Infarct-induced steroidogenic acute regulatory protein: a survival role in cardiac fibroblasts

Abstract

Steroidogenic acute regulatory protein (StAR) is indispensable for steroid hormone synthesis in the adrenal cortex and the gonadal tissues. This study reveals that StAR is also expressed at high levels in nonsteroidogenic cardiac fibroblasts confined to the left ventricle of mouse heart examined 3 days after permanent ligation of the left anterior descending coronary artery. Unlike StAR, CYP11A1 and 3β-hydroxysteroid dehydrogenase proteins were not observed in the postinfarction heart, suggesting an apparent lack of de novo cardiac steroidogenesis. Work with primary cultures of rat heart cells revealed that StAR is induced in fibroblasts responding to proapoptotic treatments with hydrogen peroxide or the kinase inhibitor staurosporine (STS). Such induction of StAR in culture was noted before spontaneous differentiation of the fibroblasts to myofibroblasts. STS induction of StAR in the cardiac fibroblasts conferred a marked resistance to apoptotic cell death. Consistent with that finding, down-regulation of StAR by RNA interference proportionally increased the number of STS-treated apoptotic cells. StAR down-regulation also resulted in a marked increase of BAX activation in the mitochondria, an event known to associate with the onset of apoptosis. Last, STS treatment of HeLa cells showed that apoptotic demise characterized by mitochondrial fission, cytochrome c release, and nuclear fragmentation is arrested in individual HeLa cells overexpressing StAR. Collectively, our in vivo and ex vivo evidence suggests that postinfarction expression of nonsteroidogenic StAR in cardiac fibroblasts has novel antiapoptotic activity, allowing myofibroblast precursor cells to survive the traumatized event, probably to differentiate and function in tissue repair at the infarction site.

Figure 1.

Morphological consequences of LAD ligation. Three days after sham operation (Sham, A–C) or LAD ligation (MI, D–F), paraffin-embedded heart sections were taken at the plane below the LAD ligature (illustration in D) and examined by immunohistochemistry staining with troponin I (A and D) and H&E (B and E) and by EM (C and F). Note a marked loss of troponin I (§) in the outer wall of the LV (D), also depicting the onset of the LV wall thinning (D) compared with that for the sham heart in A. E, Higher magnification H&E image shows MI-inflicted shrinkage of the cardiomyocytes (myo) and a marked increase in interstitial cell number compared with those for the sham heart in B. Striation of the normal myofibril sarcomeres noted in the color-enhanced inset in B is lost after MI (color enhanced inset in E), also infiltrated with numerous cells next to blood vessels (bv). RBC, red blood cells; §, monocyte/macrophage; n, cardiomyocyte nucleus; LA, left atrium; RA, right atrium. F, Electron micrograph depicting severe damage to the subcellular organization of cardiomyocytes in the MI-damaged myocard. C, Normal organization of the contractile elements in a longitudinal (inset C1, thin arrows in myofibril [myof]) and transverse orientation (inset C2). Also shown are large normal mitochondria (mit) with densely packed cristae (short arrows in mit, inset C1). Z, Z-disk; M, M-line. F, Micrograph after MI shows severe loss of sarcomeric organization (inset f2) and deformation of the mitochondria inner membrane cristae that reduce in number and appear swollen (arrows in inset f1). Asterisk denotes intramitochondrial swelling.

Figure 2.

MI induction of nonsteroidogenic StAR expression in the heart. A, 3 days after MI (or sham) operation, proteins were extracted from the outer walls of the LVs and RVs (pool of 3–4 hearts per experiment) and samples (80 μg/lane) were analyzed by Western blotting for the expression of CYP11A1, 3β-HSD (A), and StAR (B1). Ovaries from cycling mice (8 μg/lane) were used as a positive control. Shown are representative gels (n = 3) where the nonspecific (n.s.) immuno–cross-reactive protein bands served as a loading marker; this approach was taken in the face of an infarct-related massive loss of cytoplasmic proteins normally used for this purpose. B2, total RNA was extracted from the tissues described above and Star gene products were determined by semiquantitative radioactive RT-PCR and Western blot analysis. Shown are PCR products of StAR (300 ng of total RNA for RT, 75 ng of cDNA, and 26 PCR cycles) and ribosomal protein l19 transcript (21 PCR cycles) used for normalization, as well as StAR protein.

Figure 3.

