p53 signalling controls cell cycle arrest and caspase-independent apoptosis in macrophages infected with pathogenic Leptospira species
Division of Basic Medical Microbiology, State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
Department of Medical Microbiology and Parasitology, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
Department of Medical Microbiology and Parasitology, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
Division of Basic Medical Microbiology, State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
Pathogenic Leptospira species, the causative agents of leptospirosis, have been shown to induce macrophage apoptosis through caspase-independent, mitochondrion-related apoptosis inducing factor (AIF) and endonuclease G (EndoG), but the signalling pathway leading to AIF/EndoG-based macrophage apoptosis remains unknown. Here we show that infection of Leptospira interrogans caused a rapid increase in reactive oxygen species (ROS), DNA damage, and intranuclear foci of 53BP1 and phosphorylation of H2AX (two DNAdamage indicators) in wild-type p53-containing mouse macrophages and p53-deficient human macrophages. Most leptospire-infected cells stayed at the G1 phase, whereas depletion or inhibition of p53 caused a decrease of the G1-phase cells and the early apoptotic ratios. Infection with spirochaetes stimulated a persistent activation of p53 and an early activation of Akt through phosphorylation. The intranuclear translocation of p53, increased expression of p53-dependent p21Cip1/WAF1 and pro-apoptotic Bcl-2 family proteins (Bax, Noxa and Puma), release of AIF and EndoG from mitochondria, and membrane translocation of Fas occurred during leptospire-induced macrophage apoptosis. Thus, our study demonstrated that ROS production and DNA damage-dependent p53-Bax/Noxa/Puma-AIF/EndoG signalling mediates the leptospire-induced cell cycle arrest and caspase-independent apoptosis of macrophages.
Leptospirosis, caused by infection by pathogenic Leptospira species, is a zoonotic disease of global importance (Bharti et al., 2003). The disease has been spreading in South-East Asia and South America (McBride et al., 2005; Zhang et al., 2012a). However, in recent years, leptospirosis cases have also been reported in Europe and North America (Meites et al., 2004; Ko et al., 2009; Hotez and Gurwith, 2011).
After invasion into human and animal hosts, pathogenic leptospires enter the bloodstream rapidly to generate septicaemia, and then spread into internal organs such as the lungs, liver and kidneys, where they cause tissue injury (Adler and Moctezuma, 2010). Unlike the nearly asymptomatic infection in host animals, pathogenic leptospires cause disease in humans with mild to rapidly fatal forms including ‘flu-like’ clinical manifestations such as fever and myalgia, and severe syndromes such as septic shock, and respiratory and renal failure (McBride et al., 2005; Adler and Moctezuma, 2010). Pulmonary diffuse haemorrhage (PDH), a clinically severe type of human leptospirosis, usually causes mortality rates of 40% to 50% (McBride et al., 2005; Zhang et al., 2012a). However, despite the high incidence of leptospirosis and the severity of disease, the molecular and cellular basis of pathogenesis during Leptospira infection remain poorly understood.
Phagocytes play a critical role in innate immunity against microbial infection. Mononuclear macrophages and neutrophils have been shown to phagocytose leptospires, but only the former can kill the phagocytosed intracellular leptospires (Wang et al., 1984; Davis et al., 2009), indicating that mononuclear macrophages are much more important than neutrophils in defence mechanisms against leptospiral infection. However, Merien et al. first reported that Leptospira interrogans, a common species of pathogenic Leptospira, could cause macrophage apoptosis (Merien et al., 1997). Our previous studies demonstrated that L. interrogans serovar Lai strain Lai induced apoptosis of human and mouse macrophages through both the caspase-8/3-dependent and caspase-independent mitochondrion-related apoptosis inducing factor (AIF) and/or endonuclease G (EndoG) pathways, but the classifical mitochondrion-related cytochrome c (Cyt c)-caspase-9/3 pathway was not involved in apoptosis (Jin et al., 2009; Li et al., 2010). Previous studies revealed that AIF mediates apoptosis of Mycobacterium bovis-infected bovine macrophages, Porphyromonas gingivalis-infected fibroblasts, Streptococcus pneumoniae-infected brain cells and Helicobacter pylori-induced gastric cancer cells (Braun et al., 2001; Desta and Graves, 2007; Vega-Manriquez et al., 2007; Ashktorab et al., 2008), but the role of EndoG in microbe-induced caspase-independent apoptosis has not been reported yet. AIF is an apoptotic effector that causes chromatin condensation and large-scale DNA fragmentation, while EndoG is an apoptotic DNase (Daugas et al., 2000; Lily et al., 2001). Both reside within mitochondria, but are released to the cytosol during initiation of apoptosis. p53, an important tumorigenesis-associated protein, has been shown to promote the release of AIF and EndoG (Kook et al., 2007). However, the mechanisms microbe-induced p53-mediated AIF/EndoG-dependent host-cell apoptosis remain unknown.
Cell cycle arrest and apoptosis are considered to be the two major responses to a variety of genotoxic stimuli including microbial infection, because the former provides an opportunity to repair the damaged DNA while the latter eliminates the cells that have undergone irreparable DNA damage (Alenzi, 2004). Although the regulation of the cell cycle is complicated, the Akt-p53-signalling network plays a critical role in survival, growth, proliferation or apoptosis of cells in which the function of Akt is generally opposite to that of p53 protein (Oren, 2003; Wee et al., 2009). It has been reported that P. gingivalis and influenza virus activate the Akt signalling pathway in epithelial cells at early stages of infection to protect the cells from apoptotic death, but influenza virus could also activate p53-dependent or -independent signalling to trigger apoptosis of host cells with irreversible injury (Yilmaz et al., 2004; Zhirnov and Klenk, 2007).
