OUP user menu

Lipopolysaccharide induces oxidative cardiac mitochondrial damage and biogenesis

Hagir B. Suliman, Karen E. Welty-Wolf, MarthaSue Carraway, Lynn Tatro, Claude A. Piantadosi
DOI: http://dx.doi.org/10.1016/j.cardiores.2004.07.005 279-288 First published online: 1 November 2004


Objective: The responses to bacterial lipopolysaccharide (LPS) damage mitochondria by generating oxidative stress within the organelles. We postulated that LPS damages heart mitochondrial DNA and protein by oxidation, and that this is recovered by oxidative mechanisms of mitochondrial biogenesis.

Methods and results: Systemic crude E. coli LPS administration decreased mtDNA copy number and mtDNA gene transcription in rat heart caused by oxidant deletion of mtDNA. The fall in copy number was reflected in proteomic expression of several mitochondria-encoded subunits of Complexes I, IV, and V. Recovery of mtDNA copy number involved biogenesis as indicated by mitochondrial transcription factor A (Tfam) and DNA polymerase-γ expression. The transcriptional response also included nuclear accumulation of peroxisome proliferator-activated receptor-γ co-activator 1 (PGC-1) and mRNA expression for redox-regulated nuclear respiratory factors (NRF-1 and -2).

Conclusions: These novel findings disclose a duality of reactive oxygen species (ROS) effect in the heart’s response to LPS in which oxidative mitochondrial damage is opposed by oxidant stimulation of biogenesis.

  • Infection/inflammation
  • Sepsis
  • Mitochondria
  • Oxygen radicals
  • Redox signaling

1 Introduction

Current evidence indicates that host mediators, such as tumor necrosis factor-α (TNF-α), stimulate the reactive oxygen (ROS) and reactive nitrogen (RNS) species production involved in bacterial infection [33,43] and linked to experimental heart failure [12,22]. Indeed, TNF-induced ROS production in cardiac and other cells occurs mainly in mitochondria [9,16,23], and because lipopolysaccharide (LPS) presentation to immune cells stimulates TNF-α production and drives mitochondrial ROS generation, oxidants must play into inflammatory mitochondrial damage. However, little is known about cardiac mitochondrial oxidative stress induced by major bacterial products.

Mitochondrial constituents are susceptible to oxidation including the circular mitochondrial genome (mtDNA), which lacks protective histones and is located near electron transport complexes that produce ROS [8,19]. Unopposed mtDNA oxidation interferes with mitochondrial transcription and OXPHOS protein synthesis, and not only impairs respiratory capacity but may also exacerbate oxidant leakage [46]. Several groups have shown that endogenous ROS damage mtDNA, leading to altered mitochondrial gene expression and function [8,40,41,44,46].

Mitochondrial density in the cell is regulated by biogenesis, a response that depends on physiological and pathogenic factors which enable crosstalk between nuclear and mitochondrial genomes. Nuclear co-activators such as PGC-1 interact with nuclear respiratory factors (NRF-1 and NRF-2), which transactivate genes for OXPHOS, protein importation, and heme biosynthesis [6]. These factors also mediate mtDNA transcription and replication through two nuclear genes: mitochondrial transcription factor A (Tfam) and its co-factor, mitochondrial transcription factor B (mtTFB) [27,34]. In addition, PGC-1α repression by mutation causes loss of heart mitochondria and mitochondrial enzyme down-regulation [10].

We postulated that LPS induces oxidative damage that stimulates mitochondrial biogenesis by oxidant-driven mechanisms in the heart. We evaluated changes in rat heart mtDNA content, mRNA, and protein expression after systemic LPS administration to test the idea that LPS up-regulates nuclear–mitochondrial communication through oxidant-based signals originating in mitochondria. We found that LPS-induced oxidative stress damages cardiac mtDNA, decreases mtDNA copy number, and impairs mitochondrial gene transcription and protein expression, but this damage stimulates the transcription factor and nuclear gene expression required to activate biogenesis

