| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
B in cardiac pathophysiology: differences between acute and chronic stimuli in vivo
Departments of 1Pharmacology and Cell Biophysics and 2Pathology, and 3Division of Cardiology, University of Cincinnati, Ohio; and 4Department of Medicine, Jefferson Medical College, Philadelphia, Pennsylvania
Submitted 24 February 2004 ; accepted in final form 27 January 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
B in cardiac physiology and pathophysiology has been difficult to delineate due to the inability to specifically block NF-B signaling in the heart. Cardiac-specific transgenic models have recently been developed that repress NF-B activation by preventing phosphorylation at specific serine residues of the inhibitory B (IB) protein isoform IB
. However, these models are unable to completely block NF-B because of a second signaling pathway that regulates NF-B function via Tyr42 phosphorylation of IB. We report the development of transgenic (3M) mouse lines that express the mutant IB(S32A,S36A,Y42F) in a cardiac-specific manner. NF-B activation in cardiomyopathic TNF-1.6 mice is completely blocked by the 3M transgene but only partially blocked (70–80%) by the previously described double-mutant 2M [IB(S32A,S36A)] transgene, which demonstrates the action of two proximal pathways for NF-B activation in TNF--induced cardiomyopathy. In contrast, after acute stimuli including administration of TNF- and ischemia-reperfusion (I/R), NF-B activation is blocked in both 2M and 3M transgenic mice. This result suggests that phosphorylation of the regulatory Ser32 and Ser36 predominantly mediates NF-B activation in these situations. We show that infarct size after I/R is reduced by 70% in 3M transgenic mice, which conclusively demonstrates that NF-B is involved in I/R injury. In summary, we have engineered novel transgenic mice that allow us to distinguish two major proximal pathways for NF-B activation. Our results demonstrate that the serine and tyrosine phosphorylation pathways are differentially activated during different pathophysiological processes (cardiomyopathy and I/R injury) and that NF-B contributes to infarct development after I/R.
nuclear factor-B; tumor necrosis factor-; signal transduction; ischemia-reperfusion
b (nf-b) was discovered in 1986 (57), considerable research has been directed toward its characterization. NF-B exists as a dimer of related proteins classified as Rel family members due to the presence of a Rel homology domain. The five mammalian family members include p50/p105 (NF-B1), p52/p100 (NF-B2), p65 (RelA), c-Rel, and RelB. NF-B dimers typically consist of a p50/p105 or p52/p100 subunit in complex with another member of the Rel protein family; the most common form is p65/p50 (2). Generally, NF-B dimers associate with an inhibitory B (IB) protein that acts to inhibit NF-B function. Although there are seven known members of the IB protein family, regulation of NF-B activity is most frequently mediated by the two most common isoforms, IB and IB
.
A diverse range of stimuli can induce either phosphorylation of two regulatory serines (Ser32 and Ser36) or phosphorylation of a single regulatory tyrosine (Tyr42) within the amino terminus of the IB protein. In addition to the regulatory phosphorylation sites (Ser32/Ser36 and Tyr42), IB also contains 11 additional serines and 7 carboxy-terminal tyrosines, several of which are known to be phosphorylated. Phosphorylation of IB at Ser32/Ser36 by the IB kinase- (IKK-) is the predominant mechanism by which NF-B becomes activated and is operative during ischemia (31, 47). Tumor necrosis factor- (TNF-) as well as other signaling pathways activate the IB kinase complex (IKK), which in turn phosphorylates Ser32/Ser36 of IB and leads to ubiquitination and immediate degradation of IB (55). It is less well known that Tyr42 of IB is subject to phosphorylation in response to several stimuli (nerve growth factor, interferon-
, and in macrophages, TNF-; Refs. 7, 25). Phosphorylation of IB at Tyr42 may involve the protein tyrosine kinases c-Src, p56 (Lck), and ZAP-70 and causes IB to dissociate from NF-B without immediate proteolysis (1, 11, 34, 35). Degradation or dissociation of NF-B from IB via either the serine or tyrosine phosphorylation pathway enhances nuclear localization of NF-B and is necessary for activation of NF-B-dependent genes. Although phosphorylation of other serine, threonine, and tyrosine moieties within IB is known to occur, both constitutively and inducibly, these sites are thought not to affect NF-B activation directly (reviewed in Refs. 2, 25).