At the same time, p21Cip1/WAF1 and p27Kip1, the two major inhibitory proteins of cell cycle progression, can inhibit cell cycle progression by inactivation of cyclin-dependent kinases (CDKs) that induce cell division and proliferation (Hiromura et al., 1999; Gartel and Tyner, 2002). Akt-signalling promotes cell survival and cell cycle progression by inhibition of p21Cip1/WAF1 and p27Kip1 (Zhou et al., 2001; Shin et al., 2002). p53 mediates cell cycle arrest by upregulation of p21Cip1/WAF1; cell apoptosis by induction of pro-apoptotic proteins belonging to the Bcl-2 family such as Puma, Noxa and Bax; and recruitment of death receptor (Fas) from the cytosol to the cell surface (Bennett et al., 1998; Mihara et al., 2003; Meek, 2004). The disequilibrium caused by either upregulation of pro-apoptotic Bcl-2 family proteins or downregulation of anti-apoptotic Bcl-2 family proteins results in the release of Cyt c, high-temperature requirement A2 (HtrA2), second mitochondrial derived activator of caspases (Smac), AIF and/or EndoG from mitochondria (Kroemer and Martin, 2005; Li and Yuan, 2008). The Cyt c or HtrA2 and Smac mediate cell apoptosis by activation of caspase-9 or release of inhibition of caspase-3 (Li and Yuan, 2008; Jin et al., 2009), while AIF and EndoG trigger cell apoptosis through direct DNA cleavage (Daugas et al., 2000; Lily et al., 2001; Kroemer and Martin, 2005). High levels of intracellular reactive oxygen species (ROS) can cause DNA damage (Liu et al., 2008), which is a powerful activator of both p53 and Akt by phosphorylation (Wang et al., 2000; Zhou et al., 2001; Shin et al., 2002; Meek, 2004). Activated p53 mediates cell apoptosis through mitochondrion-dependent apoptotic pathways (Meek, 2004; Helton and Chen, 2007). It has been shown that infection of L. interrogans stimulates the production of high levels of ROS in rat Kupffer cells in liver tissues (Marangoni et al., 2006). Thus, we hypothesized that ROS-based DNA damage in leptospire-infected macrophages could also activate Akt and p53, and thereby lead to cell cycle arrest and AIF/EndoG-dependent cell apoptosis.
Therefore, in the present study, a wide-type p53-containing mouse macrophage line (J774A.1), and a p53-deficient (mutational) human monocyte line (THP-1) (Akashi et al., 1999) were used as host cells to characterize oxidative stress and ROS-based DNA damage due to L. interrogans infection, the roles of Akt- and p53-signalling pathways in cell cycle arrest through regulation of p21Cip1/WAF1 and p27Kip1, release of apoptotic inducers (AIF and EndoG) form mitochondria through regulation of Bcl-2 family proteins, and AIF and EndoG-mediated caspase-independent apoptosis.
Increase of ROS levels in leptospire-infected cells
It has been shown that infection of bacterial pathogens usually causes oxidative stress and an increase of ROS levels of host cells, which can be measured with the ROS-specific fluorescent dye (DCFH-DA) (Li et al., 1999; Marangoni et al., 2006). The ROS-specific fluorescent staining assay confirmed that ROS levels were rapidly increased in J774A.1 and THP-1 cells at 0.5 h after infection with L. interrogans strain Lai, and the maximal intracellular ROS levels appeared after 4 h of infection (Fig. 1A and B). The leptospire-induced ROS increase could be significantly blocked by N-acetylcysteine (NAC), an anti-oxidant, at the early stages of infection (Fig. 1A and B). Upon co-incubation of J774A.1 or THP-1 cells with L. interrogans strain Lai, the spirochaetes could be phagocytosed by the cells and form intracellular leptospire-containing phagocytotic vesicles (Fig. 1C and E). The NAC pre-treatment did not affect the spirochaete invading the cells, as the anti-oxidant in the cultures was removed prior to the infection by washing step (Fig. 1C and D). These data suggest that infection of L. interrogans induces ROS production in macrophages during phagocytosis.
Rapid 53BP1 foci formation and H2AX phosphorylation in leptospire-infected cells
ROS-dependent DNA damage has been shown to cause the intranuclear foci of p53-binding protein 1 (53BP1) and phosphorylation of histone H2AX (Rogakou et al., 1998; Schultz et al., 2000). The 53BP1 foci in the nucleus and H2AX phosphorylation occurred in the THP-1 or J774A.1 cells at the 0.5 h time point after infection with L. interrogans strain Lai (Fig. 2). When ROS in the two leptospire-infected cells was neutralized by NAC, the intranuclear foci of 53BP1 were absent (Fig. 2A and B) while the phosphorylation of H2AX was significantly reduced (Fig. 2C and D). The formation of 53BP1 foci and H2AX phosphorylation imply that infection by L. interrogans causes ROS-based DNA damage of infected macrophages.