2 Materials and methods

2.1 Animal protocols

The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Male rats (S.D., 300–400 g; Charles River) were injected with one dose of crude LPS (Escherichia coli, 055 B5; i.p. Difco) prepared in 1 ml of 0.9% NaCl. Controls received 1 ml of 0.9% NaCl. Animals were killed after 6, 24, or 48 h and the ventricles excised into cold isolation buffer. Fresh hearts were used to isolate mitochondria and snap-frozen hearts (−80 °C) for genomic studies. For physiological studies, rats were anesthetized with urethane and a Millar catheter was placed into the left ventricle (LV) via a carotid artery. Continuous LV systolic and diastolic pressure, heart rate, and positive and negative dP/dt recordings were made in control and LPS-treated rats after intravenous fluid (3–6 ml) or phenylephrine infusion to raise cardiac afterload by systemic vasoconstriction [24]. For electron microscopy, hearts of three control and three rats 24 h after LPS were flushed with PBS, perfusion-fixed with 2% paraformaldehyde+0.2% glutaraldehyde in PBS, removed and processed for EM [40].

2.2 Mitochondrial isolation and DNA extraction

Heart mitochondria (intrafibrillar and subsarcolemmal) were purified by gradient centrifugation [40]. Respiratory rates were measured with and without ADP using calibrated Clark electrodes (Diamond General, Ann Arbor, MI) [43]. mtDNA was isolated using NaI kits (Wako, Tokyo, Japan). Genomic DNA was extracted with a kit (DNA Isolator PS Kit; Wako), and DNA concentration was measured optically.

2.3 Oxidative stress

Fresh mitochondria were sonicated in 5% metaphosphoric acid, centrifuged at 1000×g for 10 min at 4 °C, supernatants frozen and stored at −80 °C. Mitochondrial oxidative stress was monitored using reduced and oxidized glutathione (GSH and GSSG) measured by enzymatic recycling assay with derivatization [40], malondialdehyde assay (Calbiochem, La Jolla, CA), and carbonyl formation by derivatization of oxidized mitochondrial protein with 2,4-dinitrophenylhydrazine (DNPH) and detection by rabbit anti-dinitrophenol (DNP) (1:250; Intergen, Purchase, NY) and HRP-conjugated anti-rabbit IgG (1:1000) visualized using the ECL system (Amersham, Piscataway, NJ). Band densities were quantified by densitometry (Bio-Rad).

2.4 Detection of mtDNA deletion

Deleted mtDNA was detected by PCR [41]. Nucleotide sequences at mtDNA breakpoints were characterized by sequencing the PCR fragments in both directions on an automated sequencer (373A Applied Biosystems) with the DYEnamic™ terminator Cycle Sequencing Kit (Amersham).

2.5 Southern blot analysis

mtDNA and nDNA blots were hybridized with mitochondrial and nuclear (18S rRNA) probes optimized for nDNA [40]. The radioactivities of mtDNA and nDNA bands were quantified by autoradiogram scanning (CS 710; Bio-Rad) and confirmed by direct quantification using a PhosphorImager (Bio-Rad) with Molecular Analyst Software.

2.6 Mitochondrial proteomics

Blue native polyacrylamide gel electrophoresis (BN-PAGE) was performed using the methods of Schagger et al. with minor modification [36–38]. Mitochondria (2 mg protein) were mixed in 6-aminocaproic acid (200 μl 0.75 M)–Bis-Tris (50 mM, pH 7.0) and 37.5 μl of DDM (10%), centrifuged (100,000×g), and 19 μl of 5% Serva Blue added to the supernatant before gradient electrophoresis (4–12%). Electron transport complexes were identified by molecular weight and known elution order [36–38]. Peptide subunits were resolved by 2-D Tricine-SDS-PAGE electrophoresis by excising BN-PAGE bands, incubating them in 1% SDS/1% mercaptoethanol for 2 h, and loading them onto a 16% separating gel. After electrophoresis, gels were stained with Coomassie brilliant blue followed by silver. Proteins were localized by apparent Mr and identified by MALDI-TOFMS. 2-D methods were worked out on pairs of samples from control rat hearts treated identically; three sets of gels with one control and one LPS heart were then carried simultaneously through each separation at each time point (6, 24, and 48 h).

2.7 Nuclear and mitochondrial mRNA expression

Total cytoplasmic RNA was extracted with TRIzol (Invitrogen), and 1 μg of sample was reverse-transcribed in reaction buffer of random hexamer primers, dNTPs, and RNasin (Promega). Gene transcripts were amplified in triplicate using gene-specific primers in Table 1 and Ref. [42]. 18S rRNA was used to control for efficiency of RNA extraction, reverse transcription, and amplification. Amplified mRNA was quantified by densitometry and normalized to 18S rRNA mRNA using image analysis software (Bio-Rad).