More than 200 genes are known to be regulated by NF-B. Thus it is not surprising that this transcription factor affects a multitude of biological processes in response to diverse stimuli and pathophysiological states. Genes of cardiovascular relevance that are regulated at least in part by NF-B include those that function in nitric oxide (NO) production, prostaglandin biosynthesis, calcium handling, cardiomyocyte function, cell death and/or survival, stress responses, natriuretic factors, growth factors, extracellular matrix, remodeling and/or cell adhesion, and antioxidant proteins (2, 4, 22, 23). The effects of NF-B on cell death and survival are complex in that it can exert both anti- and proapoptotic effects; NF-B can activate very different sets of genes depending upon the specific kinetics of its activation (1). In the heart, NF-B has been implicated in ischemia-reperfusion (I/R) injury (8, 40, 45, 52), preconditioning (19, 41, 71), unstable angina pectoris (50, 65), myocarditis (53), congestive heart failure (17), hypertrophy of isolated cardiomyocytes (17, 20, 21, 48), and dilated cardiomyopathy (6, 28, 59). Transgenic mice with cardiac-specific expression of TNF- (TNF-1.6 mice) develop severe dilated cardiomyopathy with reduced contractile function and rapidly progress to heart failure (28). NF-B is activated (19, 29), and several NF-B-dependent gene products including matrix metalloproteinases and inducible NO synthase have been implicated in the pathophysiology exhibited by these mice (14, 29, 33). Clinical studies show that NF-B activation is increased in human heart failure and there is an association between reduced NF-B activity and beneficial reverse remodeling in patients with heart failure who have left ventricular (LV)-assist devices (15, 51, 68). The transcriptional inhibitors pentoxifylline and thalidomide have yielded promising results in small clinical trials for cardiomyopathy (38, 60), and NF-B inhibition is implicated in the mechanism of each of these drugs (37, 38, 67). Unfortunately, all of the drugs that block NF-B activity are relatively nonspecific. Although pharmacological evidence supports that NF-B favors cell death in I/R injury (24, 47, 52), NF-B suppresses apoptosis and favors cell survival in isolated rat ventricular myocytes and after permanent coronary occlusion in vivo (43, 46). Thus NF-B appears to mediate opposing effects, and the role of NF-B in specific cardiac pathophysiology remains unclear.
We have engineered transgenic 3M mice that express a dominant-negative IB with mutations of the amino-terminal serines and tyrosine that mediate NF-B activation [IB(S32A,S36A,Y42F)]. Like the previously described 2M [IB(S32A,S36A)] transgenic mice (10), these mice exhibit normal cardiac morphology, histopathology, and physiology. Previous work with similar constructs in cultured cells has validated that the mechanism of action of these dominant-negative, nonphosphorylatable, mutant IB proteins is competitive titration of the endogenous IB from NF-B. Furthermore, use of the IB(S32A,S36A) and IB(S32A,S36A,Y42F) mutants has validated that the two proximal pathways that mediate NF-B activation, by phosphorylation of Ser32 and Ser36 or Tyr42, are separable and have additive effects in human glioma cells that express IB(S32A,S36A,Y42F) (44).
Activation of NF-B in response to cytokines, I/R, and TNF-induced cardiomyopathy (i.e., TNF-1.6 mice) is completely absent in 3M transgenic mice. Interestingly, the 2M transgene blocks only 70–80% of NF-B activation in TNF-1.6 cardiomyopathic mice despite that the 2M line expresses relatively more of the transgenic dominant-negative IB protein (10). This suggests that the Tyr42 of the 2M IB protein is phosphorylated in TNF-1.6/2M double-transgenic mice and thereby implicates the two distinct pathways for NF-B activation (e.g., Ser32/Ser36 and Tyr42 phosphorylation) in this cardiomyopathic model. Both 2M and 3M transgenes completely block NF-B activation after acute TNF- administration and after I/R injury, which suggests that phosphorylation of IB on Ser32/Ser36 is the predominant pathway for NF-B activation after these acute stimuli. Finally, we demonstrate that NF-B activation in the heart contributes to cell death and myocardial infarction in response to I/R injury in vivo.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
B-Y42F, a kind gift of D. W. Ballard (Vanderbilt University School of Medicine; Nashville, TN). The pCMV4-FIB-Y42F plasmid contains the cDNA for human IB with a Tyr42Phe mutation. This cDNA has been shown to block Tyr42 phosphorylation-dependent activation of NF-B (5). The cDNA was used as a PCR target in generating the IB(S32A,S36A,Y42F) cDNA, containing the Ser32Ala, Ser36Ala, and Tyr42Phe mutations using the forward PCR primer 5'-CGATGAGTCGACAATGTTCCAGGCGGCC-3' and the mutagenic reverse primer 5'-CTGCGGCTCGAGGCGGATCTCCTGCAGCTC-CTTGACCATCTGTTCGAACTCCTCGTCTTTCATGGCGTCCA-GGCCGGCGTC-3'. The PCR product was subcloned into the pMyHC vector behind the -myosin heavy chain (-MyHC) promoter (62) after sequencing to check for appropriate location of Kozak sequences and splice donor/acceptor sites. The transgene was excised from the plasmid by NotI restriction digestion, purified, and microinjected into the pronuclei of fertilized mouse oocytes as previously described (10). The oocytes were derived from mice that were F1 hybrids between the C57BL/6 and 129Sv lines (Taconic). The founder transgenic mice were genotyped by PCR using the forward primer 5'-GAAGCCTAGCCCACACCAGAAATGACAGACAGATC-3' and the reverse primer 5'-CACAGCCAGCTCCCAGAAGTGCCTCAGCAATTTC-3' (94°C for 2 min, 35 cycles of 94°C for 30 s, 63°C for 30 s, 72°C for 1 min, and 72°C for 5 min); the 475-bp product indicated the presence of the transgene. Transgenic-positive founders were then tested by Southern blot analysis to verify the PCR result and determine the transgene copy number. Transgene-positive founders were backcrossed to C57BL/6 mice to derive lines of 3M mice.