Cell cycle arrest and DNA damage in leptospire-infected cells
The results of cell cycle determination by flow cytometry confirmed that the J774A.1 or THP-1 cells displayed different cell cycle phases (G0/G1, S or G2/M) before infection with L. interrogans strain Lai. However, the leptospire-infected J774A.1 cells displayed a rapid disappearance of the G2/M phase at the 0.5 h time point and an accumulation of G0/G1-phase cells within 1 h after infection (Fig. 3). Knock-down by siRNA interference or inhibition with the p53 inhibitor (pifithrin-α) of p53 gene in the leptospire-infected J774A.1 cells led to a remarkably delayed G2/M phase disappearance, which was similar to that found in the leptospire-infected p53-deficient THP-1 cells, while blockage of Akt with the Akt-inhibitor IV in the leptospire-infected J774A.1 or THP-1 cells caused a significant decrease of S-phage cells at the early stage of infection (Fig. 3). When infected for 1 h, the sub-G1-phase cells, which correspond to cells undergoing DNA fragmentation (Kajstura et al., 2007), appeared in both the J774A.1 and the THP-1 cells, which was followed by a persistent increase in the proportion of sub-G1-phase cells and a continuous decrease of G1-phase cells (Fig. 3). The neutralization of ROS with the anti-oxidant NAC caused a remarkably delayed appearance of sub-G1-phase cells and disappearance of G2/M phase cells (Fig. 3). These data suggest that infection with L. interrogans caused DNA damage and p53-mediated early cell cycle arrest in monocytes or macrophages.
Apoptosis of leptospire-infected cells
The proportion of leptospire-infected J774A.1 and THP-1 cells undergoing early apoptosis and late apoptosis/necrosis was determined by flow cytometry. Both cell types appeared in early apoptosis after 1 h of infection, and the maximal early apoptotic ratio for J774A.1 cells (59.1%) or for THP-1 cells (35.2%) was observed at the 4 or 8 h time point of post-infection. When the caspases were blocked before infection with the pan-caspase inhibitor Z-VAD-FMK (Martinet et al., 2006), the maximal early apoptotic ratio of the leptospire-infected J774A.1 or THP-1 cells was 22.7% or 8.8% respectively (Fig. 4A and B). However, the p53-deficient cells or cells treated with the p53 inhibitor (pifithrin-α) also showed a similar decrease of the maximal early apoptotic ratios (38.7% or 41.2%) following infection of J774A.1 cells. Much lower early apoptotic ratios were observed in the leptospire-infected J774A.1 cells when both p53 and caspases were inhibited or depleted (ratios of 2.5% to 4.9% for blockage of both p53 and caspases, and 1.7% to 5.6% for depletion of the p53 gene and blockage of caspases). These data suggest that p53 mediates leptospire-induced, caspase-independent macrophage apoptosis.
Upregulation of p53-associated genes during infection
Both the RT-qPCR and the Western blot assays confirmed that the levels of mRNA transcription and protein expression of p21, p27, bax, noxa and puma genes were dramatically increased during infection of J774A.1 cells with L. interrogans strain Lai, whereas the mRNA and protein levels of bcl-2 gene were significantly downregulated (Fig. 5). In particular, the level of p53 expression increased gradually in the leptospire-infected J774A.1 cells, but the level of p53 mRNA remained unaffected. When p53 of J774A.1 cells was depleted by siRNA interference or was blocked by the p53 inhibitor pifithrin-α, the increase of mRNA transcription or protein expression of p21, bax, noxa or puma gene upon infection was diminished (Fig. 5). However, the leptospire-infected THP-1 cells displayed less upregulation of p21 or p27 expression and a significant decrease of Bcl-2 expression at the later stages of infection, compared the leptospire-infected J774A.1 cells. Furthermore, the levels of both akt mRNA and protein showed insignificant changes in the J774A.1 or THP-1 cells before or after infection. These data imply that p53 and some p53-associated genes (p21, Bax, noxa and puma) are involved in initiation of macrophage apoptosis during Leptospira infection.
p53 and Akt activation during infection
Specific serine (Ser) residues are phosphorylated during activation of either p53 or Akt, and the activated p53 needs to transfer from the cytosol into the nucleus (Oren, 2003; Meek, 2004; Wee et al., 2009). Our Western blot assays confirmed that the infection by L. interrogans strain Lai caused a rapid, persistent phosphorylation of the Ser15, and a delayed but persistent phosphorylation of Ser20 and Ser46 on the p53 protein in J774A.1 cells; but no p53 phosphorylation, as expected, in the p53-depleted or p53-inhibited leptospire-infected J774A.1 cells or leptospire-infected THP-1 cells was observed (Fig. 6). However, the wild-type or p53-deficient or p53-inhibited J774A.1 cells or THP-1 cells displayed transient Akt phosphorylation on Ser473 only during the early stages of infection (Fig. 6). These data demonstrate that infection by L. interrogans caused persistent activation and nuclear translocation of p53, as well as transient activation of Akt in macrophages.
