen, Ann Arbor, MI), mouse monoclonal anti-Hsp70 (1:2,000, Cat. No. SPA-810, StressGen, Ann Arbor, MI), or rabbit polyclonal anti-actin antibody (1:2,500, Cat. No. A-2066, Sigma, St. Louis, MO) in 5% milk in 1x TBST. On the next day, membranes were incubated in secondary horse anti-mouse horseradish peroxidase-conjugated antibody (1:10,000, Cat. No. PI-2000, Vector, Burlingame, CA) or goat anti-rabbit horseradish peroxidase-conjugated antibody (1:5,000, Cat. No. SAB-300, StressGen, Ann Arbor, MI). After being washed, the membranes were incubated in ECL Plus solution for horseradish peroxidase-labeled antibody (Cat. No. RPN2132, Amersham Biosciences) for 5 min and then washed in distilled water. Chemiluminescence was directly detected on a STORM 840 scanner with a fluorescence setting at excitation of 430 nm and emission of 503 nm [Photo Manager Two, 600 V; and Pixel size, 100 mm (STORM 840, Molecular Dynamics)]. Densitometric analyses for one-dimensional gels were done with imaging software (ImageQuant TL V.2003, Amersham Biosciences).
Confocal Immunofluorescence Microscopy
Tissue preparation. Following an overnight fixation in 2% paraformaldehyde in 0.1 M PBS (pH 7.4, 4°C), tissues were cryoprotected in 30% sucrose in 0.1 M PBS (pH 7.4). Tissue sections were then cut at 20 µm with a freezing, slicing microtome, and sections were kept in Millonig's solution at 4°C.
Immunofluorescence.
To characterize the localization of Hsp70, Hsp27, and pHsp27Ser82 in cardiac tissue and their relationship with components of the cell, sections were double labeled for either dystrophin, a cell membrane protein, or -tubulin, a cytoskeletal protein. Sections were incubated with primary rabbit polyclonal antibody against Hsp70 (1:1,000, Cat. No. SPA-812, StressGen, Victoria, Canada), primary rabbit polyclonal antibody against Hsp27 (1:1,000, Cat. No. SP-801, StressGen, Victoria, Canada), or primary rabbit polyclonal antibody against pHsp27Ser82 (1:1,000, Cat. No. SPA-524PU, StressGen, Ann Arbor, MI), and either mouse monoclonal anti-dystrophin [1:100 in 0.01 M PBS with 0.2% Triton X-100, pH 7.4 (PBST) and 1% BSA, Cat. No. D 8043, Sigma] (42) or mouse monoclonal anti--tubulin (1:1,000 in PBST and 1% BSA, Cat. No. T 5168; Sigma) (42). For immunostaining of Hsp27, pHsp27Ser82, Hsp70, tissue sections were processed as previously described (42). In brief, tissue sections were blocked with 10% normal goat serum in 1x PBST for 1 h. Tissue sections were incubated in primary antibodies at the concentration mentioned above at 4°C overnight. On the next day, the tissue sections were incubated in secondary antibodies (1:400 in PBS and 1% BSA, Alexafluor 546-conjugated goat anti-rabbit IgG and Alexafluor 488-conjugated goat anti-mouse IgG, Molecular Probe) at room temperature for 2 h. Finally, tissue sections were mounted on gelatinized slides and dried in the dark overnight at room temperature. Sections were coverslipped with ProLong Gold mounting media (Molecular Probe) and sealed. In each batch of sections stained for confocal fluorescence microscopy, some sections were incubated without primary antibody or secondary antibody to serve as controls for nonspecific staining. A Carl Zeiss Axiovert 200 laser-scanning microscope was used to Z section for confocal imaging at 1 µm. Images were captured with a CCD camera and LSM 510 META software (Version 3.2). Captured images were edited, adjusting only for brightness and contrast, and composed with Photoshop (version 7.0, Adobe Systems).