In vivo expression of StAR-bearing cells in infarcted heart tissue. Ligation of LAD was performed, and 3 days later the mouse heart was processed for indirect immunohistochemical staining of paraffin sections at the plane illustrated in Figure 1 (D, inset). W, depicts confocal tiling images compiled at low magnification of a transverse heart section stained with antitroponin I (trop) (green) and DAPI (blue). Infarct areas (§) show darker troponin I-depleted myocard. Inset, a myocard zone of the inner LV wall, where the muscle fibrils are cross-sectioned and therefore seemingly show lighter staining of troponin I. A–D, high magnifications focused on the zones depicted in W. A, Longitudinally sectioned myocytes of the RV myocard (myo) showing a troponin I pattern with typical sarcomeric striation and no StAR labeling. BC, capillaries; BV, blood vessels; rbs, red blood cells; n, nucleus. B, C, and E, StAR-positive cells (red) noted in troponin I (green)–depleted zones of the infarct-damaged myocard. Note mitochondrial StAR staining (arrows in B1, C, and F) colocalized with cytochrome c (cyt) (B2). The morphology of the StAR-positive cells is either elongated (arrows in B, B1, and C), or roundish (asterisks in B, B1, and C). E and E1 (a differential interference contrast microscopy image), inset, depict a typical border zone between a severely damaged myocard depleted of troponin I (§) and scattered myocytes still showing a striated troponin I pattern. Note ample numbers of StAR-staining cells close to blood vessels and capillaries denoted by red blood cells (RBC) (arrows in E and E1). No such StAR staining could be observed in interstitial cells (*) next to blood vessels or capillaries (arrows) noted in the healthy myocard of the RV. F, Border-zone section double stained for StAR and α-SMA. Note remnant α-SMA in the myocytes, but a lack of this actin in putative proto-myofibroblasts with StAR-bearing mitochondria (arrows). D, control section stained identically to image C but excluding the StAR first antibody, resulting in no StAR signals.

Figure 4.

Induction of nonsteroidogenic StAR expression in neonatal and adult heart cultures. A, Neonate (P2) rat hearts were dissociated with fig proteases, and cells were cultured for 4 days as described in Materials and Methods. Forskolin (10 μM) was added for 6 hours and indirect immunofluorescent staining with antisera to troponin I (green) and StAR (red) was performed. Note sarcomeric striation of troponin I (*) in cardiomyocytes, whereas StAR protein is expressed in elongated mitochondria (arrows; A, inset) of nonmyocytes. n, nucleus. B, Neonate (P2) rat hearts were dissociated as in A and after 4 days in culture, RT-PCR, and Western blot analyses were performed time dependently after StAR induction with 10 μM forskolin (Fsk). Shown are representative analyses repeated 3 times with similar results. C, Forskolin was added for 6 hours to cells as above, in the absence or presence of 20 μM mCCCP. Note mCCCP-mediated ablation of StAR import marked by loss of the mitochondrial 30-kDa mature protein (mStAR) and appearance of the remnant 37-kDa preprotein form (pStAR). n.s., nonspecific protein. D, Adult heart fibroblast cultures were treated with forskolin for 24 hours, and protein extracts (20 μg/lane) were examined by Western blot analysis for expression of StAR, CY11A1, and 3β-HSD. Ovarian granulosa cell extract (5 μg/lane) served as a positive control for the gene products. N/T, nontransfected cells. E, Neonate rat heart cultures were treated with either 10 μg/mL forskolin, 10 μM H₂O₂, or 0.1 μM STS for stimulation of apoptosis (Materials and Methods). On the following day, StAR (30 kDa) and tubulin (loading marker) were examined by Western blot analysis (20 μg/lane). F, Adult rat heart cultures were prepared (Materials and Methods) and forskolin, H₂O₂, or STS treatments were conducted and analyzed as in E.

Figure 5.

Effect of STS on apoptosis and StAR expression in primary cultures of neonate and adult rat heart. A–D, Neonate (P2) rat hearts were dissociated, and the cells were put in culture as described in the legend to Figure 4. A, Untreated control cells were immunofluorescently stained with α-SMA (red), troponin I (Cy 5, blue), StAR (green), and DAPI (blue). Note 3 cell types: cardiomyocytes (1) with a typical striated sarcomeric pattern of troponin I and SMA (purple, result of red and blue merge); myofibroblasts (2) stained red with SMA cables (arrows); and fibroblasts without SMA (arrowheads denote nuclei of such 3 cells). B, On day 5 in culture, cells were treated for 3 hours with 0.1 μM STS in serum-free medium, washed free of the drug. and fixed 8 or 24 hours later (B1 and B2, respectively) to be stained with antisera as in A. B1, Note collapse of SMA cable structures in myofibroblasts (arrows in no. 2 cells). Other cells devoid of SMA content show expression of mitochondrial StAR (no. 3 cell). B2, The number of StAR-positive cells increased at 24 hours after STS treatment (no. 3 cells), whereas many of the shrinking apoptotic cells (no. 2 cells) were SMA-expressing (arrows). A SMA-positive survivor myofibroblast with rounded StAR-expressing mitochondria is circled. A representative cardiomyocyte (no. 1) shows purple staining (merged SMA/troponin I) with loss of sarcomeric pattern. C, Survival of apoptosis-resistant StAR-expressing cells in double-treated STS culture in which a second 3-hour exposure to STS was repeated 48 hours after the standard first treatment shown in B2. The image presented was retrieved 24 hours after the second STS treatment. D1, Spontaneous differentiation of neonate cardiac myofibroblasts maintained 13 days in culture. Note the typical markers of terminal differentiation in culture, including thick SMA stress cables crossing between well-spread cells via cytoplasmic bridges (D1, inset, arrows) and multiple focal adhesion sites containing vinculin (D1a and inset). Some cells contain low basal levels of StAR decorated mitochondria (*). No add., no treatment. D2, A 3-hour exposure to 0.1 μM STS documented 1 day later as done for B results in massive labeling of StAR-bearing mitochondria in most of the cells and no signs of apoptosis or any collapse of the SMA stress cables. E1, Adult rat heart was dissociated with collagenase, and the cell monolayer was stained on day 5 for StAR, SMA, and troponin I. Shown is a myocyte-free (troponin I–negative) fibroblast monolayer with SMA-positive (no. 1) and SMA-negative (no. 2) cells; arrows denote thick SMA stress fibers. E2, Treatment with STS (as in B) induced StAR expression in apoptosis-resistant, mostly undifferentiated cells (no. 3). A minor fraction of the apoptotic cells expressed StAR (no. 4). Also note the differentiated myofibroblasts with collapsed SMA cables (arrows).

