|Home | About | Journals | Submit | Contact Us | Français|
The mechanisms by which pathogens evade elimination without affecting host fitness are not well understood. For the pathogen Pseudomonas aeruginosa, this evasion appears to be triggered by excretion of the quorum sensing (QS) molecule 2-aminoacetophenone (2-AA), which dampens host immune responses and modulates host metabolism, thereby enabling the bacteria to persist at a high burden level. Here, we examined how 2-AA trains host tissues to become tolerant to a high bacterial burden, without compromising host fitness. We found that 2-AA regulates histone deacetylase1 (HDAC1) expression and activity, resulting in hypoacetylation of lysine 18 of histone H3 (H3K18) at pro-inflammatory cytokine loci. Specifically, 2-AA induced reprogramming of immune cells occurs via alterations in histone acetylation of immune cytokines in vivo and in vitro. This host epigenetic reprograming, which was maintained for up to 7 days, dampened host responses to subsequent exposure to 2-AA or other pathogen-associated molecules. The process was found to involve a distinct molecular mechanism of host chromatin regulation. Inhibition of HDAC1 prevented the immunomodulatory effects of 2-AA. These observations provide the first mechanistic example of a QS molecule regulating a host epigenome to enable tolerance of infection. These insights have enormous potential for developing preventive treatments against bacterial infections.
Pathogens have evolved mechanisms for evading eradication by their hosts’ dynamic immune systems1,2. The establishment and maintenance of bacterial infections rely on the pathogen's ability to breach the host's innate immune responses3. Immune-driven resistance is the prevailing defense strategy against infection. However, organisms can also defend themselves through a different defense strategy, namely tolerance, defined as the ability to limit pathogen damage without controlling pathogen burden4,5. Tolerance is an evolutionarily conserved defense strategy shared by diverse plants and animals, including humans6. In tolerance (or resilience), a host copes with a pathogenic encounter without a reduction in fitness4,6,7 and avoids harmful inflammatory responses that can occur with the immune-driven resistance strategy8. The resistance and tolerance defense strategies have different ecological and evolutionary consequences: resistance reduces pathogen presence, whereas tolerance allows pathogen prevalence. Our understanding of the biological mechanisms mediating the mutual pathogen-host adaptation and the causes and consequences of variation in host tolerance is limited.
Traditionally, immunological memory has been recognized as an attribute of the adaptive immune system, while the innate immune system has been regarded as perpetually naïve. However, the ability to respond more (or less) vigorously to a second pathogen encounter has been documented in organisms lacking an adaptive immune system (i.e. without T and B cells) and attributed to innate immune “training”9,10. Innate immune training is associated with epigenetic and metabolic reprogramming of innate immune cells11,12. However, the mechanisms by which such training can be triggered and its functional outcomes for the host and pathogen have not been resolved.
The small volatile bacterial quorum sensing (QS)-regulated molecule 2-aminoacetophenone (2-AA) acts as an immunomodulatory signal that enables the host to tolerate long-term bacterial presence13. Its impact on host metabolism likely also contributes to host tolerance14. In mice, 2-AA treatment improves host survival dramatically during Pseudomonas aeruginosa (PA) infection (90% vs. 10%), despite increases in bacterial load13,15. Hence, 2-AA appears to act as a critical mediator of host tolerance to pathogen burden. 2-AA is excreted liberally by PA and other pathogens16-19 in infected tissues16,20 and its synthesis is regulated by QS, a communication system governed by population density that is used by bacteria21 to regulate virulence factors22-24.
Our prior findings showing that 2-AA promotes bacterial phenotypes that favor a persistent bacterial presence15,25 converge with emerging evidence suggesting that pathogens can repress host immunity through epigenetic reprogramming9,11,26. Epigenetic modifications are important mechanisms of gene regulation. Notably, mRNA transcription initiation and elongation are controlled by various posttranslational modifications of chromatin-associated histones, including acetylation and methylation of lysine residues on histone tails27. Bacteria can manipulate host immunity through effects on the enzymes that control epigenetic histone acetylation, namely histone acetyltransferases (HATs) and histone deacetylases (HDACs)28. Here, we examined 2-AA regulation of host cell epigenetic programs and the effects of this regulation on host inflammatory responses and bacterial burden tolerance. We found that 2-AA acts as a host training molecule by reprogramming innate immune cells via HDAC activity and alterations in histone 3 acetylation on lysine 18 (H3K18ac). This new understanding of mechanisms of host tolerance and training may provide avenues for the development of novel therapeutic interventions against bacterial infections.
