Tlr4 Deficiency Protects against Cardiac Pressure Overload Induced Hyperinflammation

Transverse aortic constriction provokes a pro-inflammatory reaction and results in cardiac hypertrophy. Endogenous ligands contribute to cardiac hypertrophy via toll-like receptor (TLR)-4 binding. A lack of TLR4 signaling diminishes hypertrophy and inflammation. Wild type mice undergoing aortic constriction respond to a lipopolysaccharide second-hit stimulus with hyperinflammation. The objective of this study was to assess whether other second-hit challenges utilizing TLR ligands provoke a comparable inflammatory reaction, and to find out whether this response is absent in TLR4 deficient mice. Assuming that cardiac stress alters the expression of pattern recognition receptors we analyzed the effects of transverse aortic constriction and second-hit virulence factor treatment on TLR expression, as well as cytokine regulation. Wild type and Tlr4 -/- mice were subjected to three days of TAC and subsequently confronted with gram-positive TLR2 ligand lipoteichoic acid (LTA, 15mg/g bodyweight) or synthetic CpG-oligodesoxynucleotide 1668 thioate (20 nmol/kg bodyweight, 30 min after D-galactosamin desensitization) signaling via TLR9. Hemodynamic measurements and organ preservation were performed 6 h after stimulation. Indeed, the study revealed a robust enhancement of LTA induced pattern recognition receptor and cytokine mRNA expression and a LTA-dependent reduction of hemodynamic pressure in TAC wild type mice. Second-Hit treatment with CpG-ODNs led to similar results. However, second-hit effects were abolished in Tlr4 -/- mice. In total, these data indicate for the first time that cardiac stress increases the inflammatory response towards both, gram-negative and gram-positive, TLR ligands as well as bacterial DNA. The decrease of the inflammatory response upon TLR2 and -9 ligand challenge in TAC Tlr4 -/- mice demonstrates that a lack of TLR4 signaling does not only prevent left ventricular hypertrophy but also protects the mice from a cardiac stress induced hyperinflammatory reaction.


Abstract
Transverse aortic constriction provokes a pro-inflammatory reaction and results in cardiac hypertrophy. Endogenous ligands contribute to cardiac hypertrophy via toll-like receptor (TLR)-4 binding. A lack of TLR4 signaling diminishes hypertrophy and inflammation. Wild type mice undergoing aortic constriction respond to a lipopolysaccharide second-hit stimulus with hyperinflammation. The objective of this study was to assess whether other second-hit challenges utilizing TLR ligands provoke a comparable inflammatory reaction, and to find out whether this response is absent in TLR4 deficient mice. Assuming that cardiac stress alters the expression of pattern recognition receptors we analyzed the effects of transverse aortic constriction and second-hit virulence factor treatment on TLR expression, as well as cytokine regulation. Wild type and Tlr4 -/mice were subjected to three days of TAC and subsequently confronted with gram-positive TLR2 ligand lipoteichoic acid (LTA, 15mg/g bodyweight) or synthetic CpG-oligodesoxynucleotide 1668 thioate (20 nmol/kg bodyweight, 30 min after D-galactosamin desensitization) signaling via TLR9. Hemodynamic measurements and organ preservation were performed 6 h after stimulation. Indeed, the study revealed a robust enhancement of LTA induced pattern recognition receptor and cytokine mRNA expression and a LTA-dependent reduction of hemodynamic pressure in TAC wild type mice. Second-Hit treatment with CpG-ODNs led to similar results. However, second-hit effects were abolished in Tlr4 -/mice. In total, these data indicate for the first time that cardiac stress increases the inflammatory response towards both, gram-negative and gram-positive, TLR ligands as well as bacterial DNA. The decrease of the inflammatory response upon TLR2 and -9 ligand challenge in TAC Tlr4 -/mice demonstrates that a lack of TLR4 signaling does not only prevent left ventricular hypertrophy but also protects the mice from a cardiac stress induced hyperinflammatory reaction.