View this table:
Table 1

RT-PCR primer sequences

GeneSequence 5′–3′Size (bp)Annealing temperature (°C)

2.8 Northern blot analysis

RNA was extracted using Trizol and 10 μg RNA/sample was used in Northern blot analysis. Uncoupling protein (UCP-2 and UCP-3) cDNA probes were generated by RT-PCR (Table 1). Blots were hybridized to cDNA probes at 65 °C and washed at moderate stringency. RNA loading was standardized by stripping, hybridization to 18S rRNA probe, and visualization by phosphoimaging.

2.9 Western blot analysis

Proteins were separated by SDS-PAGE and prepared for immunoblot analysis [41]. Membranes were incubated with polyclonal rabbit anti-PGC-1 (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), monoclonal mouse anti-CaMKII, anti-CaMKIV (1:5000; BD Biosciences, San Diego, CA) or anti-tubulin (1:1000; Sigma). After five washes in TBST, membranes were incubated at 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG (Amersham). The membranes were developed by ECL and the protein was quantified on digitized images from the middle of the dynamic range relative to tubulin expression.

2.10 Sample size and statistical analysis

Samples from four to six animals at each time point were used for analysis of PCR, RT-PCR and Western blots. Grouped data were expressed as mean±S.D. Statistical analyses were performed by ANOVA followed by Tukey's post hoc comparison using computer software. A p<0.05 was considered significant. Regression analysis was performed using Statview (SAS, Version 5.0.1)

3 Results

3.1 Myocardial structure

The animal model was characterized by examining cardiac histology, LV function, and mitochondrial respiration. Light microscopy of ventricular sections after LPS showed no obvious disturbances in mitochondrial number or typical distribution of subsarcolemmal and interfibrillar populations, but by electron microscopy mitochondrial damage was apparent after 24 h (Fig. 1A). Both mitochondrial populations showed marked structural heterogeneity including patchy disruption of inner and outer membranes, variable swelling, and some distorted cristae and electron-lucent matrix. Some myocytes showed damaged or disrupted myofibrils while others had a greater mitochondrial density and smaller mitochondria compared to control myocytes.

Fig. 1

Myocardial structure and function after LPS. (A) Electron microscopy (EM) of myocardium showing LPS-induced changes in mitochondrial ultrastructure. Left panel: control heart shows two populations of myocyte mitochondria. Subsarcolemmal mitochondria are located beneath the plasma membrane (arrow). Interfibrillar mitochondria are situated among myofibrils (*). The mitochondria are smooth and electron-dense with distinct cristae and complete outer membranes. Right panel: 24 h after LPS, some mitochondria are lost and many are swollen with loss of cristae (*). Note the presence of budding (arrow). (B) Respiratory control ratios (RCR) of isolated cardiac mitochondria. RCR were similar in control and LPS-treated rats at 6, 24, and 48 h using NADH-linked substrates; n=6 samples/group, but State 3 respiration was reduced at 6 h after LPS administration (top panel). (C) Atrial natriuretic factor (ANF) mRNA expression in control and LPS-treated rat hearts relative to constitutive 18S rRNA (top). Mean±S.D. densitometry values are provided for n=6 samples/group (*p<0.05). (D) Left ventricular (LV) pressure 24 h after LPS administration. Peak LV pressure (top panel) and +dP/dt and −dP/dt as % control (bottom panel) are shown for 8 rats/group after fluid infusion and during PE infusion to increase the cardiac afterload (Mean±S.D.; *p<0.05).

3.2 Effect of LPS on respiration

Oxygen consumption of fresh mitochondria was measured under State 4 (resting) and State 3 conditions (malate+glutamate and ADP) at 6, 24 and 48 h after LPS. Respiratory control ratios (State 3/4) were 3 to 7, and after LPS were the same statistically as control. However, State 3 (peak) respiration was depressed at 6 h and remained below control at 24 and 48 h (Fig. 1B).