An initial exposure to 2-AA led to activation of the nuclear factor (NF)-κB pathway in THP-1 human monocytes, as evidenced by increased reporter activity (Fig. 1b and Supplementary Fig. 1a), as well as increased mRNA transcript and secreted protein levels of the pro-inflammatory cytokines tumor necrosis factor (TNF)-α and interleukin (IL)-1β, as well as of monocyte chemotactic protein (MCP)-1 (Fig. 1c–h, Supplementary Fig. 1b–g). However, pretreatment of THP1 cells with 2-AA, followed by washing and 2-AA re-exposure (Fig. 1a), reduced NF-κB reporter activity (Fig. 1b) and attenuated transcription (Fig. 1c, d and e) and protein secretion (Fig. 1f, g and h) of TNF-α, IL-1β, and MCP-1. Because synthetic, endotoxin-free 2-AA was used (Supplementary Fig. 1h), the observed effects can be attributed to 2-AA rather than endotoxin-dependent tolerance. The immunomodulatory effects of 2-AA were recapitulated in primary human (Fig. 1i and j; Supplementary Fig. 2a–f) and mouse macrophages (Supplementary Fig. 2g–l), in which TNF-α, MCP-1, and IL-1β secretion were reduced upon 2-AA stimulation when cells were stimulated 6-h, 1 day, 3 days, or 7 days after 2-AA pretreatment (Fig. 1i, j and Supplementary Fig. 2a-l). These immunomodulatory effects also affected cells’ subsequent responses to unrelated pathogen-associated molecular patterns. That is, compared to non-pretreated cells, 2-AA pretreated macrophages produced reduced levels of pro-inflammatory cytokines following lipopolysaccharide (LPS) or peptiodoglycan (PGN) stimulation (Fig. 1i and j).
Immunoblot analysis with a pan-acetyl antibody revealed that acetylation of the core histone H3 peaked 3 to 6-h after 2-AA exposure and then declined (Supplementary Fig. 3a), a pattern reminiscent of 2-AA-induced pro-inflammatory cytokine expression (Fig. 1c–h). Analysis of acetylation levels of purified histones from 2-AA stimulated cells with an H3 modification multiplex assay demonstrated preferential enhancement of the H3K18ac histone mark, but unaltered levels of H3K9ac, H3K9me3, H3K14ac, or H3K56ac (Supplementary Fig. 3b). Quantitation of global acetylation of H3K18 and H3K9 showed significantly elevated levels of H3K18ac, but not H3K9ac, 1-h after first-time 2-AA exposure in human monocytes (Supplementary Fig. 3c-d) or murine macrophages (Fig. 2a and 2b). Addition of the HDAC inhibitor trichostatin (TSA) restored H3K18 global acetylation levels in 2-AA pretreated cells as seen in the non-pretreated cells following 2-AA stimulation (Supplementary Fig. 3c).
Enrichment of H3K18ac, but not H3K9ac or H3K9me3, specifically at the promoter locus of the gene that encodes the NF-κB–targeted cytokine TNF-α, was attenuated in 2-AA pretreatment in mouse macrophages (Fig. 2c, d, e, and f) and human monocytes (Supplementary Fig. 3e), compared to that in non-pretreated control cells. RNA polymerase II occupancy, consistent with attenuation of transcription, was also higher at the TNF-α promoter in stimulated cells without 2-AA pretreatment than in those with pretreatment (Supplementary Fig. 3f).