Release of AIF and EndoG from mitochondria
Release from mitochondria is a prerequisite for either AIF or EndoG to induce caspase-independent cell apoptosis (Daugas et al., 2000; Lily et al., 2001). Our Western blot assays demonstrated that both AIF and EndoG in the L. interrogans strain Lai-infected J774A.1 cells but only AIF in the infected THP-1 cells were released from the mitochondria into the cytosol (Fig. 7). When p53 of J774A.1 cells was depleted by siRNA interference or blocked by the p53 inhibitor pifithrin-α, the release of AIF and EndoG was delayed and barely decreased (Fig. 7). However, no HtrA2 and Smac could be detected in the cytosol during the infection (Fig. 7). These data indicate that there is a selectivity of apoptotic effectors released from mitochondria in macrophages during L. interrogans infection.
p53-mediated membrane translocation of Fas
p53 also promotes cell apoptosis by inducing trafficking of Fas from the cytosol to the cell surface (Bennett et al., 1998). The flow cytometric results in this study confirmed that total Fas or FasL and surface membrane Fas or FasL in the J774A.1 cells were significantly increased at the later stages of infection by L. interrogans strain Lai (Fig. 8), which were consistent with the increase of fas or fasL mRNA level determined by the RT-qPCR (Fig. 5A). However, the siRNA depletion of p53 gene or inhibition of p53 protein with the inhibitor pifithrin-α only caused a decrease of surface membrane Fas in the leptospire-infected J774A.1 cells (Fig. 8). Although the total Fas or FasL in THP-1 cells increased at the later stages of infection, no elevation of cell surface Fas or FasL was found (Fig. 8). These data suggest that p53 can mediate apoptosis of macrophages by recruiting Fas rather than FasL to the cell surface at later stages of infection of L. interrogans.
Infection results from an interaction between microbial pathogens and their hosts. In individuals with no specific antimicrobial immunity, phagocytes play a critical role in eliminating the invading pathogens by phagocytosis. However, mononuclear macrophages, but not neutrophils, can kill the phagocytosed leptospires (Wang et al., 1984; Davis et al., 2009). Therefore, the interaction between leptospires and mononuclear macrophages is an important event during leptospirosis that decides the outcome of infection. For instance, some pathogens, such as L. interrogans and M. bovis, can induce apoptosis of macrophages during infection, which contributes to pathogenicity and immune escape of the pathogens (Navarre and Zychlinsky, 2000; Vega-Manriquez et al., 2007; Jin et al., 2009).
During phagocytosis of microbial pathogens, phagocytes produce a high level of ROS to induce cellular oxidative stress and promoting the antimicrobial response (Forman and Torres, 2002; Lee et al., 2003). Oxidative stress has been observed in macrophages in the process of phagocytosing bacterial pathogens such as Mycobacterium tuberculosis, Pseudomonas aeruginosa and L. interrogans (Marangoni et al., 2006; Dympna et al., 2007; Shin et al., 2010). However, high levels of intracellular ROS can also damage host DNA (Cooke et al., 2003; Liu et al., 2008). In the present study, we show that there is a correlation between gradually elevated ROS levels with continuously increased DNA fragmentation (sub-G1 phase) in the L. interrogans strain Lai-infected human and mouse mononuclear macrophages (THP-1 and J774A.1 cells respectively), and that an antioxidant, NAC, delayed the appearance of the sub-G1 phase and disappearance of the G2/M phase (Figs 1A and B and 3). In addition, the formation of intranuclear foci of p53-binding protein 1 (53BP1) and phosphorylation of histone H2AX, two early indicators of ROS-based DNA damage (Rogakou et al., 1998; Schultz et al., 2000), were also observed in the cells during early stages of infection (Fig. 2). NAC caused the disappearance of 53BP1 foci and the decrease of H2AX phosphorylation level (Fig. 2). The data indicate that L. interrogans infection stimulates oxidative stress that results in ROS-dependent DNA damage in the infected cells.
Previous studies proved that p53 responds to DNA damage through rapid upregulation of p21Cip1/WAF1 in order to inhibit cell cycle progression for DNA repair (Cooke et al., 2003; Achanta and Huang, 2004; Meek, 2004), while the Akt kinase pathway, which is also activated by ROS, promotes cell survival and cell cycle progression through inhibition of p21Cip1/WAF1 and p27Kip1 (Wang et al., 2000; Zhou et al., 2001; Shin et al., 2002). As shown in our results, the macrophages responded to the infection of L. interrogans strain Lai by a persistent G0/G1 phase cell accumulation and a rapid G2/M phase cell disappearance, while blockage of p53 caused fewer cells to stay at the G0/G1 phase and a delayed disappearance of G2/M phase (Fig. 3). However, blockage of Akt did not influence significantly the cell cycles in the leptospire-infected cells. The RT-qPCR and immunoblot assays also confirmed the rapid and persistent increase in the p53-induced p21Cip1/WAF1 protein in leptospire-infected cells, whereas Akt-induced p27Kip1 was increased transiently with a relative lower level during the infection (Fig. 5). These data indicated that infection by the spirochaetes induces p53/p21Cip1/WAF1-dependent cell cycle arrest in macrophages.
AIF and EndoG are released from mitochondria to cause caspase-independent DNA cleavage under the action of pro-apoptotic Bcl-2 family proteins such as Bax, Noxa and Puma, which are upregulated directly by activated p53 (Miyashita and Reed, 1995; Oda et al., 2000a; Nakano and Vousden, 2001). We recently observed that AIF and/or EndoG participated in the L. interrogans-induced caspase-independent macrophage apoptosis (data not shown), but the signalling pathways responsible for release of the two apoptotic effectors was not determined. In the present study, we found that the inhibition of caspases and depletion or inhibition of the p53 gene caused a significant decrease in apoptosis in L. interrogans strain Lai-infected macrophages (Fig. 4). Consistent with these results, immunoblot analysis confirmed the upregulation of Bax, Noxa or Puma protein expression and downregulation of Bcl-2 expression, as well as the release of AIF and EndoG (but not HtrA2 and Smac) from mitochondria in the cells during infection (Figs 5 and 7). In particular, compared the wild-type control cells, the depletion or inhibition of p53 abrogated the increase in transcription of the p21, bax, noxa or puma genes, and reduced significantly the release of AIF and EndoG during infection (Figs 5 and 7). Previous studies confirmed that p53-mediated Bax-dimerization on the mitochondrial membrane could induce the release of AIF and EndoG, while Noxa and Puma activated by p53 could promote the release of AIF and EndoG through inhibition of the anti-apoptotic Bcl-2 (Miyashita and Reed, 1995; Oda et al., 2000a; Nakano and Vousden, 2001; Stambolsky et al., 2006). These data suggested that the p53-Bax/Noxa/Puma-AIF/EndoG pathway may mediate signalling during leptospire-induced caspase-independent macrophage apoptosis.