Statistical Analysis
For cardiac function data, multivariate ANOVA with repeated measures and Bonferroni multicomparison analysis were used to determine differences due to the effects of treatment, i.e., Sham, Con, INS, HS, HSINS, SBINS, and PDINS. Relative densitometric values for Hsp70, Hsp27, pHsp27, and actin were obtained from Western blot analysis. The densitometric values of each treatment group were standardized with actin and normalized to the Sham group. Statistical significance for the semiquantitative and normalized densitometric data was determined using ANOVA with or without covariant and Bonferroni multicomparison analysis. All statistics were done using SPSS, v13.0.1 (SPSS, Chicago, IL). The results are expressed as means ± SE. Significance was set at P 
0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Preischemic values for cardiac function are presented in Table 1. No significant differences (P > 0.05) were detected in the preischemic function between the experimental groups for heart rate, LVDP, left ventricular work (LVW), and ±dP/dt (Table 1). Sham hearts were not subjected to ischemia, and their function did not change significantly during 3 h of perfusion (Fig. 1). During the 30 min of no-flow global ischemia, contractility decreased quickly to zero for the Con, INS, HS, and HSINS groups. During reperfusion, there were no statistical differences (P > 0.05) between the groups or with the preischemic values for heart rate (data not shown).
|
|
Hearts from each treatment group were examined at 1 and 3 h after treatment for relative levels of Hsp70 and Hsp27 (Fig. 2). Little or no Hsp70 immunoreactivity was detected in Sham, Con, or INS hearts. Hsp70 immunoreactivity was detected at significantly elevated levels in HS (P 0.006, n = 3) and HSINS (P 0.003, n = 3) hearts compared with Sham, Con, or INS hearts at 3 h after treatment. Immunoreactivity for Hsp27, a constitutively expressed protein, was detected in all hearts, and its level was unchanged between 1 and 3 h after treatment or by the treatments.
Hsp70 and Dystrophin Immunofluorescence Microscopy
At 1 h after treatment, hearts were prepared for histology. Confocal micrographs of heart sections, double labeled with primary anti-Hsp70 and secondary antibody (AlexaFluor 546-conjugated anti-rabbit IgG, red) and primary anti-dystrophin and secondary antibody (AlexaFluor 488-conjugated anti-mouse IgG, green) are presented in Fig. 3. Dystrophin immunoreactivity was localized to plasma membranes in all experimental groups (Fig. 3, A, D, G, J, and M) and appeared to be unchanged by experimental treatments. Little or no Hsp70 immunoreactivity was detected in Sham or Con hearts (Fig. 3, B and E). After INS, Hsp70 immunoreactivity was occasionally detected and localized mainly to plasma membranes (Fig. 3, H and I). Following HS, Hsp70 immunoreactivity was detectable in cardiac tissue and appeared to be localized mostly between the cardiomyocytes and associated with microvessels (Fig. 3, K and L). After HSINS, Hsp70 immunoreactivity was mostly localized between the cardiomyocytes and associated with microvessels (Fig. 3, N and O).
Hsp27 and -Tubulin Immunofluorescence Microscopy
At 1 h after treatment, hearts were prepared for histology. Confocal micrographs of heart sections, double labeled with primary anti-Hsp27 and secondary antibody (AlexaFluor 546-conjugated anti-rabbit IgG, red), and primary anti--tubulin and secondary antibody (AlexaFluor 488-conjugated anti-mouse, green) are presented in Fig. 4. The -tubulin immunoreactivity was localized throughout the cytoplasm with some concentration along cell membranes, and microtubular networks were apparent in cardiomyocytes in all experimental groups (Fig. 4, A, D, G, J, and M). Hsp27 immunoreactivity was detectable in Sham (Fig. 4, B and C), Con (Fig. 4, E and F), INS (Fig. 4, H and I), HS (Fig. 4, K and L), and HSINS (Fig. 4, N and O) hearts and was localized throughout the cytoplasm with limited concentration along cell membranes. In INS hearts, some Hsp27 immunoreactivity appears to be concentrated on 3–5-µm-diameter structures. In HS hearts, some Hsp27 immunoreactivity appears to be concentrated with -tubulin on the cytoskeleton. No other obvious change in Hsp27 distribution was noted.