Induction of HAT activity was observed following 2-AA stimulation in human monocytes (Fig. 3a and Supplementary Fig. 4a) and mouse macrophages (Supplementary Fig. 4c and e), and this induction was reduced in 2-AA pretreated groups. Meanwhile, 2-AA pretreatment amplified 2-AA stimulation effects on HDAC activity (Fig. 3b, Supplementary Fig. 4b, d, and f). Hence, an initial exposure of 2-AA induced HAT activity whereas 2-AA re-exposure reduced HAT activity and resulted in increased and sustained HDAC activity, which suppresses pro-inflammatory gene expression.
To identify which HDACs repress pro-inflammatory gene expression in 2-AA pretreated cells, we employed a number of inhibitors with differing HDAC specificities. As shown in Figure 3b, 2-AA-induced HDAC activity in 2-AA pretreated cells could be blocked with TSA (inhibitor of Class I/II HDACs, i.e. HDAC 1-3 and 8/4, 5, 6, 7, 9, and 10), valproic acid (VPA; inhibitor of all Class I/II HDACs except HDAC 6 and 10)29,30, and Ms275 (preferential inhibitor of HDAC1)31, but not with TMP269 (inhibitor of Class IIa HDACs 5, 7, and 9)32, RGFP966 (HDAC3 inhibitor)33, PCI-34051 (HDAC8 inhibitor)34, HPOB (HDAC6 inhibitor)35, or nicotinamide (NIC; inhibitor of Class III HDACs, i.e. SIRT1-7)36 (Supplementary Fig. 4f and 5). This selectivity profile points to the involvement of HDAC1 (Class I HDAC) in 2-AA promoted deacetylation.
HDAC1 protein was highly accumulated in the nuclear fraction of 2-AA pretreated cells, whereas HDAC2 levels were unaltered relative to levels in non-pretreated controls (Fig. 3c and Supplementary Fig. 6a-e). HDAC1 enrichment was also increased at TNF-α promoters in the 2-AA pretreated cells compared to that in control cells (Fig. 3d and e, Supplementary Fig. 6f and g).
HDAC1 KD in murine macrophages by shRNA (Supplementary Fig. 6h) restored H3K18 acetylation (Fig. 4a and Supplementary Fig. 6i), TNF-α secretion (Fig. 4b), and NF-κB reporter activation (Fig. 4c) in 2-AA pretreated macrophages compared to untreated cells following 2-AA stimulation. Taken together, these data indicate that HDAC1 is involved in 2-AA induced immune tolerance training and regulation of H3K18 acetylation.
The Class I/II HDAC inhibitor TSA rescued H3K18ac levels (Supplementary Fig. 3c) and TNF-α secretion (Fig. 4b, 4d, Supplementary Fig. 7a) in 2-AA pretreated cells following 2-AA stimulation, but not in HDAC1 KD cells (Fig. 4b). Ms275 (preferential HDAC1 inhibitor) and VPA (Class I/II HDAC inhibitor) also restored TNF-α production in 2-AA pretreated THP1 cells (Fig. 4d and Supplementary Fig. 7a). Conversely, agents that inhibit HDACs other than HDAC1 (i.e. NIC, TMP269, RGFP966, PCI-34051, and HPOB) did not rescue TNF-α secretion in 2-AA pretreated cells (Supplementary Fig. 7 j). TNF-α production in 2-AA pretreated cells was also unaffected by paragyline (histone demethylase inhibitor), 5'azacytidine (DNA methyltransferase inhibitor), and anacardic acid (HAT inhibitor) (Supplementary Fig. 7k). Taken together, we show that reversal of the immunomodulatory epigenetic effects of 2-AA pretreatment is achieved by HDAC1 inhibition.
Daily administration of the Class I/II HDAC inhibitor TSA extinguished the survival outcome benefit of 2-AA pretreatment in burn-PA infection (BI) model mice dramatically (p < 0.0001; Fig. 5a). Furthermore, the TSA treatments reduced the survival of BI animals (12%) to a level even lower than that of untreated BI animals (56%)(p < 0.044; Fig. 5a). The TSA treatment decreased bacterial burden (p = 0.031 vs. no TSA) in 2AA-pretreated BI animals (Fig. 5b); it had a weaker (non significant trend) effect on bacterial load in BI animals without 2-AA pretreatment (p = 0.059 vs. no TSA).