Both p53 and Akt are activated through phosphorylation on certain serine residues (Ser) (Wee et al., 2009). Phosphorylation at Ser473 (Ser473-P) is an indicator of Akt activation (Zhou et al., 2001; Shin et al., 2002). Among the possible phosphorylation sites in p53 protein, Ser15 phosphorylation (Ser15-P) is critical because it allows the subsequent phosphorylation of Ser20 (Ser20-P) and Ser46 (Ser46-P) and mediates many important biological effects of p53 such as upregulation of p21Cip1/WAF1, resistance of p53 cleavage, and induction of nuclear translocation and intranuclear stability of p53 (Shieh et al., 1997; Nakagawa et al., 1999). Ser20-P also contributes to the intranuclear stability of p53 (Meek, 2004), and Ser46-P promotes the transcription and oligomerization on the mitochondrial membrane of Bax (Oda et al., 2000b). In the present study, we observed that infection by L. interrogans strain Lai triggered the rapid phosphorylation of Ser15 and subsequent phosphorylation of Ser20 and Ser46 of p53 (Fig. 6), which enables p53 to mediate the cell cycle arrest at G0/G1 phase and caspase-independent apoptosis in the macrophages during infection. Compared the persistent phosphorylation of p53, Akt activation by phosphorylation at the Ser473 occurred transiently within 1 h of post-infection (Fig. 6). In the Akt-p53-signalling network, p53 signalling mediates cell apoptosis, whereas Akt signalling induces cell survival and proliferation (Shin et al., 2002; Oren, 2003; Wee et al., 2009). Previous studies have shown that Akt activation may be the primary reaction of host cells against microbial infection, but the infected host cells can die due to the irreversible cell injury mediated by p53 signalling (Yilmaz et al., 2004; Zhirnov and Klenk, 2007). Therefore, persistent p53 activation and early transient Akt activation in the leptospire-infected macrophages indicates the irreversible cell injury and subsequent apoptosis upon prolonged infection.
Another mechanism of p53-induced apoptosis is the upregulation of Fas and translocation of Fas from the cytosol to cell surface (Owen-Schaub et al., 1995; Bennett et al., 1998). Ligation of Fas by FasL on adjacent cells typically induces apoptosis through the FADD-caspase-8/3 pathway (Brunner et al., 1995). Ligation of Fas by FasL on adjacent cells typically induces apoptosis through the FADD-caspase-8/3 pathway (Brunner et al., 1995). In this study, we showed that the expression and membrane translocation of both Fas and FasL in infected macrophages were increased at the later stages of infection (Fig. 8). However, in the p53-depleted or -inhibited macrophages, only Fas membrane translocation was decreased significantly, not the total Fas, FasL expression or FasL membrane translocation (Fig. 8). This is different from a previous report showing that p53 upregulates Fas expression (Owen-Schaub et al., 1995). Thus, our findings suggest that p53 signalling in macrophages during L. interrogans infection promotes the Fas membrane translocation, but not responsible for the upregulation of total Fas or FasL as well as FasL membrane translocation.
It is well known that p53 mediates cell cycle arrest and apoptosis in tumour cells (Lai et al., 2007). The results from our study also showed that p53 mediates cell cycle arrest and apoptosis of macrophages upon L. interrogans infection. At early stages of leptospirosis, leptospires can be found in patients' bloodstream associated with macrophages, inside or outside (Faine et al., 1999); and we also found that not all leptospires are phagocytosed by J774A.1 or THP-1 macrophages. It has been shown that binding of bacterial adhesins to their receptors on macrophages is the first step for phagocytosis, and both the adherence and the phagocytosis can cause the elevation of intracellular ROS levels (Süßmuth et al., 2000; Rosenberger and Finlay, 2003). In this study, we confirmed that the Leptospira-induced ROS can promote p53 activation and subsequent p53-dependent apoptotic signalling in macrophages. Taken together, the results from this study demonstrated that infection by pathogenic Leptospira species, probably due to adherence and phagocytosis of the spirochaete, leads to the production of ROS and DNA damage in infected macrophages, and that the DNA damage-activated p53 mediates mitochondria-dependent caspase-independent macrophage apoptosis through the p53-dependent Bax/Noxa/Puma-AIF/EndoG signalling pathway.
Leptospiral strain and culture conditions
Pathogenic L. interrogans serogroup Icterohaemorrhagiae serovar Lai strain Lai was provided by National Institute for Control of Pharmaceutical and Biological Products in Beijing, China. The leptospiral strain was cultured at 28°C in Ellinghausen–McCullough–Johnson–Harris (EMJH) liquid medium supplemented with 5% albumin bovine fraction V (Sigma, USA) and 0.05% Tween-80 (Difco, USA).
