Widgets Magazine- ericyue.info
FJ Express Summaries

Hydrogen sulfide is a novel mediator of lipopolysaccharide-induced inflammation in the mouse

    Published Online:ericyue.info/10.1096/fj.04-3583fje


    The aim of this work was to investigate the part played by the novel gaseous mediator, hydrogen sulfide (H2S) in E. coli lipopolysaccharide-induced inflammation in the mouse and to examine plasma concentrations of this mediator in patients with septic shock.


    1. LPS-induced endotoxic shock is associated with increased plasma H2S concentration and tissue H2S tissue synthesizing activity

    LPS administration in animals has been widely used over many years by numerous researchers to induce a state of endotoxic shock which bears similarities to the clinical condition of septic shock in humans. Lung and liver myeloperoxidase (MPO) activity were measured as a marker of tissue neutrophil infiltration. Lung MPO activity was significantly (P<0.05) increased (148.2±2.6%, n=6) 6 h after injection of LPS (10 mg/kg). This increase was even more marked (218.3±4.9%, n=6) after a higher dose of LPS (20 mg/kg). In liver, a similar (but smaller) significant (P<0.05) rise in MPO activity was detected 6 h after administration of the same two doses of LPS (78.8±8.2% and 121.1±4.1% respectively, n=6). Lung sections from LPS-treated animals (Fig. 1 B) exhibited characteristic signs of inflammatory damage which included interstitial edema, alveolar thickening, and the presence of numerous leukocytes (lymphocytes and neutrophils) in both the interstitium and the alveoli. Liver (Fig. 1E ) and kidney (Fig. 1H ) sections from LPS-treated mice showed similar signs of damage while the histological appearance of all organs were essentially normal in saline-treated animals (Fig. 1A, D, G ).

    Figure 1.

    Figure 1. Lung, liver, and kidney hematoxylin and eosin stained sections from control (saline-injected), LPS injected (10 mg/kg, i.p., 6 h), and PAG-pretreated LPS-injected (50 mg/kg i.p., administered 30 min before LPS, 10 mg/kg, 6 h) animals. Results show representative sections from 6 animals in each group. Horizontal bar indicates 100 μm.

    LPS administration caused a time-dependent elevation of plasma H2S concentration. At a dose of 10 mg/kg of LPS, plasma H2S concentration was significantly (P<0.05) increased to 40.9 ± 0.6 μM and 65.3 ± 0.9 μM at 6 h and 24 h, respectively (c.f. 34.1±0.7 μM in saline-injected animals, n=6). The ability of liver and kidney homogenates prepared from LPS-injected animals to synthesize H2S from added L-cysteine was also assessed. When compared with saline-injected controls, liver (0.64±0.007 nmol/mg protein, n=6) and kidney (0.16±0.006 nmol/mg protein, n=6) from animals administered LPS (10 mg/kg) formed significantly (P<0.05) larger amounts of H2S. The enhancement in H2S biosynthesis was dependent upon time (after LPS injection) and dose of LPS injected. In separate experiments, administration of LPS significantly (P<0.05) elevated the plasma concentration of nitrate/nitrite (NOx) (e.g., 38.2±0.4 μM, 6 h after 10 mg/kg LPS, i.p., c.f. 15.2±0.50 μM, n=6, in controls).

    2. LPS increases tissue CSE expression

    CSE mRNA was clearly expressed in liver and kidney from control mice and a significant up-regulation (94.2±2.7% and 77.5±3.2% increase, respectively, n=6) of CSE mRNA was detected in liver and kidney removed from LPS-challenged (10 mg/kg, 6 h) mice.

    3. PAG reduces H2S formation and is anti-inflammatory

    In vitro, the CSE inhibitor (DL-propargylglycine, PAG) caused dose dependent inhibition of the conversion of L-cysteine to H2S by mouse liver homogenate (IC50, 93.0±1.9 μM, n=4). In vivo, PAG (50 mg/kg, i.p.) given to LPS-injected mice reduced plasma H2S (22.6±1.4 μM c.f. 40.9±0.6 μM in LPS-alone group, n=6, P<0.05). PAG pretreatment blocked H2S biosynthesis from added L-cysteine in liver and kidney homogenates determined ex vivo (Fig. 2 a, B). PAG pretreatment reduced the LPS-mediated rise in lung and liver MPO activity (Fig. 2C ) and examination of tissue sections suggested a lesser degree of interstitial leukocyte infiltration and tissue damage in lungs (Fig. 1C ) and to a lesser extent in liver and kidney (Fig. 1F, I ). PAG pretreatment significantly (∼36%) reduced the LPS-induced rise in plasma NOx concentration (38.2±0.4 μM c.f. 24.6±1.0 μM, n=6, P<0.05).

    Figure 2.

    Figure 2. Effect of PAG (50 mg/kg, i.p.) pretreatment on (A) liver and (B) kidney H2S-synthesizing activity (shown as nmol/mg protein) and (C) MPO activity shown as percentage increase over enzyme activity in controls in lung and liver in untreated, saline- and LPS- (10 mg/kg, i.p., killed 6 h thereafter) injected mice. Results are mean ± se, n= 6, *P < 0.05 (c.f. control) and +P < 0.05 (c.f. animals not administered PAG).