Biochemical analysis showed that 2-AA pretreatment of BI animals reduced H3K18 acetylation and HAT activity (Fig. 5c and d), while enhancing HDAC activity (Fig. 5e) and dampening pathogen-induced cytokine production (Fig. 5f and g). Even without exogenous 2-AA, in BI mice, H3K18 acetylation, HAT activity, and pro-inflammatory cytokine secretion fell gradually over time, while HDAC activity was increased and sustained (Fig. 5c-g). TSA administration restored global acetylation levels of H3K18ac (Fig. 5c) and HAT activity in the spleen (Fig. 5d), dampened HDAC activity (Fig. 5e), and increased serum levels of TNF-α and MCP-1 (Fig. 5f and g) in 2-AA pretreated BI animals. Hence, TSA treatment increased PA infection susceptibility while disrupting 2-AA mediated immunosuppression in vivo.
Here, we uncovered mechanisms by which a pathogen-excreted molecule leads to mutual pathogen-host adaptation wherein the host immune response is trained to tolerate a high pathogen burden. We found that PA, which infects plants37,38, insects39, nematodes40, mice38 and humans41, promotes host-tolerance training through epigenetic reprogramming via the QS molecule 2-AA. Although there are several examples of immunity repression via infection-induced host epigenetic reprogramming26,28,42,43 and small molecules favoring acute infection have been identified22, no pathogen-derived small molecule to our knowledge has been shown previously to alter immune responses toward tolerance to pathogen burden without compromising host survival.
We report four principal results demonstrating the effects of 2-AA on host epigenetic reprogramming (Fig. 5h). Firstly, 2-AA promoted long-lasting effects that impacted inflammatory responses non-specifically. Secondly, reduction of histone acetylation in 2-AA pretreated cells attenuated transcription of pro-inflammatory cytokines in vitro and in vivo. Thirdly, HDAC1 expression and activity were elevated in 2-AA pretreated cells. Finally, HDAC1 inhibition counteracted 2-AA's immunomodulatory effects in vitro and in vivo. These results show that 2-AA triggers epigenetic modulation that switches innate immune cells from a transcriptionally permissive to a more silent chromatin state that in turn promotes pathogen survival while augmenting host tolerance.
Previous cell culture studies have pointed to histone deacetylation as a mechanism of host immunomodulation during infection with other pathogens28,42,44,45. However, the in vivo relevance of these studies for both host and pathogen, as well as the bacterial component mediating the effect, were unknown. Consistent in principle with our findings, Eskandarian and colleagues found that Listeria monocytogenes infection—which under normal circumstances triggers H3K18 deacetylation—was reduced in mice when a critical deacetylase was knocked out45.
Our findings suggest that 2-AA trains innate immune cells to limit pathogen-induced inflammation via histone deacetylation. This phenomenon differs from prior reports of so-called immunity/memory training, such as that seen with Bacillus Calmette-Guerin vaccination, Candida albicans infection, or fungal cell wall component exposure, which induce histone acetylation and provide non-specific protection through an augmentation of pro-inflammatory responses and clearance9,10,46. On the contrary, 2-AA dampens inflammatory responses and allows maintenance of a high bacterial burden, adaptations that benefit the pathogen as well as the host. Interestingly, it has been reported that nasal treatment with live attenuated Bordetella pertussis offers long-term protection against influenza viruses by limiting virus-induced inflammation in an antigen-independent manner, without affecting viral load47.
The mechanism mediated by 2-AA is also distinct from previously described effects of other bacterial molecules that dampen inflammatory responses while promoting bacterial clearance48,49. That is, 2-AA triggers immunomodulation that does not promote clearance. Although both dampen immune responses, the infection outcomes are opposite. The particular HDACs involved also differ between 2-AA and LPS effects50.