Introduction
Severely ill patients demonstrate an increased post-traumatic susceptibility towards secondary bacterial infections [1,2]. Recovering homeostasis after a primary insult is the result of complex mechanisms involving activation and suppression of the immune system. Secondary insults can easily perturb this process. This may cause an inappropriate immune response and increase morbidity and mortality.
Furthermore, it has been demonstrated that a variety of endogenous ligands signals via TLRs. Those damage associated molecular patterns (DAMPs) such as extracellular matrix components, heat shock proteins, mitochondrial DNA are released upon tissue and cell injury, and activate the immune system via TLRs [6]. In mice, DAMP-induced injury, following cardiac pressure overload and myocardial infarction can be attenuated via inhibition of TLR4 signaling [7][8][9][10].
Myocardial injury modulates the cardiac immune system and enhances the subsequent effect of the bacterial virulence factor lipopolysaccharide (LPS) [11,12]. It has been shown that transverse aortic constriction (TAC) alone increases the expression of the TLR4 co-receptor CD14 [12]. However, it remains elusive whether PRRs other than CD14 are also upregulated after TAC. In a second-hit model of pressure overload-induced cardiac hypertrophy followed by endotoxin stimulation, the subsequent LPS challenge induced higher NFκB activation and cytokine expression in the TAC group compared to sham mice, and elevated CD14 expression even further. Polymicrobial sepsis confronts the immune system with a variety of exogenous TLR ligands. In a peritonitis model, the increase of TLR2, 4, and 9 as well as CD14 has been observed [13]. Therefore, we expect that TAC might also alter the response towards ligands signaling via TLR2 and -9. Tlr4 -/mice are protected from extensive cardiac hypertrophy. It remains unknown whether TAC influences their cardiac immune response towards TLR ligands such as lipoteichoic acid (LTA) and CpG-ODN 1668 thioate signaling via TLR2 and TLR9.
The purpose of this study was to (i) determine whether TAC changes the sensitivity towards the virulence factors LTA and CpG-ODN, (ii) detect alterations in the expression of pattern recognition receptors after pressure overload-induced hypertrophy and second-hit stimulation; (iii) investigate whether TLR4 deficient mice with reduced susceptibility towards cardiac stress respond with attenuated inflammation towards second-hit LTA or CpG-ODN stimulation.

Experimental animals
Experiments were performed on female mice at an age of about 10-12 weeks with an approximate weight of 18-22g. C57BL/6 mice were purchased from Charles River (Sulzfeld, Germany). Breeding pairs of Tlr4 -/mice on C57BL/6 genetic background were kindly provided by S. Akira [14]. All animals employed in the present study were housed in individually ventilated pathogen-free cages with free access to water and standard rodent chow. The animals were handled according to the principles of laboratory animal care (NIH publication No. 85-23, revised 1996). The animal procedures carried out in this study were in accordance with German legal guidelines and were specifically approved by the responsible local authority for animal care (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Recklinghausen, Germany, animal protocols: #50.203.2-BN43 38/06, 84-02.04.2011-A313).

First hit: transverse aortic constriction
Animals were separated into two subgroups, undergoing TAC or sham operation. Aortic banding induced cardiac hypertrophy in mice. Surgery for TAC was achieved as published previously during anesthesia at 2.5 Vol.% isoflurane [11,15]. Mice were intubated in a supine position and mechanical ventilation was initiated (MiniVent 845, Hugo Sachs Elektronik, March-Hugstetten, Germany). Ventilation was adapted to physiological parameters. A left parasternal incision was performed. Retractors were used to achieve a clear sight into the thorax and a suture was passed underneath the aorta and tied down on a 27G needle, which was immediately removed and allowed to achieve a standardized and previously validated decreased diameter of the aorta [11,15]. For sham operation procedure the suture was passed underneath the aorta without ligation. After surgery mice were monitored daily for clinical signs of infection. None of the mice included showed any kinds of healing problems following the surgery. For analgesia mice received a single intraperitoneal (i.p.) injection of 0.065 mg/kg BW buprenorphin.

Biometric measurements
The impact on cardiac biometric parameters was investigated 3 d after TAC or sham operation. Body weight was registered. Heart and lung were excised under 2.5 Vol.% isoflurane anesthesia, prepared and total heart weight, left ventricular as well as lung weights and tibia lengths were recorded immediately. Ventricles were snap frozen in liquid nitrogen and kept at -80°C.

Hemodynamic measurements
Hemodynamic parameters were recorded after 3 d of aortic banding followed by 6 h of stimulation using a 1.4 French nylon single pressure micro tip catheter (AD Instruments GmbH, Sperrbach, Germany). Catheter insertion was performed during anesthesia at 2.5 Vol.% isoflurane. Data recordings were performed at 1 Vol.% isoflurane and 1 L/min oxygen flow. For the recording of arterial and left ventricular blood pressure the catheter was inserted into the right carotid artery. First, the catheter was pushed forward to a position 4 mm in front of the aortic valve for peripheral blood pressure recordings and was then further advanced into the left ventricle. Data were analyzed using a power lab data acquisition system (AD Instruments; Software: LabChart for Windows v.6Power Lab).