Western Blot Analysis of Phosphorylated Hsp27
Hearts from each treatment group were examined at 1 and 3 h after the treatments for phosphorylated Hsp27 (Fig. 5). The antibodies specific for pHsp27Ser15 (Fig. 5A) and pHsp27Ser82 (Fig. 5B) gave immunoreactive product at 1 and 3 h after treatment in various groups. Little or no pHsp27 immunoreactivity was detected in Sham or Con hearts. Although pHsp27Ser15 immunoreactivity appeared to be about five times elevated at 1 h in the HS and HSINS groups (Fig. 5C) compared with the Sham group, these normalized data were not significantly different (ANOVA, P > 0.05, n = 3). Similarly, pHsp27Ser82 immunoreactivity appeared to be about five times elevated at 1 h in the INS, HS and HSINS groups (Fig. 5C) compared with the Sham group; however, these normalized data were not significantly different (ANOVA, P = 0.073, n = 6).
Dephosphorylation of Hsp27
Hearts from each treatment group were examined at 3 h after treatment for pHsp27 and for abolition of pHsp27 immunoreactivity after treatment of samples with alkaline phosphatase (Fig. 6). The antibodies specific for pHsp27Ser15 (Fig. 6A) and pHsp27Ser82 (Fig. 6B) gave immunoreactive product in various groups. After incubation of samples with alkaline phosphatase, no pHsp27 immunoreactivity was detected.
Phosphorylated Hsp27 and Dystrophin Immunofluorescence Microscopy
At 1 h after the various treatments, hearts were prepared for histology. Confocal micrographs of heart sections, double labeled with primary anti-pHsp27Ser82 and secondary antibody (AlexaFluor 546-conjugated anti-rabbit IgG, red), and primary anti-dystrophin and secondary antibody (AlexaFluor 488-conjugated anti-mouse, green) are presented in Fig. 7. Dystrophin immunoreactivity was localized along cell membranes in all experimental groups (Fig. 7, A, D, G, J, and M). Little or no pHsp27Ser82 immunoreactivity was detectable in Sham (Fig. 7, B and C) or Con (Fig. 7, E and F) hearts. Diffuse pHsp27Ser82 immunoreactivity was detected in the cytoplasm and associated with cell membranes in the INS (Fig. 7, H and I), HS (Fig. 7, K and L), and HSINS (Fig. 7, N and O) hearts. pHsp27Ser82 immunoreactivity appeared to be less abundant and more aggregated in HS hearts compared with INS hearts.
pHsp27 Levels After 3 h of Perfusion
At 1 h after treatment, hearts were isolated and perfused as in Fig. 1. All hearts were subjected to 30 min of ischemia, except the Sham hearts that were continually perfused for 3 h. Hearts from each treatment group were examined for pHsp27, total Hsp27, and Hsp70 (Fig. 8). The antibodies specific for pHsp27Ser15 (Fig. 8A) and pHsp27Ser82 (Fig. 8B) gave immunoreactive product in various groups. Hsp27 immunoreactivity was unchanged, whereas the HS and HSINS groups had elevated Hsp70 immunoreactivity compared with the Con, Sham, and INS hearts. pHsp27Ser82 immunoreactivity in the INS group was significantly different (P = 0.046, n = 3) from that of the Sham group of hearts (Fig. 8C).