Based on our results, we propose a model (Fig. 5h) in which acute 2-AA stimulation results in increased HAT activity, resulting in histone acetylation at pro-inflammatory cytokine promoter loci that potentiate transcription, whereas prolonged exposure to 2-AA induces deacetylation by HDAC1, leading to a loss of active transcription, thus promoting immunosuppression. The CCAAT/enhancer-binding protein (C/EBP)β is activated following 2-AA treatment in mouse macrophages, which inhibits transcription13, and histone deacetylation is initiated by HDAC recruitment through C/EBPβ at specific cytokine promoter sites during persistent mycobacterial infection28. Our results suggest that the presently observed host tolerance to infection is a result of sustained deacetylation leading to long-lasting epigenetic alterations that favor long-term bacterial presence and host tolerance of a bacterial burden.
Collectively, our results show that the QS derived molecule, 2-AA, modulates both the source pathogen and its host. In the pathogen, 2-AA permits survival and may promote persistence via the silencing of acute virulence functions15,25. In the host, 2-AA triggers immunomodulation that prevents an overly robust inflammatory response, thus enabling the host and pathogen to coexist. Based on our findings, we propose that 2-AA is a critical mediator of the intricate interplay between host and pathogen. Given that QS molecules are widely conserved across bacterial genera and the common occurrence of polymicrobial settings, it is important to note that 2-AA impacts other Gram-negative microbes25. Elucidation of patho-epigenetic changes is critical for understanding host tolerance to commensal microbes and may guide the development of treatments that may improve resilience to pathogen damage in patients with serious infections.
The animal protocol was approved by Institutional Animal-Care and Use Committee (IACUC) of Massachusetts General Hospital (Permit: 2006N000093). No randomization or exclusion of data points was used.
All cells were maintained in 5% CO2 at 37°C. THP-1 Blue cells (human monocytic leukemia line with an NF-κB–inducible reporter (Invivogen) and RAW264.7 cells carrying the NF-κB luciferase plasmid (IMGENEX, USA) were maintained in RPMI 1640 medium (Invitrogen) and Iscove's modified Dulbecco's medium (IMDM, Invitrogen), respectively. The media were supplemented with 10% heat-inactivated FBS (endotoxin-free Certified FBS; Invitrogen), 2% penicillin/streptomycin, 2 mM L-glutamine, and 10 mM HEPES (all from Gibco). The cells were seeded in T-75 tissue culture flasks (Falcon, USA) and used between passages 2 and 3.The THP-1 and RAW264.7 cell lines were extensively validated by the Invivogen and IMAGENEX respectively. For quality control, we have tested the cells for mycoplasma with PlasmoTest™ kit (Invivogen).
Peripheral blood mononuclear cells (PBMCs) were separated from buffy coats of healthy donors (from MGH blood components lab) by Lymphoprep (Stemcell tech) density gradient centrifugation and purified by adherence 51. The cells were allowed to differentiate in complete XVIVO medium containing 100ng/mL human M-CSF (Peprotec) for 7 days at 37°C, 5% CO2. On day 4, cultures were supplemented with one volume of complete XVIVO medium containing 100ng/mL human M-CSF. Macrophages were seeded onto 24 well (1.5 × 105 cells/well) tissue culture plates for the treatments.
Bone marrow was flushed from mouse tibias and femurs and cells (previously spun at 1500 rpm for 10min to remove debris) seeded at a density of 50 × 106 cells (yield from 1 mouse)/non treated 10cm dish with complete DMEM/F12 (Gibco) medium containing 5 ng/mL of IL-1 (Peprotec) and 5ng/mL CSF-1 (Peprotec) and incubated at 37°C, 5% CO2. After 24-h, non-adherent cells were seeded into 24-well (1.5 × 105 cells/well) tissue culture dish with fresh complete medium containing 5 ng/ml of IL-1 and 5ng/mL CSF-1.Cells were matured over 5 days.