Cardiomyopathic TNF-1.6 mice with cardiac-specific expression of TNF- have been described previously (28) and were crossed with 2M and 3M transgenic mice in these studies. F1 mice from crosses between the TNF-1.6 mice and either the 2M (mouse line 44, Ref. 10) or the 3M (mouse line 49) mice were obtained. The four expected genotypes were found in the expected proportions (i.e., 
25% each). Mice of each genotype were euthanized, and the hearts were used for preparation of nuclear extracts and EMSA analysis of NF-B activity as described (see Electrophoretic mobility shift assay). Because double-transgenic, single-transgenic, and nontransgenic (wild-type control) mice from each of these crosses were siblings and were analyzed in the F1 generation, all mice from a particular cross were of the same genetic background.
All mice were maintained in accordance with institutional guidelines and the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, revised 1985), and all procedures were approved by the University of Cincinnati Institutional Animal Care and Use Committee.
Southern blots.
Southern blots were performed as previously described (10, 26). Briefly, 30 µg of genomic DNA were digested with PstI; electrophoresed on a 0.8% agarose-Tris, acetic acid, and EDTA gel; and washed and transferred to a nitrocellulose filter. The membrane was then baked, blocked, and probed with a fragment that extends from base pairs 4267 to 4802 of the promoter sequence (GenBank no. 1621436) after it was labeled by random priming. After it was washed, the membrane was exposed to a PhosphorImager screen and analyzed using a Storm 840 image analyzer and ImageQuant software (Molecular Dynamics; Sunnyvale, CA). The transgene copy number was determined by densitometric analysis relative to the endogenous 3535-bp -MyHC fragment (one copy per haploid genome; Fig. 1).
|
B as described previously (10). For TNF-, the procedure was the same with modifications. Briefly, lysates were prepared and run on 13% polyacrylamide gels, and protein was transferred to 0.45-µm nitrocellulose filters (Amersham). Filters were blocked and incubated for 2 h with a primary polyclonal goat anti-TNF- antibody (Sigma; 1:500 dilution) in Tris-buffered saline [10 mM Tris·HCl (pH 7.2) and 0.15 M NaCl] that contained 5% nonfat dry milk (Carnation) and 0.1% Tween 20, washed, and then incubated for 1 h with a secondary antibody (donkey anti-goat antibody; 1:5,000 dilution; Santa Cruz Biotechnology). To determine TNF- specificity, duplicate blots were incubated with primary antibody, one in the presence of an excess of recombinant murine TNF- (5 mg/ml) to compete for the TNF--specific signal. Blots were washed, and signals were detected by enhanced chemiluminescence (ECL) using an ECL Kit (Amersham) or an NEN Renaissance Kit (NEN Life Sciences Products). The blots were exposed to film (Kodak BioMax ML), and signal intensities corresponding to bands were quantitated using National Institutes of Health (NIH) Image software. Controls were used to verify that signals were within the linear range of detection. The membranes were then stained with Ponceau S dye, overall protein loading was assessed with NIH Image software, and the chemiluminescent signals were normalized to account for protein loading and transfer variability. Experiments were conducted three times with samples from two individuals.
Cytokine stimulations of NF-B.
For stimulation analysis, mice were injected intraperitoneally with TNF- (0.10 µg/g), a cytomix (0.1 µg/g TNF-, 0.008 µg/g IL-1, and 0.2 µg/g IFN-), or a sterile water vehicle (all cytokines were from Peprotech). After 30 min, the mice were euthanized by CO2 inhalation. The hearts were removed, dissected into atrial and ventricular pieces, flash frozen in liquid N2, and stored at –80°C. Ventricular samples were processed and analyzed by EMSA.
Electrophoretic mobility shift assay. Nuclear extracts from ventricular tissue samples were prepared as described by Dignam et al. (12) with modifications. Briefly, ventricular tissue was pulverized at liquid N2 temperatures, homogenized at low speed in buffer A [10 mM HEPES, (pH 9), 1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT, 25 µg/ml leupeptin, 0.2 mM sodium orthovanadate, and 0.1% (vol/vol) Triton X], vortexed, and incubated on ice for 10 min. After centrifugation (5,000 g for 10 min), the pellet was suspended in solution C [20 mM HEPES, (pH 7.9), 25% (vol/vol) glycerol, 0.6 M KCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM PMSF, 0.5 mM DTT, 25 µg/ml leupeptin, and 0.2 mM sodium orthovanadate] and vortexed. This suspension was incubated on ice for 40 min with rigorous vortexing every 10 min. After centrifugation (10,000 g for 15 min), the supernatant was retained as a crude nuclear extract. Protein concentrations were determined using a Bio-Rad protein assay with bovine serum albumin as a standard.