Inhibition of p38 MAPK With SB-203580 Suppresses Insulin-Induced Myocardial Protection
To investigate whether insulin increased phosphorylation of Hsp27 through activation of MAPK pathways, rats were treated with either SB-203580 or PD-098059, inhibitors of p38 MAPK and ERK1/2 pathways, respectively. Five groups of hearts, Sham, Con, INS, SBINS, and PDINS, were isolated and perfused as described for hearts in Fig. 1. Sham hearts were not subjected to ischemia, and their function did not change significantly during 3 h of perfusion (Fig. 9). During the 30 min of no-flow global ischemia, contractility decreased quickly to zero for the Con, INS, SBINS, and PDINS hearts. During reperfusion, there were no statistical differences (P > 0.05) between the groups or with the preischemic values for heart rate (data not shown).
Following ischemia, during reperfusion, recovery of LVDP (Fig. 9A), LVW (Fig. 9B), +dP/dt (Fig. 9C), and –dP/dt (Fig. 9D) was blocked for SBINS hearts compared with the INS hearts. PDINS hearts had a recovery of function that approached that of the INS and Sham hearts. The average percent (±SE) recovery of heart function during 2 h of reperfusion and significant differences among the treatment groups for treatment effects are summarized in Table 3.
|
At the end of the 3-h perfusion period (Fig. 9), hearts were examined for total and pHsp27 (Fig. 10). The antibodies specific for pHsp27Ser15 (Fig. 10A) and pHsp27Ser82 (Fig. 10B) gave immunoreactive product in various groups. SBINS and PDINS hearts appeared to have diminished pHsp27Ser15 immunoreactivity (Fig. 10A) compared with INS hearts. SBINS hearts appeared to have diminished pHsp27Ser82 immunoreactivity (Fig. 10B) compared with INS and PDINS hearts. Hsp27 immunoreactivity (Fig. 10, A and B) appeared to be unchanged among the hearts. Although densitometric analysis suggested that SBINS hearts had diminished pHsp27Ser82 immunoreactivity compared with INS hearts, significant differences (P = 0.004 to P = 0.019, n = 3 for each group) were evident only between PDINS hearts and Sham, Con, and SBINS hearts (Fig. 10C).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Insulin in a cocktail of glucose-insulin-potassium (GIK) has beneficial effects through changes in energy metabolism (56). During ischemia, the components of GIK favor a shift from the anaerobic free fatty acid metabolism with generation of toxic fatty acyl-CoA to the less toxic anaerobic glucose-dependent metabolism and facilitates glycolysis to produce more energy. Most clinical studies suggest GIK reduces both morbidity and mortality (67, 38). However, some studies show that GIK is of minimal benefit (9, 51). In animal studies, either GIK or insulin infusion protects cardiac contractile function when given before and immediately following ischemia. The benefits of insulin are thought to be through insulin intracellular mechanisms or signaling pathways (26).
Previously we have shown that at 6 h after a single physiological dose of insulin, hearts have elevated expression of Hsp70 and improved functional recovery against ischemia-reperfusion injury (42). This dose of insulin had no effect on blood glucose levels (43). Although evidence suggests that transgenic overexpression of Hsp70 is directly involved in protecting the heart from ischemia-reperfusion injury (47, 60), at 6 h after insulin injection, it remains an open question as to whether Hsp70 is contributing to the functional recovery of the hearts after ischemic injury. In the present experiments, little or no Hsp70 was detected in the hearts at 1 or 3 h after insulin injection, suggesting that Hsp70 is not involved in the improved functional recovery (Fig. 1) at 1 h after insulin treatment. Although Hsp70 was detected in hearts at 3 h after heat shock (HS and HSINS groups, Fig. 2) and in hearts after 1 h after heat shock followed by 3 h of isolated perfusion (HS and HSINS groups, Fig. 8), recovery of function was not evident in these groups (Fig. 1 and Table 2).