Chemically synthesized 2-AA was purchased form Sigma-Aldrich (titration by HClO4, 98.0–102.0%; purity (GC) > 98.0%). The exact amount of 2-AA in human tissues has not been quantified due the compound's high volatility. The basis for selecting the 2-AA concentrations in our studies is based on the assessment of 2-AA levels in bacterial cultures of several clinical Pseudomonas aeruginosa isolates from burned patients. LC/MS found that 2-AA levels were in the range of 37-225 μM. The concentration used in our murine studies, namely 50 μM (6.75 mg/kg), is at the low end of this range, and thus should be considered a reasonable approximation of “physiological” levels. Because 2-AA is a highly volatile molecule and tolerization assays of human monocytes and murine macrophages run for ~ 24-h in microtiters (prone to evaporation), we use 1-2 (200–400 μM) and 1.8-3.5 (400–800 μM) orders of magnitude higher concentrations in our culture studies than that measured in bacterial cultures, respectively.
2-AA was serially diluted in PBS. For positive control, LPS was used. The endotoxin unit was measured using the Limulus Amebocyte Lysate (LAL) technique with LAL chromogenic Endotoxin Quantitation Kit (Thermo-scientific) according to the manufacturer's instructions.
THP-1 cells were plated at a density of 105/mL in 24 well plates and grown overnight at 37 °C in 5% CO2. Cells in the treatment groups were pretreated with 200 μM 2-AA for 24-h; treated and non-treated cells were washed with 1X phosphate buffered saline (PBS) and kept in fresh medium. Cells were stimulated with 200 μM 2-AA for the durations indicated in the figures. Similarly, murine macrophage RAW264.7 cells were pretreated (or not) with 800 μM 2-AA for 48-h then stimulated with 400 μM 2-AA.
For memory experiments, human and mouse primary macrophages were used. Human macrophages (105/mL) were pretreated with 400 μM 2-AA for 24-h; treated and non-treated cells were washed out and after 1 day, 3 days or 7 days, the cells were stimulated with 200 μM 2-AA for 6-h. Mouse primary macrophages (105/mL) were pretreated with 800 μM 2-AA for 48-h; treated and non-treated cells were washed out and after 1 day, 3 days or 7 days, the cells were stimulated with 400 μM 2-AA for 6-h.
THP-1 Blue cells express a secreted embryonic alkaline phosphatase (SEAP) under the control of NF-κB transcription factors. To quantify secreted SEAP, the culture supernatant was incubated with QUANTI-Blue colorimetric assay reagent (Invivogen) according to the manufacturer's instructions. All assays were run in triplicate.
After treatment, Raw 264.7 cells were washed with PBS and then lysed in the buffer provided with the Luciferase Assay Kit (Promega) as described earlier 13.
Total RNA from approximately 1.2 × 106 cells was extracted with the RNeasy Mini Kit (Qiagen). cDNAs were prepared with the RETROscript™ Kit (Ambion® Life Technologies), according to the manufacturer's protocol. Real-time PCR was performed using the Brilliant II SYBR green super mix (Agilent) and primer sets specific for human or murine TNF-α, IL-β, MCP-1, and GAPDH and β-actin (Supplementary Table 1). Expression of cytokines was normalized to GAPDH or β-actin with the ΔCt Method and relative expression was calculated with non-pretreated and unstimulated cells as references. The assay was conducted in triplicate; means and standard deviations were calculated for each group.
Histones were isolated from cells and tissues as described previously52. To study the histone modifications, 5 μg of isolated histones were analyzed by immunoblotting using acetyl-pan-H3 (cat no: 06-599), acetyl-pan-H4 (cat no: 04-557) (Millipore), acetyl-Histone H3 (Lys18) (cat no: 9675) and H3 (cat no: 4499) (Cell Signaling Technology).
Quantification of global level of lysinie-specific histone acetylation was done using Global Acetyl Histone H3-K9/K18 colorimetric assay kits (Abnova) according to manufacturer's protocol.
Nuclear extracts were obtained from cells at experimentally indicated time points with a Nuclear Extract kit (Active Motif, Carlsbad, CA) as described earlier 13.
The binding specificity of purified histone to modified H3 residues was analyzed by using Histone H3 Modification Multiplex kit (Epigentek) according to manufacturer's protocol.