A double-stranded, 22-bp oligodeoxynucleotide (5'-AGTTGAGGGGACTTTCCCAGGC-3'; Promega) that contained a consensus NF-B-binding site (underlined) was end labeled using [-32P]ATP and T4 polynucleotide kinase (Promega) and was purified using a G-50 Sephadex column (Pharmacia Biotech). The binding reactions were performed in a final volume of 10 µl that contained 10 µg (for TNF- and cytomix administration or TNF-1.6 mice) or 20 µg (for I/R) of nuclear protein, 10 mM Tris·HCl (pH 7.5), 50 mM NaCl, 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 4% glycerol (vol/vol), and 1 µg of poly(dI-dC). After a 10-min preincubation at room temperature, the labeled probe (1 x 105 cpm/reaction) was added to each reaction and the reactions were incubated for an additional 20 min at room temperature. The DNA-protein complexes were separated on 6% nondenaturing polyacrylamide gels in 1x Tris borate-EDTA buffer. Gels were vacuum-dried and exposed to X-ray film at –20°C using intensifying screens. Competition assays with 100-fold molar excess of unlabeled NF-B consensus oligodeoxynucleotide or control nonspecific oligodeoxynucleotide (Promega) were performed to ensure that the signal was specific for NF-B. Each reaction was repeated a minimum of three times.
Morphological and histological analyses. The 3M and nontransgenic (wild-type) mice at 15 wk of age were used for morphological and histological analyses by an experienced pathologist who was blinded with regard to the ages and genotypes of the animals. The 3M and nontransgenic (wild-type) mice at 15 wk of age were utilized for histological analysis. The mice were euthanized by CO2 inhalation, and the hearts were flushed, relaxed, and fixed under pressure as previously described (10). The hearts were postfixed overnight at 4°C, embedded in paraffin, and sectioned, stained, and examined by an experienced pathologist who was blinded to the ages and genotypes of the animals.
Echocardiography. Transgenic 3M and nontransgenic siblings (n = 5) at 30 wk of age were subjected to two-dimensional M-mode and pulsed-wave Doppler echocardiography while in the conscious state. In the M-mode echocardiography, measurements of end-diastolic and end-systolic dimensions and septal and LV wall thicknesses were made. Calculations of fractional shortening, mean velocity of circumferential fiber shortening, and ejection time were made from recorded data (Table 1). The individuals who performed the echocardiography and the resulting calculations were blinded with regard to the ages and genotypes of the mice.
|
B activation by EMSA.
For collection of LV tissue for EMSA analyses, reperfusion was for 15 min; for infarct-size determination, reperfusion was for 24 h. In the latter instance, the chest was closed in layers, and the mice were allowed to regain consciousness. The mice were euthanized after 24 h of reperfusion, and the hearts were removed after aortic cannulation with polyethylene-10 tubing. The hearts were perfused through the aortic root with a vital stain (1% triphenyltetrazolium chloride, pH 7.4, 3 ml over 3 min) at 60 mmHg. The suture was then tied at the site of previous occlusion, and the heart was perfused with a 5% solution of phthalo blue dye (Heucotech) to delineate the nonischemic region. The heart was frozen and razor cut into 5–7 transverse slices. The slices were weighed, fixed in 10% neutral-buffered formaldehyde, and photographed using a Nikon Coolpix 880 digital camera fitted with a UR-E2 macro lens. The images were downloaded to a PowerMac G4 computer, and computerized digital planimetry was performed using the latest version of NIH Image software. The areas of the region at risk for ischemia (red and white), the infarct region (white), and the nonrisk region (blue) were measured for each section; the slices were weighed, and the average measures of these regions were calculated according to the method of Fishbein et al. (13). The infarct size was calculated as a percentage of the region at risk for each heart. Power analysis indicated that n = 6 was adequate to determine whether the mean infarct size differed significantly between transgenic and wild-type (nontransgenic) groups (power = 0.95; = 0.05).
Statistical analysis.
Results are reported as means ± SE. Differences between groups and pair-wise multiple comparisons of intergroup data were analyzed by two-way ANOVA. Overall differences were further analyzed by post hoc contrasts with unpaired Student's t-tests using Bonferroni correction. Differences were considered significant at P 
0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Western blot analysis of IB and TNF- protein levels.
The transgenic IB(S32A,S36A,Y42F) protein can be distinguished from the endogenous IB protein by polyacrylamide gel electrophoresis (Fig. 2, A and B). After normalization for protein loading, the relative levels of IB(S32A,S36A,Y42F) protein were found to be 2.6-fold (mouse line 49) and 3.0-fold (mouse line 50) higher than endogenous IB levels. Levels of TNF- were not found to be significantly different in either 2M [IB(S32A,S36A)] or 3M [IB(S32A,S36A,Y42F)] transgenic mice relative to wild-type mice (Fig. 2C; 2M, 92 ± 0.053%; 3M, 100.2 ± 0.0067% vs. WT, 100.0 ± 0.067%; P > 0.05) at 40 wk of age (endogenous TNF- was not detectable before this age). Results from duplicate blots, one of which was incubated with excess recombinant murine TNF- (Fig. 2D), demonstrate that the 27-kDa intracellular form of TNF- was detected. The 18-kDa secreted form was presumably lost in the extracellular milieu.
|
B activation in response to exogenous cytokines.