At 6 h after insulin injection, Hsp70 is localized mainly along plasma membranes of cardiomyocytes (42). In the present study, at 1 h after insulin injection, little or no Hsp70 is detected in the hearts (Fig. 3). In contrast, at 1 h after heat shock, Hsp70 was detectable in cardiac tissue and appeared to be localized mostly in small capillaries and in perivascular cells between the cardiomyocytes (Fig. 3), and this localization is not different at 6 or 24 h after heat shock (40, 42, 43). At 1 h after induction, it is unlikely that sufficient mature Hsp70 has accumulated in the cells to provide protection from ischemia-reperfusion injury. The difference in detectable Hsp70 at 1 h (Fig. 3) and 3 h (Fig. 2) after INS or HS treatment may be related to the activation of signaling pathways. Heat denatures protein that recruits Hsc70/Hsp70 into freeing the HSF-1. Free HSF-1 is phosphorylated, trimerized, and translocated to the nucleus (52, 65) where it binds to a regulatory upstream promoter heat shock element (HSE) on heat shock genes (28, 59). Insulin may be signaling through GSK3 to release its inhibition of the heat shock transcription factor. The overexpression of GSK3β resulted in significant reduction in heat-induced HSF-1 activities. The activation of HSF-1-HSE binding under non-heat shock conditions was also observed after inhibiting GSK3β with LiCl (LiCl mimics insulin inhibition of GSK3β) (73). When comparing our contractile data for HS and HSINS hearts, we think that the insulin is rescuing the contractile function even after heat shock. We are convinced that any recovery of contractile function at 1 h after insulin is not due to Hsp70 but might be due to phosphorylation of Hsp27.
Hsp27 is abundant in the heart and is associated with contractile elements. At 24 h after heat shock treatment, the Hsp27 distribution and abundance are minimally changed (40). In the current study, the abundance of Hsp27 is not significantly changed in the heart after insulin, heat shock, or heat shock and insulin treatments (Fig. 2). At 1 h after insulin injection, there appears to be some localization of Hsp27 between cardiomyocytes and possibly around blood vessels that is not obvious in the other treatment groups (Fig. 4). As with Hsp70, an increase in protective proteins in blood vessels could be the first line of defense in protecting the heart from reactive oxygen species during reperfusion injury (40). After heat shock, Hsp27 appears to be localized with -tubulin (Fig. 4) and may be acting as a chaperone to restore -tubulin and microtubule function (44).
As mentioned above, at 1 h after insulin injection, as expected there was no obvious increase in the abundance of either Hsp70 or Hsp27. However, survival kinase pathways such as PI3K-Akt, MEK1/2-ERK1/2, and the p38 MAPK pathways are activated by ischemic preconditioning (26) and are also regulating heat shock proteins and phosphorylation of Hsp27 (18). The MAPK pathway is activated by many stimulatory agents and phosphorylates downstream signaling molecules (16). Insulin also activates MAPK (24, 32, 49), and MAPK activation leads to Hsp27 phosphorylation (1, 25, 35, 39, 63). Thus, in the present study, we used phosphospecific antibodies to detect phosphorylated isoforms of Hsp27. pHsp27Ser15 was detected mostly in HS and HSINS hearts, and pHsp27Ser82 was detected repeatedly in INS, HS and HSINS hearts (Figs. 5 and 6). The variation was high, and significant differences were not achieved. However, we are confident that we detected phosphorylated isoforms of Hsp27 because the immunoreactivity was abolished by dephosphorylation with alkaline phosphatase (Fig. 6). The difference in distribution of pHsp27 between INS and HS hearts may be related to damage to cellular proteins caused by high temperature and pHsp27 stabilizing the cytoskeleton (Fig. 7).
During 3 h of isolated perfusion, hearts treated with insulin had the strongest recovery of function after ischemic injury (Fig. 1 and Table 2; and Fig. 9 and Table 3). Most interestingly, at the end of the perfusion period, the INS hearts had significantly elevated levels of pHsp27Ser82 compared with the Sham hearts (Fig. 8). Once insulin activates the insulin receptor, the PI3K/Akt and MAPK pathways are activated (55). Recent evidence indicates that insulin activates ERK1/2 and p38 MAPK (3, 25, 35, 41). p38 MAPK, in turn, activates MAPKAPK-2 that directly regulates the phosphorylation of Hsp27 (1, 39, 69). Ser15 and Ser85 phosphorylation of Hsp27 by MAPKAPK-2 is the key mechanism in reduction of apoptosis and facilitation of F-actin remodeling that protects cells from doxorubicin toxicity (69). Further evidence for phosphorylation of Hsp27 having a role in myocardial protection is provided by studies on activation of the p44/42 MAPK (ERK1/2) and p38 MAPK pathways by atorvastatin (20) and on activation of the adenosine A1 receptor and p38 MAPK (14).