Cell lines and culture conditions
A human monocyte line (THP-1) with mutated p53 gene, and a mouse mononuclear macrophage-like cell line (J774A.1) containing wide-type p53 gene, were provided by the Cell Bank of Institute of Cytobiology, Chinese Academy of Science. The cells were maintained in RPMI-1640 liquid medium (GiBco, USA) supplemented with 10% fetal calf serum (FCS, GiBco), 100 U ml−1 penicillin (Sigma) and 100 μg ml−1 streptomycin (Sigma) in an atmosphere containing 5% CO2 at 37°C. For differentiation of THP-1 cells into macrophage-like cells, the cells were pre-treated with 10 ng ml−1 phorbol 12-myristate 13-acetate (PMA, Sigma) at 37°C for 48 h before use in all experiments, except for experiments for cell cycle detection (DeCoursey et al., 1996).
Detection of intracellular ROS
Dichlorofluorescein diacetate (DCFH-DA) (Sigma), an ROS-specific fluorescent dye, was used to measure the ROS levels in leptospire-infected cells (Li et al., 1999). Briefly, freshly cultured L. interrogans strain Lai was harvested by a 17 200 g centrifugation step at 15°C for 15 min. After washing with autoclaved 0.01 M phosphate-buffered saline (PBS, pH 7.4), the leptospiral precipitate was suspended in antibiotic-free 2.5% FCS RPMI-1640 liquid medium for counting leptospires under a dark-field microscope with a Petroff-Hausser counting chamber (Fisher Scientific, USA) (Schreier et al., 2009). THP-1 or J774A.1 cells (1 × 105 per well) were seeded in 12-well culture plates (Corning, USA) containing a 12 × 12 mm coverslip per well for incubation overnight at 37°C. The coverslips with THP-1 or J774A.1 cell monolayers were washed thoroughly with PBS and then infected with the harvested L. interrogans strain Lai (1 × 107) at a multiplicity of infection (moi) of 100 (100 leptospires per host cell) for co-incubation at 37°C for 0.5, 1, 2, 4 or 8 h (Jin et al., 2009; Li et al., 2010). After washing thoroughly with PBS, the leptospire-infected cells were incubated in antibiotic-free 2.5% FCS RPMI-1640 medium containing 5 μM DCFH-DA for 30 min at 37°C. The fluorescence intensity reflecting ROS levels in the cells was detected using a laser confocal microscope (LSM510-Meta, Zeiss, Germany) with 488 nm excitation and 530 nm emission wavelengths. For some experiments to study the effects or ROS, the THP-1 or J774A.1 cells were pre-treated for 1 h with 10 mM NAC (Sigma), an anti-oxidant (Macip et al., 2002; Yun et al., 2009), and the subsequent experimental steps such as infection with the spirochaete and detection of intracellular ROS were the same as described above.
Detection of intracellular leptospires
Laser confocal microscopy was used to observe the leptospires in THP-1 or J774A.1 cells after infection with L. interrogans strain Lai as well as to determine the influence of the anti-oxidant (NAC) upon the infection. Briefly, THP-1 or J774A.1 cells (1 × 105 per well) were seeded in 12-well culture plates (Corning) for incubation overnight at 37°C, and then the cell monolayers were infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4 or 8 h as described above. After treatment with 0.25% trypsin-PBS for 5 min to detach the cell-adhered leptospires, washing three times with PBS and centrifugation at 100 g for 10 min at 4°C, the cell pellets were fixed with 4% glutaraldehyde-PBS for 30 min, and then permeabilized with 0.1% Triton X-100-PBS for 10 min to allow antibody penetration into the cells (Zhang et al., 2012b). After washing with PBS and blocking with 5% BSA-PBS, the cells were incubated with 1:500 diluted rabbit anti-L. interrogans strain Lai-IgG prepared by our laboratory, followed by incubation with 1:1000 diluted Alexa Fluor594-conjugated donkey anti-rabbit-IgG (Invitrogen) for 1 h to stain intracellular leptospires. After washing with PBS again, the cells were incubated with 1 μg ml−1 DAPI (Sigma) for 10 min to stain cell nucleus. Finally, the cells were smeared on glass slides and then observed under a laser confocal microscope (LSM510-Meta, Zeiss) with 590/617 or 355/460 nm wavelengths (excitation/emission) for Alexa Fluor594 or DAPI detection. In the confocal microscopic detection, the cell monolayers that pre-treated with NAC as described above were used as the control to determine the potential influence of NAC on the infection efficiency of the spirochaete. In order to show the leptospire-containing phagocytotic vesicles, J774A.1 or THP-1 cells were infected with the spirochaete at an moi of 100 for 1 h. After washing with PBS and fixation with 2.5% glutaraldehyde-PBS for 2 h, the cells were scraped from the wells for a 10 min centrifugation at 250 g. The cell pellets were post-fixed, dehydrated, embedded in TAAB resin, ultrathin sectioned and stained (Zhang et al., 2012b). The intracellular leptospire-containing phagocytotic vesicles were observed under a transmission electron microscope (type TECNAI-10, Philips, Holland).