    4. Injection of the H2S donor NaHS causes inflammation

    To investigate a possible proinflammatory effect of H2S, we injected the H2S donor drug, sodium hydrosulfide (NaHS), into mice and evaluated its effect on lung and liver MPO activity, tissue structure, and plasma concentration of the cytokine TNF-α. There have been few attempts to evaluate the biological activity of NaHS in animals. However, a vasodilator effect of injected NaHS (14 μmol/kg, i.p.) in rats with pulmonary hypoxic hypertension has been reported. Accordingly, mice were injected with this dose of NaHS and killed 1 h later. A rise in plasma TNF-α level (4.5±1.4 ng/mL, c.f. 0 in controls, n=6) was observed. NaHS caused histological evidence of inflammatory damage in the lung but not liver as well as a significant (P<0.05) increase in MPO activity (c.f. saline-injected animals) in lung (303.7±26.0% increase, n=6) and liver (136.5±14.3% increase, n=6).

    5. Plasma H2S concentration is increased in humans with septic shock

    Plasma H2S concentration was measured in 5 patients suffering from septic shock and compared with 5 age- and sex-matched controls. Plasma H2S was significantly increased (approx. 3.4-fold) in septic shock compared with healthy controls (150.5±43.7 μM c.f. 43.8±5.1 μM, n=5, P<0.05).


    The present findings point to an important role for H2S in endotoxin-induced inflammation. There is little information in the literature concerning the possible role of H2S in this condition.

    In the present work, we show that LPS injection in mice caused significant inflammatory damage (measured biochemically and histologically) in lung and liver along with increased plasma H2S levels, liver and kidney H2S synthesizing activity, and CSE mRNA expression. We propose that LPS up-regulates tissue CSE expression which, in turn, is reflected in enhanced conversion of L-cysteine to H2S in tissue homogenates and consequently raised plasma levels of this mediator. The transduction mechanism(s) which underlie the control of CSE gene expression and enzyme activity, are poorly understood. Recently, a Sp1 binding consensus in the CSE core promoter region was identified which suggests that Sp1 may play a part in regulating CSE transcription. Further work is needed to investigate this possibility. However, it is interesting to note that NO modulates the DNA binding activity of Sp1.

    Pretreatment with PAG (CSE inhibitor) in animals administered LPS reduced leukocyte sequestration in lung and liver and provided at least partial protection against tissue inflammatory damage. It seems logical to assume that the improvement seen in such endotoxic animals following PAG administration is correlated with pharmacological inhibition of CSE-mediated H2S biosynthesis.

    Since inhibition of endogenous H2S biosynthesis offered some protection toward LPS-induced inflammatory damage, we determined whether H2S itself provoked an inflammatory reaction. Injection of NaHS (H2S donor) produced an inflammatory response characterized by lung and liver neutrophil infiltration (determined by measurement of MPO) and tissue damage to the lung.

    Both LPS and NaHS injection caused pronounced lung damage which, in the case of LPS, was reduced by pretreatment with PAG. We detected little H2S synthesizing activity in lung homogenates and the enzyme activity which was measurable was not increased after injection of LPS. Others have failed to detect CSE mRNA or protein in mouse lung. The cell/tissue source of H2S which is formed by CSE after LPS injection to cause lung damage is therefore unlikely to be the lungs. It is suggested, therefore, that neutrophils are “activated” by H2S in the bloodstream (or perhaps H2S formed in other organs such as the liver) and are then “trapped” in the dense pulmonary vasculature, thereby leading to signs of lung inflammatory disease and tissue destruction.

    The underlying cellular targets for H2S in LPS-mediated inflammation remain an important but unsettled question. Recent reports in the literature show that H2S activates ERK and MAPKp38 in human aortic vascular smooth muscle cells and in human embryonic cell (HEK293) cultures transfected with CSE cDNA. This may have relevance for the present study since LPS is known to trigger phosphorylation of MAPKp38 which leads to phosphorylation and activation of the MAPKAP-K2 (MK-2) pathway and phosphorylation of Hsp27 to cause the biosynthesis of TNF-α. It is therefore possible that overproduction of H2S activates the MAPKp38/MK-2/Hsp27 pathway, thereby promoting cytokine production which then exacerbates the resulting endotoxic shock. In our experiments we did detect large amounts of TNF-α in plasma after NaHS administration. Whether other proinflammatory cytokines are released under these circumstances remains to be seen.

    These results suggest that H2S plays a part in the inflammatory disease associated with endotoxic shock and highlight a potential new approach to the treatment of this condition. In this context, it is particularly interesting that plasma H2S concentration was elevated in patients suffering from septic shock. Although this part of the study is still preliminary in that it involves relatively few patients, the present data do point to a probable role for H2S septic shock in humans. Whether PAG (or a related CSE inhibitor) will ultimately prove useful in the clinic is not yet clear.

    To read the full text of this article, go to ; doi: 10.1096/fj.04-3583fje

    Figure 3.

    Figure 3. Schematic diagram of the role of H2S in LPS-induced inflammation.