HAT activity was measured in nuclear lysates using a non-radioactive HAT fluorescent assay kit (Active Motif), according to the manufacturer's instructions. The fluorescence intensity was measured using Tecan plate reader (excitation wave length 360-390nm, emission wavelength 450-470nm). HAT activity was expressed as arbitrary fluorescent units (AFU).
Nuclear fraction were used to measure deacetylase activity with a non-isotopic assay that used a fluorescent derivative of epsilon-acetyl lysine (HDAC Fluorescent Activity Assay Kit, Active Motif), according to the manufacturer's instructions. The fluorescence intensity was measured using Tecan plate reader (excitation wave length 340-360nm, emission wavelength 440-460nm). HDAC activity was expressed in arbitrary fluorescent units (AFU).
The protein content of HDAC1 and HDAC2 was measured in nuclear lysate using colorimetric EpiQuick HDAC1 and HDAC2 assay kits (Epigentek) according to the manufacturer's protocol. Absorbance was read at 450 nm. Results were calculated using a standard curve and expressed as nanogram/mg of protein. The assay was conducted in triplicate.
Cells were washed in 1X PBS, cross-linked in 1% methanol-free formaldehyde for 10 min, then glycine was added to a final concentration of 0.125M for 5min at RT. Using the truChIP™ High Cell Chromatin Shearing kit (Covaris), cells were prepared for sonication according to the manufacturer's protocol. Approximately 1×107 cells were placed in a 12×12-mm tube and subjected to genomic DNA shearing with the Covaris S220 sonicator for 8min (140 peak power, 5 duty factor, and 200 cycles/burst). ChIP was conducted with the Magna ChIP™ A/G Kit (Millipore) according to the manufacturer's protocol. Briefly, chromatin from approximately 1.2×106 cells was incubated overnight at 4°C with 4μg of anti-acetyl H3K18, acetyl H3K9, H3K9me3, HDAC1 (Abcam) or anti-RNA polymerase II phospho S CTD (Millipore) ChIP-grade antibodies and 20 μl of A/G magnetic beads. The beads were washed serially (5 min each) with low-salt wash buffer, high-salt wash buffer, LiCl wash buffer, and TE buffer from the kit at 4°C. Chromatin was eluted with elution buffer containing Proteinase K at 62°C for 4-h, then incubated at 95°C for 10min. DNA was isolated by column purification. Real-time PCR was performed with the Brilliant II SYBR green super mix (Agilent) and primer sets to amplify various regions of TNF-α and β-actin loci (Supplementary Table 1). Normalized values were calculated by the percent-input method. β-actin percent input normalized values are shown. The assay was conducted in triplicate; means are reported with standard deviations.
HDAC1 (sc-29344-V) and non-targeting control (sc-108080) shRNA lentivirus clones were purchased from Santa Cruz Biotechnology. Murine macrophage Raw264.7 cells were infected with lentiviral shRNA. After selection with puromycin, a pool of infected cells was expanded and efficacy of HDAC1 KD was validated by Western blot analyses.
Cellular extracts were prepared in RIPA buffer and analyzed by standard immunoblot techniques. The primary antibodies used were specific for HDAC-1(cat no.5356) (Cell Signaling Technology), and mouse anti-β-actin (cat no: sc-47778) (Santa Cruz Biotechnology). 20μg of proteins were subjected to immunoblot analysis, fractionized by polyacrylamide gel electrophoresis (PAGE) and transferred to membrane. The bands were detected by SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) reaction, according to the manufacturer's instructions.
TNF-α, IL-β, and MCP-1 protein levels in culture supernatants of 2-AA pre-treated and non pretreated cells were measured by ELISA using the Quantikine human/mouse TNF-α, IL-β, and MCP-1 kits (R & D Systems) according to the manufacturer's instructions. Mouse serum was isolated and cytokine secretion assayed by ELISA using the Quantikine TNF-α and MCP-1 kit (R & D Systems) according to the manufacturer's instructions.