Transgenic and wild-type mice (nontransgenic siblings) from mouse lines 49 and 50 were injected intraperitoneally with recombinant murine TNF- (0.1 µg/g), cytomix (0.1 µg/g TNF-, 0.008 µg/g IL-1, and 0.2 µg/g IFN-), or vehicle (sterile water). Results of EMSAs demonstrated high levels of NF-B DNA binding using nuclear extracts from wild-type mice injected with TNF- or cytomix, whereas only low levels of DNA binding were detectable in mice injected with vehicle (Fig. 3). EMSA of cardiac nuclear extracts from 3M mice showed no detectable NF-B activation in response to either TNF- or cytomix injection. Furthermore, in 3M mice, we were unable to detect even the low levels of NF-B activation observed in wild-type mice in response to vehicle injection. The results with TNF- are comparable to those previously published for 2M transgenic mice (10) and are shown in Fig. 3 (lanes 12–14) for comparison.
|
Echocardiographic analysis of cardiac function.
Transgenic 3M and nontransgenic mice at 30 wk of age were subjected to two-dimensional M-mode and pulsed-wave Doppler echocardiography while in the conscious state. There were no significant differences between measurements of LV end-diastolic and end-systolic dimensions, anterior and posterior wall thicknesses, and LV mass. Furthermore, calculations of the LV functional parameters of fractional shortening, velocity of circumferential fiber shortening, and ejection time were not significantly different between the two groups (see Table 1). Thus cardiac morphological and functional parameters are not altered by chronic repression of NF-B in cardiomyocytes (see Table 1 and online supplement Fig. 2).
NF-B activation in TNF-1.6 transgenic mice.
LV tissue samples from hearts of F1 mice derived from crosses between TNF-1.6 and either the 2M or the 3M mice were collected at 3–6 wk of age and used to prepare nuclear extracts. At this age, the hearts of TNF-1.6 mice are significantly hypertrophic (28), and there is robust activation of NF-B. Representative results of EMSAs of these samples are shown in Fig. 4. NF-B was robustly activated in TNF-1.6 mice, whereas only a low basal level of NF-B activation was observed in hearts of wild-type mice (Fig. 4, A and B). We noticed that the upper NF-B complex is actually a doublet that consists of two bands. Supershift experiments showed that both anti-p65 and anti-p50 antibodies shifted the upper band (Fig. 4C), which demonstrates that this complex contains both p65 and p50, and thus it is the p65/p50 heterodimer implicated as a potent transcriptional activator.
|
B oligodeoxynucleotide but is unaffected by nonspecific oligodeoxynucleotide, which suggests that it does contain Rel family members (Fig. 4A, lane 6 and Fig. 4B, lane 9). The lower band of the upper complex (small gray arrow) and the lower, constitutive complex (large gray arrow) are not blocked by either transgene.
NF-B activation after I/R.
To determine the effect of I/R (30 min of ischemia and 15 min of reperfusion) upon activation of NF-B, wild-type, and transgenic mice (2M mouse line 44 and 3M mouse line 49) were subjected to either I/R or sham surgery. Nuclear extracts were prepared from ischemic and control tissues and were subjected to analysis by EMSA; representative results are shown in Fig. 5. In wild-type mice, I/R activates NF-B relative to sham open-chest surgery. However, experiments using extracts from 2M and 3M mice exhibit no detectable NF-B activation after I/R. Competition assays (Fig. 5, lanes 11 and 12) demonstrate that the indicated complex binds specifically to the NF-B DNA-binding site. Previously published results (29, 45, 69) showed that this complex is the p65/p50 heterodimer. Thus, within the limits of detection of the assay, I/R-induced activation of NF-B is completely blocked by genetic blockade of NF-B in both 2M and 3M mice.
|
B blockade upon infarct size after I/R injury.
Of 16 mice subjected to I/R injury in this experiment, 14 survived and showed ECG evidence that confirmed 30-min ischemia and verified reperfusion. Expressed as a percentage of the region at risk, wild-type mice (n = 8) exhibited an infarct size of 20.3 ± 2.8%. The 3M mice (n = 6) exhibited a significantly smaller infarct size of 6.1 ± 1.3% (P 0.05; Fig. 6). There was no significant difference between the region at risk in the two groups (P > 0.05) and no correlation was found between normalized infarct size and the size of the risk region for either group (r2 = 0.39 and 0.22 for wild-type and 3M mice, respectively). Neither group exhibited any abnormalities of gross coronary architecture.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
B activity in cardiovascular disease. In any case, the existence of multiple NF-B-subunit proteins that may be functionally redundant would severely confound results with individual Rel family-knockout mice. Previously described cardiac-specific transgenic models have been shown to exhibit a specific blockade of NF-B activation in response to administration of TNF- and LPS (10, 19). However, none of these models is sufficient to delineate the role of NF-B activation via phosphorylation of Tyr42 of IB. Until now, there was no murine model that exhibited complete in vivo blockade of NF-B activation in cardiomyocytes.