3.3 Left ventricular function

Based on the role of ANF in cardiac ion homeostasis and control of myocardial contraction, we tracked changes in myocardial function by monitoring ANF expression [18]. ANF was induced rapidly in response to LPS at 6 h, and remained elevated for 24 h (Fig. 1C). On this basis, LV function relative to controls was assessed 24 h after LPS administration (Fig. 1D). LV measurements in LPS-treated rats showed no significant increase in heart rate or LV end diastolic pressure, but greater variability in +dP/dt and decreases in −dP/dt consistent with diastolic dysfunction.

3.4 Mitochondrial oxidative stress

Depletion by LPS of the major mitochondrial antioxidant, GSH, occurred within 6 h even in well-coupled rat heart mitochondria (Fig. 2A). Mitochondrial lipid peroxidation product, malondialdehyde, began to accumulate by 6 h and increased fourfold 24 h after LPS (Fig. 2B) accompanied by greater protein carbonyl accumulation than controls (Fig. 2C).

Fig. 2

Cardiac oxidative stress after LPS. (A) GSH/GSSG ratio in cardiac mitochondria. Values are mean±S.D.; n=6; *p<0.05. (B) Malondialdehyde concentration in mitochondria. Values are mean±S.D.; n=6; *p<0.05. (C) Representative dot blot of mitochondrial proteins (20 μg) showing increased protein carbonyls. Post-LPS densitometry values were different from control at 6 and 24 h (n=6; mean±S.D.; *p<0.05).

3.5 mtDNA copy number and mtDNA damage

The mtDNA copy number by Southern blot analysis, expressed as the ratio of mtDNA to nuclear DNA (18S rRNA), decreased by 31% (p<0.05) 24 h after LPS compared with control rats, followed by recovery (Fig. 3A). The mtDNA damage was localized using primers encompassing two GC-rich direct repeats that undergo oxidative modification resulting in a bulky deletion detectable by PCR in the heart after LPS administration (Fig. 3B). The quantity of mtDNA deletion correlated negatively with mitochondrial GSH/GSSG (r2=0.89) and directly with malondialdehyde content (r2=0.91).

Fig. 3

mtDNA damage after LPS. (A) Southern blot analysis of cardiac mtDNA depletion after LPS. The autoradiography of the mtDNA fragment, top band, and bottom bands show signals from the nuclear DNA fragment for the 18S rRNA gene. The numbers below the bands show the mtDNA/nuclear 18S intensity ratio as a percentage. (B) Identification and verification of mtDNA deletion. mtDNA deletion demonstrated by PCR on GelStar-stained agarose gel. Numbers below the gel show % deletion as a percentage of wild-type mtDNA. The deletion was undetectable in control mitochondria but was readily detectable 24 h after LPS administration. Densitometry values represent measurements for 4 rats/group expressed as mean±S.D.; symbols indicate p<0.05 relative to control.

3.6 Mitochondrial proteomics

The mtDNA damage raised the issue of whether the LPS response affected only mitochondrial-encoded OXPHOS protein expression or also nuclear-encoded OXPHOS proteins. Fig. 4A shows representative first dimension BN gels at 6 and 24 h after LPS indicating substantial decreases in staining intensity of Complexes I, IV, and V. Complex III was variably affected and Complex II (encoded in the nucleus) was stable on most first dimension gels. The response of Complex IV was complex; mobility in the first dimension at 24 h was enhanced by a change in subunit composition that was demonstrated in the second dimension gel. Fig. 4B shows proteomic maps of a control heart and one 24 h after LPS administration. LPS administration transiently affected many proteins but nine mitochondrial peptides showed consistent and significant changes (Table 2). After LPS, two Complex I proteins (identified by MALDI-TOFMS and labeled a, b (Fig. 4B) decreased by more than 50%, one in Complex V decreased by 29% (c) while several Complex IV protein subunits decreased by 22–45% (d–i). The gels indicated that mitochondrial damage after LPS is complex, and affected the expression of both mitochondrial and nuclear encoded proteins. Some proteins affected by LPS were of non-mitochondrial origin, but these are not described here. By 48 h after LPS administration, however, the mitochondrial 2-D gels were indistinguishable from control heart gels (data not shown).