For the HDAC inhibition assay, cells were pretreated with TSA (5nM, Sigma-Aldrich). For the TNF-α secretion assay, human THP1 or murine macrophage RAW264.7 cells were treated with TSA, VPA (0.1 μM), Ms275 (5 μM), NIC (0.5 μM) HPOB (0.5 μM) (Sigma-Aldrich), TMP269 (0.1 μM), PCI-34051 (0.1 μM), RGFP966 (0.5 μM) (Selleckchem), paragyline (3 μM), 5'azacytidine (10 μM) or anacardic acid (10 μM) (Cayman Chemical). The concentration of HDAC inhibitors used did not induce cytotoxicity (Supplementary Fig.5 a-h). It is important to note that at high concentrations, HDAC inhibitors blocked TNF-α secretion (Supplementary Fig. 7b-i) 30.
The cytotoxicity of cells treated with TSA (Sigma-Aldrich), NIC, Ms275, HPOB (Sigma-Aldrich), TMP269, RGFP966, PCI-34051 (Selleckchem) or VPA (Cayman Chemical) was measured by MTT assay as described previously 13.
A P. aeruginosa strain known as RifR human clinical isolate UCBPP-PA14 (a.k.a. PA14) was used 37,38. They were grown at 37 °C in Luria-Bertani (LB) broth under shaking and aeration or on LB agar plates containing appropriate antibiotics. Over day PA14 cultures were grown in LB from single colony to O.D.600nm 1.5 and diluted 1:50,000,000 in fresh LB media and grown overnight to O.D.600nm 3.
A thermal injury mouse model 37 was used to assess 2-AA effects on survival and bacterial burden in 6-wk-old CD1 male mice (Charles River; Boston, MA). Four days prior to thermal injury and infection mice were injected intravenously (IV) with 2-AA (6.75 mg/kg). Following administration of anesthesia, a full-thickness thermal burn injury involving 5–8% of the total body surface area was produced on the shaved mouse abdomen dermis, and an inoculum of 1 × 104 PA14 cells in 100 μl of MgSO4 (10 mM) was injected intradermally into the burn eschar. Immediately after BI, a group of mice received intraperitoneally (IP) TSA (0.5 mg/kg,) once per day for up to 10days. Mice survival was assessed over the course of 11 days.
Because of the aggravated mortality rates as a result of TSA treatment daily for 10 days, CFU assessment for all groups was performed at 5 days instead of 10 days post-infection. CFU counts were assessed in muscle samples obtained from underneath the burn eschar and site of bacterial cell inoculation. Samples were homogenized in 1 mL of PBS, diluted and plated on LB-agar plates containing rifampicin (50 mg/L). For animal experiments, the sample size analysis was performed to ensure that our sample size was more than sufficient for necessary power of 0.8 using the PASS software, version 12 (NCSS, LLC, Kaysville, UT, USA). Investigators were not blinded to experimental conditions, and no randomization or exclusion of data points was used.
Spleen and blood samples were collected from all the above mice groups at 1 day, 5 days and 10 days post-BI for cytokines, histone acetylation and HAT/HDAC assays.
Wherever applicable, at least three independent experiments were performed, and the data were analyzed using the Student's t test or a one-way analysis of variance (ANOVA) with Tukey's HSD Post-hoc test, as appropriate. Animal data were analyzed using the Kaplan-Meier survivability test. Bacterial CFU counts were analyzed using the Kruskal-Wallis non-parametric test with Dunn's post-test. For all experiments P values < 0.05 were considered significant.
We thank Stuti Mehta and Megha Basavappa for their help with the isolation of human primary macrophages and mouse primary macrophages respectively. We extend a special thanks to Alicia Ballok for editing the manuscript. This work was supported by Shriners grant # 87100, 85200 and in part by NIH R33AI105902 and Cystic Fibrosis Foundation grant CFF#11P0.
AB, designed the study and did the experiments, performed data analysis and wrote the manuscript. AT, performed ChIP-PCR experiments and analysis of the data and wrote the manuscript. DM contributed to the in vivo experiments. KLJ supervised ChIP-PCR experiments. LGR designed the study and supervised the research and wrote the manuscript.