We demonstrate that NF-B is completely inhibited in 3M mice subsequent to cytokine administration and in response to TNF--induced cardiomyopathy, and that there are no adverse effects of genetic blockade of baseline NF-B activity. These findings confirm two aspects of our previous results (10), namely, that 1) cardiac NF-B blockade does not lead to cardiac morphological or functional abnormalities, and 2) cardiac NF-B activation occurs predominantly in cardiomyocytes as evidenced by the fact that cardiomyocyte-specific, dominant-negative blockade prevents detectable activation of NF-B in myocardium. That NF-B activity is not required for maintenance of cardiac structure and/or function probably reflects the fact that basal NF-B activity is low and activation of NF-B in the heart occurs predominantly in response to stimuli such as cytokine exposure, mechanical stretch, I/R, hypoxia, and -adrenergic receptor stimulation (8, 40, 42, 52).
Activation of NF-B in TNF-1.6 transgenic mice.
Recent studies (17, 20, 21, 32, 48) have shown that NF-B plays an important role in the hypertrophy of isolated cardiomyocytes in response to various types of stimuli, including TNF-, myotrophin, and G protein-coupled agonists. Several investigators (18, 28, 29) have shown that transgenic mice expressing TNF- specifically in cardiac tissues exhibit TNF--induced cardiomyopathy that is associated with activation of NF-B. TNF-1.6 mice, which overexpress TNF- in a cardiac-specific manner, develop severe dilated cardiomyopathy with reduced contractile function, exhibit alterations in fetal and extracellular matrix gene expression, and progress rapidly to heart failure (6, 28, 59). Our results with TNF-1.6/2M and TNF-1.6/3M double-transgenic mice show that the 2M transgene only partially blocks activation of NF-B (70–80%; see Fig. 4A), whereas the 3M transgene completely blocks its activation (see Fig. 4B). In contrast, although the level of activation of NF-B is similar to that in TNF-1.6 mice (see Fig. 3 vs. Fig. 4), both the 2M and 3M transgenes completely block NF-B activation consequent to acute activation by exogenous TNF- administration (see Fig. 3). Moreover, the 2M line used (mouse line 44) expresses the transgenic IB(S32A,S36A) protein at levels that are 6.5-fold higher than endogenous IB (10), whereas the 3M line used (mouse line 49) expresses IB(S32A,S36A,Y42F) protein at levels only 2.6-fold higher than endogenous IB. Taken together, this evidence suggests that the complete and partial blockades of NF-B activation exhibited in hearts of TNF-1.6/3M and TNF-1.6/2M mice, respectively, are due to functional differences between the transgenic 2M and 3M IB proteins (i.e., the Y42F mutation) rather than quantitative differences in protein expression or levels of NF-B activation. Because the transgenic 2M and 3M proteins act by titrating endogenous IB from NF-B, we interpret these results to mean that a fraction of the transgenic 2M protein is tyrosine phosphorylated at amino acid position 42, and this mediates some amount of NF-B nuclear localization that results in NF-B activation in the range of 20–30% of that seen in TNF-1.6 mice. This Tyr42-mediated activation could be indirectly mediated by TNF--induced expression of other cytokines such as IFN- that elicit Tyr42 phosphorylation of IB (7, 25). Alternatively, it is possible that other signal transduction cascades activated during the development of cardiac hypertrophy or progression to cardiac failure mediate this tyrosine phosphorylation. Other factors that are known to induce Tyr42 phosphorylation of IB and activation of NF-B are nerve growth factor and the Src and Lck tyrosine kinases (35). Increased activation of Src is associated with pressure-overload and endothelin-1-induced cardiomyocyte hypertrophy (27, 63), whereas Lck is activated during development of chronic dilated cardiomyopathy due to Coxsackie virus infection (30). Furthermore, many of the "classical" hypertrophic signaling cascades (phosphatidylinositol 3-kinase/Akt, p38, MAPK, Ras, and PKC) can activate NF-B in a variety of cell types (25). Thus it seems likely that NF-B is activated in chronic cardiomyopathy through multiple parallel signal transduction pathways. NF-B thereby serves as a signaling integrator for a diverse set of signals, and it regulates a wide array of downstream genes (25). The effects of partial and complete NF-B blockade upon pathophysiology and remodeling in cardiomyopathic TNF-1.6 mice are complicated. The analysis of the TNF-1.6/2M mice has revealed that NF-B blockade in TNF--induced cardiomyopathy results in reduced hypertrophy and improved survival (unpublished observations). Analyses with 3M mice are ongoing and will be presented elsewhere in full.
Role of NF-B in I/R injury.