Fig. 4

Mitochondrial proteomics. (A) Non-denaturing Blue-native (BN) gel, first dimension. Note distinct decreases in Complex I, IV, and V 24 h after LPS. (B) Second dimension silver-stained PAGE of rat cardiac mitochondria. Numerals indicate the positions of the five respiratory complexes. Spots of sequenced mitochondrial proteins affected in LPS treated hearts are indicated by letters (a–i).

View this table:
Table 2

Decreases in selected cardiac mitochondrial proteins after LPS

Gel bandProtein ID (Swiss-Port)Complex of originPercent decrease (mean±S.D.)
aNADH dehydrogenase II61±4
bNADH ubiqinone oxidoreductaseI53±9
cATP synthase subunit 6V29±6
dCytochrome c oxidase subunit IIIV22±7
eCytochrome c oxidase subunit IIIIV26±4
fCytochrome c oxidase subunit IVIV36±7
gCytochrome c oxidase subunit Va or VbIV45±7
hCytochrome c oxidase subunit VIcIV32±4
iCytochrome c oxidase subunit VIIIV42±7
  • Nine mitochondrial proteins affected most consistently on analysis of 2D gels by densitometry 24 h after LPS administration (peak response).

3.7 Mitochondrial mRNA transcription

We measured the effect of decreased mtDNA copy number on mtRNA and selected mtRNA transcripts by semi-quantitative PCR. At 24 h after LPS, significant decreases were found in ATPase 6, COX I, and ND1 and ND2 transcripts (30–45%). In contrast, steady-state mRNA for the 18S rRNA nuclear transcript was stable after LPS administration (Fig. 5A).

Fig. 5

Steady-state DNA transcription in LPS treated rats. (A) Gel Star-stained 2% agarose gels demonstrating RT-PCR products of mtDNA transcription in rat hearts (left panel). Mitochondrial gene expression was measured by RT-PCR using gene-specific primers for rat mitochondrial ATPase 6, COX1, ND1, ND2. The nuclear mRNA for 18S rRNA was used to control for RNA integrity and efficiency of the RT-PCR. Mitochondrial gene expression after normalization to 18S rRNA levels is shown in right panel. 18S rRNA is constitutively expressed and used to control for the RNA integrity and the efficiency of the RT-PCR. (B) Expression of oxidant-sensitive nuclear respiratory factors (NRF-1 and -2) after LPS administration. Gel Star-stained 2% agarose gels demonstrate RT-PCR products from hearts of rat (left panel) and after normalization to 18S rRNA (right panel). (C) Genes regulating mtDNA replication. Nuclear expression of Tfam and pol-γ was measured using RT-PCR. Right panel shows gene expression after normalization to 18S rRNA. (D) UCP-2 and UCP-3 expression by Northern blot analysis in hearts of LPS-treated rats. 18S rRNA was used to control for the amount and quality of the RNA. Densitometry values are mean±S.D. (n=4; *p<0.05).

3.8 Mitochondrial DNA replication

We examined nuclear regulation of two oxidant-sensitive transcription factors necessary for mtDNA replication and mitochondrial biogenesis, nuclear respiratory factors-1 and -2 (NRF-1 and -2), and two nuclear genes regulated by these factors that encode proteins of mtDNA replication, DNA polymerase (pol-γ) replicates mtDNA and Tfam regulates mtDNA transcription and replication. NRF-1 and -2 transcripts increased significantly at 24 and 6 h, respectively (Fig. 5B), and pol-γ and Tfam transcripts increased significantly by 6 h and increased further at 24 h (Fig. 5C).

3.9 UCP expression

UCP proteins have variable capacities to dissipate the mitochondrial proton gradient. Northern blot analysis detected little UCP-2 mRNA in control rat hearts while LPS produced an appreciable increase in UCP-2 mRNA that peaked at 48 h (Fig. 5D). UCP-3 mRNA level doubled 6 h after LPS and then began to subside.

3.10 PGC-1 protein expression

PGC-1 is a nuclear co-activator regulated by energy homeostasis and glucose metabolism. Therefore, we assessed nuclear PGC-1 expression to explore the mechanism of differences in nuclear and mitochondrial transcription. PGC-1 protein was elevated 6 h after LPS (Fig. 6A) coordinately with expression of downstream genes, Tfam, UCP-2 and UCP-3. Tubulin expression did not change in either group.