Quantitative real-time PCR
In order to achieve total RNA extraction, left ventricles were homogenized and RNA was isolated using the thiocyanate-phenol-chloroform method [16]. RNA was dissolved in 100μl of RNase-free water, and concentration was determined photometrically (absorbance at 260 nm) before storage at -80°C. RNA was transcribed reversely according to the manufacturer's protocol using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA, Part No. 4368814). 25 μl RNA were mixed with 25μl master mix, containing 5 μl 10x reverse transcriptase buffer, 2 μl 25x dNTPs, 2 μl 10x random primers, 2.5 μl multi scribe reverse transcriptase and 10.5 μl nuclease free water.
We used specific pre-made TaqMan 1 Gene Expression Assays (Applied Biosystems) for GAPDH, ANP, BNP, IL-1β, IL-6, TLR1, -2, -4, -6, -9 (Table 1). Real-time PCR was performed according to the manufacturer's protocol. 5.5 ng of cDNA was mixed with 5 μl 2xTaqMan 1 Universal Master Mix (Applied Biosystems, #4304437), 0.5 μl TaqMan 1 Gene Expression Assay and 2.3 μl nuclease free water to a final volume of 10 μl in a 384-well optical reaction plate. Each sample was measured in triplicate wells and underwent 40 cycles of amplification on an ABI PRISM 1 Sequence Detection System (Applied Biosystems). C T values were determined with SDS Software 2.2 (Applied Biosystems) and relative quotients (RQ) were calculated following the ΔΔC T method (RQ target gene / GAPDH).

Data analysis and statistical procedures
All values are expressed as mean (M) ± SEM. For tests of significance between the groups, oneway analysis of variance (ANOVA) and Newman-Keuls post-hoc testing were performed. Statistics were calculated using Prism 4.05 (GraphPad Software Inc., San Diego, CA, USA). Differences between experimental groups were considered to be significant with p < 0.05.

LTA affects hemodynamic function of TAC operated mice
We studied the effect of TAC induced cardiac stress on a subsequent LTA challenge 3 d after surgery in wild type (WT) mice (Fig 1). We confirmed a good reproducibility of TAC surgery. Left ventricular systolic pressure as well as mean arterial blood pressure increased significantly in TAC PBS mice (p < 0.01, Fig 2A and 2B). Six hours after i.p. application, LTA did not cause cardiac depression in the sham group. However, left ventricular systolic and mean arterial blood pressure were reduced in LTA TAC compared to PBS TAC mice (p < 0.05, Fig 2A  and 2B) without affecting the heart rate ( Fig 2C).

Cardiac stress changes pattern recognition receptor mRNA expression after LTA treatment
Previously published data demonstrated an elevation of CD14 protein expression after 3 d of TAC, which accounts for increased susceptibility towards an LPS challenge [12]. We hypothesized that myocardial injury and stress could also alter the inflammatory response towards LTA treatment and augment pattern recognition receptor (PRR) expression and signaling.
Indeed, we observed an elevation of TLR-1, -2, -4, -6, -9 and CD14 mRNA amounts in those WT TAC mice confronted with a second hit LTA stimulus, reaching the level of significance except for TLR9 (Table 2, Fig 3). In contrast no major differences between sham PBS and TAC PBS expression were detected on mRNA levels.

TAC leads to increased cardiac inflammation upon second hit lipoteichoic acid treatment
A previously conducted study elucidated an increased NFκB activation in TAC mice exposed to LPS, followed by enhanced cytokine expression. We assumed that the observed differences in PRR expression presented above might also result in a hyperinflammatory response towards LTA and CpG-ODN.
In line with previously published findings, cytokine expression was not altered anymore 3 d after TAC surgery in wild type mice [11] (Table 3, Fig 4A). Six hours of LTA treatment alone elevated the cytokine mRNA expression, with IL-1β mRNA levels increasing 13-fold compared to PBS sham specimens (p < 0.01). Highest amounts of TNFα, IL-1β, IL-6, and IL-10 mRNA were measured in the TAC LTA second hit group. IL-10 mRNA values significantly exceeded the expression levels in the sham LTA group (p < 0.05), and IL-1β, IL-6 and IL-10 mRNA amounts were significantly elevated compared to TAC PBS (p < 0.001, p < 0.05 and p < 0.01, respectively).