I/R results in a complex series of events that includes production of reactive oxygen species (ROS), NO and NO derivatives, calcium overload, release of cytokines and chemoattractant molecules, and inflammation/infiltration. I/R has been shown to induce activation of NF-B in heart (8, 40, 52). NF-B is known to regulate expression of antiapoptotic genes in cardiomyocytes after ischemic stress (39, 40, 52), and downregulation of these genes is implicated in inducing apoptosis associated with I/R injury in heart (39, 40). Furthermore, dominant-negative inhibition of NF-B in isolated cardiomyocytes results in increased levels of TNF--induced apoptosis (46). These studies support the conclusion that NF-B predominantly promotes cell survival after I/R. Conversely, NF-B is implicated in the regulation of multiple genes with pivotal roles in I/R injury (2, 14a, 66) such as intracellular adhesion and chemoattractant molecules, Fas ligand, and inflammatory cytokines (IL-1, IL-6, IL-8, and TNF-), which suggests that NF-B can also promote cell death after I/R. Although pharmacological studies using diethyldithiocarbamate (DDTC) to inhibit NF-B support this interpretation (7, 24, 52), this agent is relatively nonspecific; both DDTC and pyrrolidine dithiocarbamate also act as ROS scavengers and iron chelators (56), and DDTC also inhibits superoxide dismutase (3). Because it is known that ROS are involved in mediating I/R injury, this activity may mask the effect of NF-B inhibition, and the reduction of NF-B activity by these agents may indirectly result from reduced ROS levels. Thus these pharmacological studies do not directly query the role of NF-B in I/R. Two studies have shown more directly, using decoy oligodeoxynucleotide strategies, that NF-B blockade can protect the heart against ischemic injury produced during cardioplegic arrest (54) and after I/R in rats (45). However, it is difficult to obtain homogenous decoy-dependent blockade. Moreover, because other transcription factors potentially bind or overlap with NF-B DNA-binding elements, these results may not be highly specific. Thus the exact role played by NF-B in I/R injury in the heart in vivo remains controversial.
We show that both 2M and 3M mice completely lack measurable I/R-induced activation of NF-B in the heart (see Fig. 5). If phosphorylation of IB Tyr42 were significant after I/R, we would expect to partially activate NF-B much as occurs in TNF-1.6 transgenic mice. Thus although we cannot rule out a small, undetectable level of Tyr42-dependent activation of NF-B, Tyr42 phosphorylation does not detectably contribute to NF-B activation after I/R in our model. In contrast to the situation in cardiomyopathic myocardium (TNF-1.6), in vascular endothelial cells after oxidative stress (6), or in human glioma cells (44), Tyr42 phosphorylation of IB is not a significant factor in myocardial I/R-induced activation of NF-B in vivo. Although previous studies have shown that IB is tyrosine phosphorylated after preconditioning, IB has a total of eight tyrosines, several of which are known to be phosphorylated, and specific IB phospho-Tyr42 antibodies do not exist. A recent study by Zhang et al. (70) showed that late ischemic preconditioning is abrogated by tyrosine kinase inhibition in association with reduced generalized tyrosine phosphorylation of IB in rabbits. However, the experiments of Zhang et al. do not demonstrate a cause-and-effect relationship between tyrosine phosphorylation of IB and NF-B activation by late preconditioning. Furthermore, this study did not distinguish between phosphorylation of the regulatory Tyr42 and other tyrosines in the IB protein. Finally, interpretation of this study is complicated by the fact that tyrosine kinase inhibition will block multiple signaling cascades including MAPK pathways that are known to be involved in cardioprotection and to activate NF-B (25, 41). Although the other seven tyrosines in IB may indeed be phosphorylated in a regulated manner, there is no evidence that phosphorylation of these tyrosines modulates NF-B activation. To date, our studies address the effects of the 2M and 3M mutant IB proteins after I/R only. Whether tyrosine phosphorylation of IB plays a functional role in NF-B activation after ischemic preconditioning will be addressed elsewhere.
We found a highly significant 70% reduction of infarct size in 3M transgenic vs. wild-type mice (see Fig. 6) and previously have reported a 61% reduction with 2M transgenic mice (25). These results clearly demonstrate that the overall action of NF-B in the heart after I/R is injurious. Furthermore, we found that the reduction in infarct size was not different between 2M and 3M mice (P > 0.05). Thus although we are easily able to detect additive effects in this model (49), we found no additive effect upon infarct size of the 3M vs. the 2M transgene [IB(S32A,S36A) vs. IB(S32A,S36A,Y42F)]. We would expect such an additive effect if both the Ser32/Ser36 and Tyr42 pathways significantly contributed to I/R injury. In contrast, using similar constructs, Miyakoshi et al. (44) showed an additive effect that indicated there are separate effects of the Ser32/Ser36 and Tyr42 phosphorylation pathways upon NF-B activation in human glioma cells. Our infarct-size results support the EMSA results discussed herein and further suggest that Tyr42 phosphorylation of the 2M protein does not contribute measurably to NF-B activation and cell death after I/R. This conclusion is strengthened by a recent study (47) that showed that IMD-0354, a relatively specific inhibitor of IKK-, which is the specific Ser/Thr kinase that phosphorylates IB at Ser32/Ser36, prevents increased NF-B activation after myocardial I/R and reduces infarct size by 59% in rats. Our results are consistent with the hypothesis that acute stimuli including TNF- administration and I/R act predominantly via pathways that mediate Ser32/Ser36 phosphorylation of IB, likely IKK-dependent pathways. We cannot, of course, rule out the possibility that our findings reflect unknown or compensatory effects of chronic NF-B blockade. Ruling out such effects will require studies employing a method of cardiac-specific NF-B blockade that can be regulated. However, because TNF- is regulated by NF-B, a specific possibility was that chronic NF-B inhibition reduces endogenous levels of cardiac TNF-. Because TNF- is involved in I/R injury, reduced baseline levels of TNF- would affect I/R injury. Our result that NF-B blockade does not affect levels of endogenous TNF- (see Fig. 2) rules out the possibility that reduced basal levels of TNF- indirectly underlie the cardioprotective effects of NF-B genetic blockade. Whether reduced activation of TNF- and/or other genes post-I/R mediates the cardioprotective effect of NF-B blockade is the subject of ongoing research.