Fig. 6

Mitochondrial biogenesis. Western blot analyses of cardiac biogenesis markers referenced to tubulin. (A) PGC-1, a nuclear marker of mitochondrial biogenesis. Control lanes (0 h) and after LPS treatment (6–48 h) are shown. Histogram shows densitometry for PGC-1/ tubulin expressed as mean±S.D. (n=4/ group; *p<0.05). (B) Changes in nuclear expression of CaMKIV in control rats (0 h) and after LPS. Histogram shows densitometry values for CaMKIV/tubulin (mean±S.D.; n=4/group; *p<0.05). (C) Changes in nuclear expression of CaMKII in controls and after LPS. Histogram shows densitometry values for CaMKII/ tubulin as mean±S.D. (n=4/group).

3.11 CaMK expression

Calcium/calmodulin-dependent protein kinases (CaMK) have been implicated as sentinel molecules in cardiac hypertrophy, and CaMK IV induces PGC-1 expression [47]. In LPS-injected rats appreciable nuclear CaMK IV accumulation was noted in the myocytes (Fig. 6B), however, CaMK II nuclear protein did not change after LPS (Fig. 6C)

4 Discussion

This study reports three novel aspects of heart mitochondrial injury during inflammation. First, a single LPS exposure associated with mitochondrial GSH depletion without necrosis oxidizes mtDNA within 6 to 24 h, and mtDNA copy number decreases via a 3.8-kb oxidant-driven deletion spanning coding regions for ND1, ND2, COX1, and COX 2 [41]. Second, the oxidative stress decreases not only specific mitochondria-encoded proteins of Complexes I and IV, but several nuclear-encoded proteins as well, and before substantial loss of in vitro coupling and peak respiratory capacity. Third, specific nuclear-encoded mitochondrial genes are activated by this stress, including genes of mitochondrial biogenesis and their key transcription factors and co-activators.

Normally, mitochondrial thiols, most notably GSH, prevent mtDNA oxidation but LPS administration decreases mitochondrial GSH in the heart. Concurrently, mtDNA copy number decreases within 6 h, and by 24 h, heterogeneous loss of OXPHOS proteins is apparent, implicating both mtDNA and mitochondrial protein oxidation in injury pathogenesis. The selective depletion of particular mitochondrial and nuclear-encoded Complex I and IV subunits also suggests that they are early targets of oxidative/nitrosative damage, which precedes the loss of functional compensation by the electron transport system. ROS are generated at several sites in mitochondria, notably Complex III during oxidation of NADH-linked substrates, and in some cases, Complex I inhibition may protect by limiting electron flow to Complex III [7]. It is therefore difficult to predict the effects of selective loss of Complex I proteins on mitochondrial ROS production. Loss of certain subunits may also generate additional ROS due to electron leakage [9,16], and we have not defined the role of RNS, which also affect Complex I and IV functions.

Cardiac mtDNA was more susceptible than nuclear DNA to damage after LPS, reported previously only with focal liver necrosis [40]. As nuclear DNA is more resistant to oxidation than mtDNA, DNA oxidation products accumulate more rapidly in mitochondrial than nuclear DNA [41]. That mtDNA is especially susceptible to lipophilic oxidants is due to a high mitochondrial lipid-to-DNA ratio and mtDNA's inner membrane attachments, while divergent ROS and antioxidant defense localization may evoke differential DNA and protein damage. Extensive and persistent mtDNA damage also suggests secondary ROS generation caused by defective OXPHOS.

Mechanistically, this study also showed that the most pronounced cardiac mitochondrial protein losses occurred from Complexes I and IV, but the data are insufficient to estiamte precisely how much of this effect was due to the loss of mtDNA coding for subunits of these complexes. Because some nuclear-encoded OXPHOS proteins were also diminished, we must consider susceptibility to ROS, RNS, or damage to mitochondrial protein importation as potential causes of protein loss superimposed on the effects of mtDNA damage.

LPS-mediated mitochondrial damage was clearly sufficient to stimulate corrective responses. Mitochondrial DNA transcription and replication require Tfam and DNA pol-γ importation and their transcripts were stimulated by LPS, but mitochondrial transcript levels (ND1 and ND2) had not recovered completely by 48 h when mtDNA copy number had returned to 82% of control. This apparent lag in transcriptional recovery is certainly predicted by the persistent 3.8-kb deletion, but that LPS-induced mtDNA damage would impair transcriptional efficiency more than replication also fits the notion that low Tfam levels stimulate mtDNA replication while higher levels stimulate transcription [15]. We must thus still quantitatively correlate the losses in mitochondrial DNA and RNA with changes in the corresponding OXPHOS proteins.