Moderate effect of aortic banding on second-hit CpG induced cardiac inflammation
After observing enhanced inflammation following stimulation with gram-positive and gramnegative bacterial cell wall components, we challenged the immune system with the synthetic bacterial CpG-ODN 1668-thioate signaling via TLR9. Initially, an amount of 10 nmol CpG-ODN/mouse was utilized but was discontinued due to high mortality. 30% of TAC mice died after stimulation. Out of the surviving mice, 57% died during catheter measurement. Therefore, the concentration was reduced to 5 nmol/mouse with 92% survivors after stimulation and no further loss during catheter procedure.
Expecting a CpG effect on PRR expression we measured the mRNA levels 6 h after 1668-thioate treatment. Highest levels of all PRRs tested were detected in the TAC CpG group. However, the CpG-ODN stimulus was less potent than LTA and the detected mRNA amounts did not reach the level of significance (Table 2).
CpG-ODN caused an increase of cytokines 6 h after stimulation in the sham group. In contrast to previously published data using higher concentrations, significant levels were not detected [3]. Pro-inflammatory cytokine expression data showed an overall elevation but no significant differences between sham and TAC CpG-ODN mice (Table 3). However, IL-10 mRNA levels of TAC CpG-ODN mice exceeded those obtained from the other treatment groups (p < 0.01).
Five nmol CpG-ODN per mouse had no influence on left ventricular systolic pressure ( Fig  5A) whereas high concentrations led to cardiac depression in previous experiments [17]. A moderate increase of heart rate from 454.2 ± 18.32 s -1 in PBS to 531.5 ± 17.05 s -1 in CpG-ODN sham mice was detected (n.s.; Fig 5B).

Cardiac function of TAC operated Tlr4 -/mice is not altered after LTA or CpG-ODN stimulation
It has been shown that a lack of TLR4 signaling diminishes cardiac hypertrophy following aortic banding [7,8]. We confirmed this finding after 3d of TAC in female mice and observed a significantly reduced left ventricular weight/tibia length index (p < 0.05, Fig 6A) as well as left ventricular systolic pressure (p < 0.05, Fig 6B and 6C) in TAC Tlr4 -/mice compared to wild type mice. Second-hit treatment with LTA or CpG-ODNs did not alter left ventricular systolic pressure or other hemodynamic parameters in the TAC Tlr4 -/group (Fig 6B and 6C).

TLR4 deficiency prevents hyperinflammatory state in response to second-hit virulence factor challenge
In Tlr4 -/mice second-hit LTA treatment induced only a moderate alteration of cytokine mRNA expression. Detected levels did not significantly exceed the amounts measured in sham LTA or TAC PBS Tlr4 -/mice ( Fig 4A). Synthetic bacterial DNA also did not cause a hyperinflammatory reaction. CpG-ODN elevated TNF and IL-1β mRNA values in sham mice (p<0.05, Fig 4B). However, mRNA amounts in TAC CpG-ODN mice did not rise above those measured in the sham CpG-ODN group.