In the present experiments, the elevated level of Hsp27 phosphorylation is associated with improved cardiac function that was diminished by SB-203580. The elevated level of pHsp27 appears to be contributing to the improved cardiac function, and the p38 MAPK pathway is responsible for the phosphorylation of Hsp27 induced by insulin. Phosphorylation of Hsp27 is thought to be important for its protective functions. Phosphorylation of Hsp27 alters its quaternary structure and chaperone function. Hsp27 exists in cells in either monomer and dimer forms or as multimer forms up to 600 to 800 kDa. Phosphorylation facilitates the movement of Hsp27 between the various pools (21). Larger multimers are thought to have a chaperone function and to regulate antioxidative activity (62). Phosphorylation of Hsp27 downregulates multimer size and the chaperone function in favor of dimer and monomer Hsp27 regulating actin cytoskeleton stabilization. In our experiments it may be that insulin treatment mobilized some pHsp27 from unphosphorylated Hsp27 pools. The pHsp27 may be regulating actin dynamics and stabilizing the cytoskeleton during the ischemia-reperfusion injury. Multimer forms of Hsp27 are likely functioning as chaperones and also reducing reactive oxygen species damage (61). In addition to insulin stimulating an increase in the amount of Hsp27 phosphorylation, it is also likely that the ischemia-reperfusion injury stimulated an increase in the amount of pHsp27 (71). Moreover, Hsp27 phosphorylation may be the key mechanism in insulin-induced myocardial protection. Insulin protects the cardiomyocyte by not only regulating glucose and free fatty acid metabolism but also regulating the phosphorylation state and, hence, the function of Hsp27 through p38 MAPK pathway.
Inducible myocardial protection is clearly complex and unlikely to be due to one mechanism. Rapidly acquired protection appears to be related to the activation of survival kinases (26), including phosphorylation of Hsp27. Delayed protection is dependent on protein synthesis and accumulation of various protective proteins, such as Hsp70. Insulin stimulation appears to regulate rapidly acquired protection as shown in this study and with a longer-term protection seen 6 h after insulin treatment that may involve an expression of Hsp70 (42).
In conclusion, insulin induces an apparent rapid phosphorylation of Hsp27 through the p38 MAPK pathway that is associated with improved functional recovery of cardiac contractile function after ischemia-reperfusion injury. After ischemia-reperfusion injury, only the insulin-treated hearts had significantly elevated levels of pHsp27. Insulin stimulation appears to regulate the phosphorylation of Hsp27 that may be providing rapid myocardial protection.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
induces changes in the phosphorylation, cellular localization, and oligomerization of human hsp27, a stress protein that confers cellular resistance to this cytokine. J Cell Biochem 58: 248–259, 1995.[CrossRef][Web of Science][Medline]- and oxidative stress-induced cytotoxicity in stably transfected murine L929 fibroblasts. J Immunol 154: 363–374, 1995.[Abstract]B-crystallin inhibit actin polymerization. Eur J Biochem 268: 2083–2090, 2001.[Web of Science][Medline]This article has been cited by other articles:
|
|
 |
|
  B. Laczy, B. G. Hill, K. Wang, A. J. Paterson, C. R. White, D. Xing, Y.-F. Chen, V. Darley-Usmar, S. Oparil, and J. C. Chatham Protein O-GlcNAcylation: a new signaling paradigm for the cardiovascular system Am J Physiol Heart Circ Physiol, January 1, 2009; 296(1): H13 - H28. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