Detection of 53BP1 intranuclear foci and H2AX phosphorylation
THP-1 or J774A.1 cells were seeded in 12-well culture plates (Corning) containing a 12 × 12 mm coverslip per well for overnight at 37°C, and then infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4 or 8 h as described above. The coverslips with THP-1 or J774A.1 cell monolayers were washed with Tris-buffered saline (TBS, pH 7.4) and then fixed with 4% paraformaldehyde for 10 min at room temperature. After washing with TBS containing 0.1% Triton X-100 (TBST), the coverslips were incubated with blocking solution (5% BSA-TBST) for 4 h at 4°C, and then incubated with 1:500 diluted rabbit anti-53BP1-IgG (AbCam, USA) or rabbit anti-pSer139/H2AX-IgG (Cell Signaling Technology, USA) for overnight at 4°C. After washing thoroughly with TBST, the coverslips were incubated with 1:1000 diluted Alexa FluorTM488-conjugated anti-rabbit-IgG (Cell Signaling Technology), and then stained with 0.1 μM rhodamine phalloidin (Cytoskeleton, USA) and 1 μg ml−1 DAPI (Cell Signaling Technology) for 1 h in the dark. Finally, the coverslips were observed under a laser confocal microscope (LSM510-Meta, Zeiss) with 488/510, 535/585 or 355/460 nm wavelengths (excitation/emission) to detect the 53BP1 foci or phosphorylated H2AX, or to show cytoplasm and nucleus respectively (Rogakou et al., 1998; Schultz et al., 2000). For oxidative stress experiments, THP-1 or J774A.1 cells were pre-treated for 1 h with 10 mM NAC (Sigma) (Macip et al., 2002; Yun et al., 2009). The subsequent infection with the spirochaetes, and detection of 53BP1 and H2AX, were the same as described above.
p53 gene depletion in J774A.1 cells
The p53 gene of J774A.1 cells was depleted by siRNA interference. The control siRNAs (catalogue No.: D-001206-13) and mouse p53 siRNAs (catalogue No.: M-040642-02) from siGENOME SMARTpool database and a siRNA Transfection Kit used in this assay were provided by Thermo Scientific (USA). Briefly, J774A.1 cells (5 × 105 per well) were seeded in six-well culture plates (Corning) for incubation at 37°C. When the cells were 60% confluent, 25 nM p53 siRNAs were transfected into cells for p53 gene silencing according to the manufacturer's protocol. A Western blot assay using rabbit anti-mouse p53 IgG (Cell Signaling Technology) as the primary antibody and horseradish peroxidase (HRP)-conjugated goat anti-rabbit-IgG as the secondary antibody (Jackson ImmunoResearch, USA) was performed to confirm the depletion of the p53 gene (Niu et al., 2005).
Determination of cell cycle status
THP-1 and wild-type or p53-depleted J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4, 8, 12 or 24 h were harvested by a 250 g centrifugation step after trypsinization. After washing thoroughly with cold PBS and centrifugation, the cell precipitates were suspended in PBS and then fixed with cold 70% ethanol at −20°C for 30 min followed by rehydration with PBS for 15 min at room temperature. The cells were stained with propidium iodide (PI) cell cycle staining solution (100 μg ml−1 RNase A and 50 μg ml−1 PI in PBS, Sigma). The cell cycle status of the stained cells was examined using a flow cytometer (FC500 MCL, Beckman, USA) and determined using Multicycle software (Phoenix Flow Systems, USA) (Hsu et al., 2004). On the other hand, THP-1 and J774A.1 cells were pre-treated with 10 μM Akt inhibitor IV (Sigma) (Ruhland et al., 2007) or 10 mM NAC (Sigma) (Macip et al., 2002; Yun et al., 2009), and J774A.1 cells were pre-treated with 30 μM p53 inhibitor pifithrin-α (Sigma) (Martinet et al., 2006). The subsequent infection with the leptospiral strain, cell staining and cycle detection were performed the same as described above.
Detection of cell apoptosis
THP-1 and wild-type or p53-depleted J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 1, 2, 4, 8 or 12 h were harvested by a 250 g centrifugation step after trypsinization. Apoptotic ratios of the cells were detected using the Annexin V-FITC Apoptosis Detection Kit (BioVision, USA) according to the manufacturer's protocol. Briefly, the harvested cell pellets were washed three times with PBS and then resuspended in annexin-binding buffer. The cell suspensions were treated with both annexin V and PI dyes for 15 min at room temperature. The stained cells were detected using a flow cytometer (FC500 MCL, Beckman) to distinguish the cells in early apoptosis (annexin V+/PI−) from those in post-apoptosis/necrosis (annexin V+/PI+). To study the effects of p53, wild-type J774A.1 cells were pre-treated with 30 μM p53 inhibitor pifithrin-α (Sigma) or 100 μM pan-caspase inhibitor Z-VAD-FMK (Sigma) or both 30 μM pifithrin-α and 100 μM Z-VAD-FMK (Martinet et al., 2006), while p53-depleted J774A.1 cells or wild-type THP-1 cells were pre-treated with 100 μM Z-VAD-FMK, and the subsequent steps for cell apoptosis detection were performed the same as described above.