Recovery of energy capacity by biogenesis requires coordinated expression of genes by two independent genomes including transcription factor activation and binding to promoter consensus sequences in nuclear genes encoding for mitochondrial proteins [20]. Relevant transcription factors have been identified including NRF-1, NRF-2, PPARα and PPARγ, and Sp1 [20,35]. In response to biogenesis-inducing stimuli, transcription for these proteins changes before or coincident with increases in the expression of their target genes [17,49]. Cooperation among DNA-binding proteins and promoter diversity among nuclear genes for mitochondria uniquely regulate biogenesis [31], and evidence to date implicates PGC-1 as the most dominant of these regulatory proteins. LPS exposure concurrently elevates mRNA expression for PGC-1, NRF-1, and NRF-2, as well as Tfam and pol-γ, thereby creating a favorable environment for mitochondrial biogenesis. A computer-assisted transcription factor analysis (DNAsis, Hitachi Software) reveals that rodent PGC-1 promoters have multiple sites for redox-sensitive transcription factors, e.g., CREB and AP1 [1,4], as predicted from the PGC-1 human promoter sequence [11]. Thus, oxidant regulation of PGC-1 should exploit the simultaneous co-activation of NRF-1 and NRF-2, which are oxidant-sensitive transcription factors of nuclear genes for mitochondria including Tfam [35,42]. Tfam ensures coordinate induction of mtDNA transcription and replication, maintaining vital OXPHOS protein expression.

In various cells, Ca2+ influx activates transcription factor myocyte enhancer factor-2 (MEF2), which activates mitogen-activated protein kinase p38 [26]. This may also be mediated by CaMK because transfection of constitutively active CAMK IV strongly activates p38, and animals overexpressing active CaMK IV display enhanced PGC-1 activation [47]. Thus, CaMK-mediated phosphorylation regulates PGC-1 protein content [14] and nuclear translocation of CaMK IV likely contributes to the increase in PGC-1 protein [32]. Because PGC-1 is directly downstream of p38, it may be possible to establish a novel redox link from it to CaMK IV in the heart.

PGC-1 also co-activates MEF-2 [25,35], which activates slow twitch or oxidative fiber gene expression [48]. PGC-1 promotes GLUT-4 expression [32] and activates UCP-2 but not UCP-3 gene expression [28,29]. LPS-stimulated increases in PGC-1 expression correlated well with sustained UCP-2 and transient UCP-3 mRNA expression. Thus, we raise the intriguing thesis that PGC-1 influences UCP-2 and UCP-3 gene expression and oxidant production in the heart. Indeed, UCP-3 activation does depend on the peroxisome proliferator-activated receptor (PPARα, PPARγ) [39].

Early studies and sequence homology of UCP-2 and -3 with UCP-1 suggested they too stimulate mitochondrial proton conductance and uncouple electron transport from ATP synthesis [2]. UCP-2 now appears to be involved only in superoxide-induced uncoupling [13], and may serve an anti-oxidant function by decreasing mitochondrial ROS leakage [30]. UCP-2 regulates pancreatic insulin secretion by controlling ATP concentration [5], and its mRNA levels increase in skeletal muscle after LPS [3]. UCP-3 facilitates glucose transport and metabolism in cardiac and skeletal muscle cells [21], and mitochondria lacking UCP-3 overproduce ROS [45]. Thus, our findings may reflect anti-oxidant or signaling properties of these proteins in the heart.

In conclusion, LPS administration in vivo rapidly oxidizes mtDNA in the rat heart. Ensuing losses in mitochondrial RNA and protein support the idea that ROS generation in the defense against bacterial infection mediates cellular damage stemming from a decline in OXPHOS capacity. The stipulation is that this oxidative stress stimulates redox signaling of biogenesis, which drives mitochondrial recovery from the deleterious effects of LPS.


This study was supported by grants from NHLBI and the Veteran's Administration (CAP).


  • Time for primary review 16 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
View Abstract