Of relevance to the role played by NF-B in ischemic injury is a study recently presented by Misra et al. (43), who demonstrated that mice expressing a cardiac-specific IB dominant-negative transgene [IB(
36)] that is functionally similar to the 2M mutant transgene (10, 19) produce larger infarcts after permanent coronary ligation (PL) relative to nontransgenic controls. This study showed an increase in incidence of apoptosis (14.5 vs. 2.3%, 6 h postligation) and a 50% increase in infarct size (34 vs. 17%) in transgenic vs. nontransgenic mice. Misra et al. (43) conclude that NF-B is antiapoptotic after PL, whereas our results show that NF-B promotes cell death after I/R. Although at first difficult to reconcile, similar disparate results were reported with mice bearing genetic ablation of TNF- or simultaneous ablations of both TNF- receptors (TNFR1/2). Kurrelmeyer et al. (30) showed that TNFR1/2-knockout mice have increased infarct size after PL (30). However, a more recent study (36) showed that TNF--knockout mice have smaller infarcts after I/R. We have independently confirmed that infarct size is reduced after I/R in both TNF--knockout (49) and TNFR1/2-knockout (W. K. Jones and X. Ren, unpublished observations) mice in our model. Taken together, these results suggest that TNF- and NF-B primarily promote cell death after I/R and yet promote cell survival after PL. The differences between the outcomes of I/R and PL experiments likely result from differences in the ischemic insults. After permanent occlusion of a coronary artery, a significant portion of the risk region will infarct and cannot be acutely rescued due to the permanent blockage of flow. Different levels of ROS, NO, and NO derivatives likely occur in the border and infarct zones in the two models and may differentially activate signaling pathways with differential effects upon NF-B-dependent gene-expression programs. NF-B is known to regulate both proapoptotic and antiapoptotic genes, and different NF-B-dependent gene-expression programs affect the decision between cell death and survival (25). We believe it is likely that NF-B exerts both beneficial (prosurvival) and injurious effects that may be differently balanced in distinct regions of the ischemic zone; after I/R, the summative effect is injurious, whereas after PL, the summative effect is protective. Additional work with these models will allow us to sort out the beneficial/prosurvival and the injurious/pro-cell death actions of NF-B. Nevertheless, it is clear that NF-B plays an important role in mediating cell death and survival decisions during the development of myocardial infarction.
Prospectus for IB(S32A,S36A,Y42F) transgenic mice.
The 3M mice offer the first specific means of completely inhibiting NF-B activation in cardiomyocytes in vivo. As such, they offer a number of advantages over pharmacological inhibitors and double-stranded DNA decoys as Jones and colleagues (10) have previously delineated. Here, we demonstrate that the 3M transgene completely blocks activation of NF-B in TNF-1.6 mice with cardiac-specific expression of TNF-. This provides the foundation for future investigations into the role played by NF-B in the development of cardiac hypertrophy and cardiomyopathy and will allow us to delineate the NF-B-dependent genes involved in these processes. In addition to studying cardiomyopathy, the 3M mice may be used to investigate NF-B action and signal transduction in various models of cytokine signaling, heart failure, septic shock, cardiac allograft rejection, viral myocarditis, toxic substance or drug exposure, and autoimmune disease. Furthermore, the 2M and 3M mice will be useful for pharmacological experiments or in experiments using transgenic mice or other approaches to block potential upstream activators. Such experiments will allow us to distinguish the signaling pathways that activate NF-B via phosphorylation of Ser32/Ser36 and/or Tyr42. Thus the relative contribution of various signaling cascades to activation of NF-B can be determined for specific stimuli. Delineation of NF-B signaling will improve our understanding of the multiple parallel signaling pathways that activate NF-B in cardiac pathophysiology and eventually provide rationale for developing novel therapeutic approaches to treat human heart disease.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
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.
* M. Brown and M. McGuinness contributed equally to this study. 
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
activation of nuclear transcription factor-B in marrow macrophages is mediated by c-Src tyrosine phosphorylation of IB. J Biol Chem 273: 29417–29423, 1998.