Figure 6.

StAR-expressing heart fibroblasts are less sensitive to STS-induced apoptosis. Adult heart fibroblasts were grown on glass coverslips and treated with 0.1 μM STS as described in the legend to Figure 5E2. A, Individual cells stained for StAR (green) and DAPI (artificial magenta) were identified by a differential interference contrast microscopy image (not shown), and apoptosis (fragmented nucleus) was defined in cells containing mitochondria with or without StAR expression (Materials and Methods). Shown is a representative image depicting several types of cellular statuses: nonapoptotic StAR-negative cells (no. 1); nonapoptotic StAR-positive cells (no. 2); and apoptotic StAR-positive cells (no. 3). Arrows point to the apoptotic StAR-negative cells. B, Apoptosis is presented as a percentage of cells with a fragmented nucleus, with or without StAR-bearing mitochondria (StAR⁺ or StAR⁻, respectively). More than 500 cells were counted in each experiment. Bars represent means ± SD of 3 independent experiments. *, P < .05 compared with StAR⁻ apoptotic cells.

Figure 7.

StAR knockdown increases heart fibroblast susceptibility to apoptosis. Adult heart fibroblasts were transfected with either siStAR or siCtrl negative control (Materials and Methods). Cells without any transfection (N/T) served as an additional control. One day later, the cells were treated for 7 hours with STS (0.1 μM), followed by FACS analyses of cells stained with annexin V-FITC and PI. A, Representative Western analysis of StAR knockdown (average 55%–60% of normalized siCtrl) after STS treatment. B, FACS analysis. Annexin V-PI–stained apoptotic cells were counted after STS treatment (panels 3–4, 7–8, and 11–12) or none (No add., panels 1–2, 5–6, and 9–10). Average values of 3 independent experiments are presented in C (mean ± SD). *, P < .05 relative to the cognate siCtrl.

Figure 8.

Antiapoptotic activity of recombinant StAR in HeLa cells. StAR was transiently expressed in HeLa cells, and 48 hours after transfection the cells were treated for 6 hours with STS (0.5 μM) and double stained with antibodies to StAR (red) and cytochrome c (Cyt c, green); DAPI-stained DNA is blue or artificial white. A, Representative image showing a normal StAR-positive cell with an intact nucleus (top cell in A and B) and an apoptotic StAR-negative cell with a fragmented nucleus (bottom cell in A and B). Color-merged image A shows the elongated morphology of StAR- and cytochrome c–containing mitochondria (A1, inset, arrowheads), whereas in the apoptotic StAR-negative cell, cytochrome c leaked to the cytosol and the mitochondria underwent fission (A2, inset, arrowheads). C, Apoptosis was calculated as a percentage of cells with a fragmented nucleus. More than 100 cells were screened in each experiment. Bars represent means ± SD of 3 independent experiments. *, P < .05.

Figure 9.

Attenuated BAX activation by StAR in cardiac fibroblasts. Adult heart fibroblasts were transfected with either StAR RNAi (siStAR) or negative control RNAi (siCtrl) oligomeres (Materials and Methods). One day later, the cells were treated for 3 hours with STS (0.1 μM), washed, and incubated for an additional 18 hours in serum-free medium without STS. A, Representative Western analysis of StAR protein knockdown examined in serum-free medium with or without STS. Also shown is the lack of STS effect on the cellular levels of total BAX and the mitochondria content (heat shock protein 60 [HSP60]); tubulin served as the loading marker. B, Mitochondria were isolated, and purity was assessed by Western blot analysis depicting VDAC1 (porin) and cytochrome c (Cyt c) as mitochondrial (m) markers, and phosphoglycerate kinase (PGK1/2) as a cytosolic (c) protein marker. C. Mitochondria isolated from the cells treated as in A were mildly digested with trypsin as described in Materials and Methods and subjected to Western blot analysis of BAX cleavage. Shown is a representative blot; VDAC1 was used as a loading control. D, Data obtained from lanes 3 and 4 of C were plotted as a percentage of active (cleaved) BAX of total (mean ± SD) in 3 independent experiments. **, P < .01.