Discussion
We hypothesized that cardiac stress and injury after aortic banding enhance the immune response towards bacterial TLR2 and TLR9 ligands. Furthermore, we assumed that Tlr4 -/mice responding to TAC with reduced cardiac hypertrophy and inflammation are protected from a LTA or CpG-ODN induced hyperinflammatory cardiac response. In accordance with these hypotheses our findings indicate that cardiac pressure overload enhances the inflammatory response towards the gram-positive cell wall component LTA and synthetic bacterial DNA. After aortic banding and second-hit LTA stimulation blood pressure decreased while PRR and cytokine expression were elevated. Furthermore, this study demonstrates for the first time that a lack of TLR4 signaling does not only reduce cardiac hypertrophy and inflammation but also protects Tlr4 -/mice from hyper-responsiveness towards subsequent virulence factor challenges.
Measurements of hemodynamic and molecular biological parameters were performed 6 h after virulence factor treatment. In a previously conducted investigation applying LPS as a second-hit stimulus, strong alterations of hemodynamic function as well as cytokine secretion were observed 6 h after second-hit application [12]. Due to matters of comparability, we decided to also perform hemodynamic measurements and tissue harvesting 6 h after TLR ligand application. Recent reports revealed that application of ultrapure LTA seems to be a less potent stimulus for cardiac inflammation, depression and vascular contractility compared to  LPS [5,13]. NFκB activation and pro-inflammatory cytokine mRNA expression and protein secretion reached their maximum 2-4 h after LTA stimulation and returned to baseline after 6 h. This observation is in line with our results and might explain why only modest alteration of cytokine protein expression was detectable after LTA treatment. Interestingly, in the study presented here application of LTA as a second-hit significantly reduced systolic left ventricular pressure as well as mean arterial pressure. Thus, our investigation revealed that the inflammatory and cardiac depressive effect of LTA is potentiated in a previously injured heart. Therefore, gram-positive bacterial pathogens might be a major contributor to organ dysfunction and inflammation after chronic stress and tissue damage. Compared to LTA or LPS second-hit stimulation we observed only a moderate inflammatory response and no cardiac depression upon synthetic bacterial DNA challenge. Instead a pronounced increase of anti-inflammatory IL-10 was detected after TAC. In this survey, the effect of CpG-ODN was generally weak. Compared to previous studies, we used a considerably lower dose (200 nmol/kg vs. 400 or 800 nmol/kg bodyweight). Formerly published studies performed with higher CpG-ODN dosages (800 nmol/kg bodyweight) also revealed only a moderate induction of cytokines 6 h after stimulation [3,17], whereas strongest effects were observed after 1-2h post treatment. Cardiac depression was detected in mice injected with 800 nmol CpG-ODN/kg bodyweight after 6h [17]. Therefore, stronger CpG-ODN effects on pro-inflammatory cytokine expression and potential cardiac depression may have been detectable at earlier time points. However, in this study we chose low CpG-ODN concentrations since we observed increased mortality after treatment with 200 nmol CpG-ODN/kg bodyweight.
Former publications demonstrated, that NFκB transcription factor activity and pro-inflammatory cytokine secretion have returned to baseline 3 d past TAC surgery [11,12]. Hence, the hyperinflammatory response is not caused by prolonged increased activity of proinflammatory  mediators after the onset of TAC. However, sustained alteration in leukocyte populations occurred during the pathogenesis of LV hypertrophy [18]. CD45 + immune cells remained elevated up to 21 d after surgical intervention with monocyte/macrophage and neutrophil numbers increasing significantly on day 3 after TAC. The major proportion of macrophages in the heart was of the Ly6C low /M2 type, cells secreting anti-inflammatory cytokines and contributing to tissue repair [19]. In contrast, TAC elevated the amount of Ly6C high monocytes circulating in the blood [18]. These monocytes exhibit phagocytic and pro-inflammatory functions. Thus, an accelerated and increased pro-inflammatory response towards systemically administered bacterial pathogens might arise from a higher abundance of leukocytes in TAC mice. Depending on ligand and concentration, bacterial pathogens induce trained immunity or cause tolerance, partially by epigenetic modification of monocytes [20]. The effects of endogenous TLR ligands remain unknown, but may also alter cellular functional programs and epigenetic events.
We observed that TAC Tlr4 -/mice responded with an attenuated inflammatory response towards bacterial pathogen stimulation. TLR4-mediated recognition of host-derived molecules such as fibrinogen activates the immune system and sustained TLR4 activation induces cardiac hypertrophy [21]. Recent studies utilizing different models of cardiac hypertrophy revealed, that MCP-1 and monocyte/macrophage recruitment depended on TLR4 signaling [22,23]. Therefore the lack of leukocyte accumulation in TAC Tlr4 -/mice might explain the absence of a hyperinflammatory response upon second-hit treatment in our model. A recent publication by Jiang et al. reported that cardiac signal regulatory protein (SIRP)-α protects against cardiac hypertrophy via the disruption of toll-like receptor 4 signaling [24]. Contrariwise, TLR4-disruption could compensate for the detrimental effects on cardiac hypertrophy in SIRP-α knockdown hearts. SIRP-α is also abundant in immune cells and negatively regulates macrophage activation [25]. These data may suggest a link between TLR4 signal transduction and SIRP-α in our second-hit model. If TLR4-deficient mice exhibit higher SIRPα levels, this down-regulates innate immune responses. In contrast, TAC wild type mice with reduced SIRP-α levels may develop an augmented, prolonged inflammatory response upon second-hit TLR ligand treatment.
We assumed that pressure overload alters the expression of PRRs. Release of endogenous ligands after TAC with subsequent TLR activation may lead to selective or cross-regulation of TLRs. In TAC wild type mice, LTA TLR2 agonist treatment caused substantial cross-regulation of all PRR mRNA levels measured whereas Tlr4 -/mice responded with a differential upregulation of TLR2 mRNA expression. The transcriptional regulation of TLR remains widely unknown. Studies elucidating the negative regulation of TLRs mostly dealt with the dissociation of adaptor complexes, degradation of signal proteins and transcriptional regulation of mediators downstream of TLR [26]. This study now extends this knowledge to the transient regulation of PRRs in a second hit model. Our data here complement the previous findings of PRR regulation in the vasculature during inflammation [13]. Based on our findings, prospective studies will interrogate the regulation of leukocyte recruitment, activation, and function in models of tissue injury predisposing to secondary infections. Mechanistical analyses need to discover whether a transient modulation of TLR4 signal transduction might offer new possibilities for the better use of safe and efficient TLR4 agonists.