Real-time quantitative RT-PCR
THP-1 and wild-type or p53-depleted or p53-inhibited (by pifithrin-α) J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4, 8, 12 or 24 h were added with TRIzol reagent (Invitrogen, USA) to extract the total cellular RNAs according to the manufacturer's protocol. cDNAs from the RNAs were synthesized by reverse transcription (RT) using a PrimeScript RT reagent Kit (TaKaRa, Japan). Using the cDNAs as the templates, p53, akt, p21, p27, bax, bcl-2, noxa, puma, fas and fasL mRNAs were measured by real-time quantitative PCR (qPCR) using a SYBR Premix Ex-TaqTM II Kit (TaKaRa) in an ABI-7500 Real-Time PCR System (ABI, USA). The primers used in qPCR from the RTPrimerDB database or designed using Primer3 software are shown in Table 1 (Lefever et al., 2009). In the RT-qPCR, the β-actin-encoding gene was used as the internal reference. The obtained RT-qPCR data were analysed using both the ΔΔCt model and the randomization test in REST 2005 software (Pfaffl et al., 2002).
Table 1. Primers used in the real-time quantitative PCRs
Sequence (5′ to 3′)
F, forward primer; R, reverse primer.
Western blot assay
THP-1 and wild-type or p53-depleted or p53-blocked (by pifithrin-α) J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4, 8, 12 or 24 h were harvested by a 250 g centrifugation after trypsinization and then lysed with RIPA buffer (0.15 M NaCl, 1% NP-40, 0.1% SDS, 1 mM DTT and 1 mM PMSF in 50 mM Tris-HCl buffer, pH 7.5) for 5 min on ice. Insoluble substances were removed by a 5 min centrifugation step at 17 200 g at 4°C. The protein concentration in each of the supernatant specimens was estimated using a BCA protein assay kit (Thermo Scientific). The specimens with equivalent protein amounts were spread by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and then electron-transferred onto polyvinylidene difluoride (PVDF) membrane (Millipore, USA). Using rabbit anti-mouse or human-p53-, Akt-, p21-, p27-, Bax-, Bcl-2-, Noxa- or Puma-IgG (Cell Signaling Technology) as the primary antibody, and HRP-conjugated goat anti-rabbit-IgG as the secondary antibody (Jackson ImmunoResearch), several Western blot assays were performed to detect the expression of p53, Akt, p21, p27, Bax, Bcl-2, Noxa and Puma proteins in the specimens. The immunoblotting signals reflecting protein expression levels were quantified by densitometry (grey scale determination) using an Image Analyser (Bio-Rad, USA). In the assay, the normal rabbit IgG (Sigma) instead of the primary antibody mentioned above and β-actin were used as the controls.
Detection of p53 and Akt phosphorylation
The soluble protein specimens of THP-1 and wild-type or p53-depleted or p53-blocked (by pifithrin-α) J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 0.5, 1, 2, 4, 8, 12 or 24 h were prepared using RIPA buffer and centrifugation, and then the protein concentration in each of the specimens was determined as described above. After SDS-PAGE and electro-transferring onto PVDF membrane, several Western blot assays were performed to detect the phosphorylation of p53 and Akt in the specimens using a Phospho-p53 Antibody Sampler Kit (Cell Signaling Technology) and a Phospho-Akt Pathway Antibody Sample Kit (Cell Signaling Technology). The immunoblotting signals reflecting protein phosphorylation were quantified by densitometry (grey scale determination) using an Image Analyser (Bio-Rad).
Detection of AIF, EndoG, HtrA2 and Smac release from mitochondria
Fractions of mitochondria and cytosol from the THP-1 and wild-type or p53-depleted or p53-blocked (by pifithrin-α) J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 1, 2, 4, 8 or 12 h were separated using a Mitochondria/Cytosol Fractionation Kit (BioVision, USA) according to the manufacturer's protocol. The protein concentration in each of the fractions was estimated using a BCA protein assay kit (Thermo Scientific). The fractions with equivalent protein amounts were subjected to SDS-PAGE in 12% gel and then electro-transferred onto PVDF membrane (Millipore). Using rabbit anti-mouse or human-AIF-, EndoG-, HtrA2- or Smac-IgG as the primary antibody (AbCam), and HRP-conjugated goat anti-rabbit-IgG as the secondary antibody (Jackson ImmunoResearch), several Western blot assays were performed to detect AIF, EndoG, HtrA2 and Smac proteins in the mitochondrial or cytosol fractions, in which β-actin was used as the inner reference. The immunoblotting signals were quantified by densitometry (grey scale determination) using an Image Analysor (Bio-Rad).
Detection of expression and translocation of Fas and Fas ligand
THP-1 and wild-type or p53-depleted or p53-blocked (by pifithrin-α) J774A.1 cells infected with L. interrogans strain Lai at an moi of 100 for 2, 4, 8, 12 or 24 h were harvested by a 250 g centrifugation after trypsinization. After washing with PBS, the cells were fixed with 4% paraformaldehyde at 4°C overnight. The cells were washed with PBS, and then treated with 0.1% saponin (Sigma) in PBS for 10 min at room temperature for permeabilization (Monari et al., 2005). After washing with PBS again, the cells were stained with phycoerythrin (PE) anti-mouse or human-Fas, or PE anti-mouse-FasL or human-FasL antibody (eBioscience, USA) for 15 min at room temperature in the dark. Total PE-binding Fas or FasL in the cells was measured by flow cytometry as described above.
Data from a minimum of at least three independent experiments were averaged and presented as mean ± SD (standard deviation). One-way analysis of variance (anova) followed by Dunnett's multiple comparisons test were used to determine significant differences. Statistical significance was defined as P ≤ 0.05.
This work was supported by grants (No.: 81171534 and 81261160321) from the National Natural Science Foundation of China and a grant (No.: 2010ZZ09) from the National Key Laboratory for Diagnosis and Treatment of Infectious Diseases of China.