LOOHs (lipid hydroperoxides) in oxLDL [oxidized LDL (low-density lipoprotein)] are potentially atherogenic compounds. Recently, H2S was identified as the third endogenous gasotransmitter in the vasculature. H2O2 is known to be destroyed by H2S. Assuming that H2S may also react with LOOHs, the results show that H2S can destroy LOOHs in oxLDL. The ability of LOOH-enriched LDL to induce HO-1 (haem oxygenase 1) in endothelial cells was abolished by H2S pretreatment. HPLC analysis showed that 9-HPODE [(9S)-hydroperoxy-(10E,12Z)-octadecadienoic acid], a compound found in oxLDL, was reduced to 9-HODE [(9S)-hydroxy-(10E,12Z)-octadecadienoic acid] in the presence of H2S. Thus H2S may act as an antiatherogenic agent by reducing LOOHs to the less reactive LOHs and could abrogate the pathobiological activity of oxLDL.
The oxidative modification of LDL (low-density lipoprotein) may play a role in atherogenesis [1–4]. The formation of the potentially atherogenic LOOHs (lipid hydroperoxides) in LDL is an early event in LDL lipid oxidation. Thus destroying LOOHs in LDL can limit its pathobiological potential.
Sodium hydrogen sulfide (NaHS), TPP (triphenylphosphine) and Xylenol Orange were from Sigma–Aldrich. H2O2 (30% solution) was supplied by Merck.
9-HPODE [(9S)-hydroperoxy-(10E,12Z)-octadecadienoic acid] and 9-HODE [(9S)-hydroxy-(10E,12Z)-octadecadienoic acid] were obtained from Cayman Chemicals (Ann Arbor, MI, U.S.A.) as ethanolic solutions (1 mg/ml).
NaHS was used to generate H2S in solution. H2S concentration was taken as 30% of the NaHS concentration according to Beauchamp et al. .
Prior to LDL oxidation, the lipoprotein was passed over a gel column equilibrated in PBS (pH 7.4). Routinely, LDL (0.2 mg/ml PBS) was incubated for 2 h in the presence of 10 μmol/l Cu2+ at 37 °C and the reaction was stopped by the addition of EDTA (100 μmol/l). Alternatively, LDL (1 mg/ml; 10 mmol/l Mops and 150 mmol/l NaCl, pH 7.4) was dialysed for 24 h at 4 °C against Mops buffer containing 5 μmol/l Cu2+ and oxidation was stopped by dialysis against PBS containing 100 μmol/l EDTA for 24 h at 4 °C by the method of Gerry et al. .
H2S treatment of oxLDL (oxidized LDL) samples
Cu2+ oxLDL was incubated at 37 °C in the absence or presence of NaHS for up to 1 h. For cell experiments, samples of Cu2+ oxLDL (0.2 mg/ml) were brought to 0.4 mg/ml by the addition of native LDL, and after incubation with or without NaHS for 30 min at 37 °C, the preparations were subjected to gel chromatography to get rid of NaHS.
Estimation of LDL oxidation
Estimation of lipid peroxides. LOOHs were estimated using the ferrous oxidation in FOX assay (Xylenol Orange assay) . A 50 μl portion of the sample was mixed with 450 μl of FOX reagent and after 30 min at room temperature (25 °C), absorbance was estimated at 560 nm. H2O2 was used to calculate concentrations. NaHS at the highest concentrations used did not interfere with the assay as no inhibition of colour development was observed when NaHS was added to FOX reagent/LOOH mixtures.
Diene formation. LDL oxidation was monitored as the increase in conjugated diene formation by measuring A234 .
Measurement of MDA (malondialdehyde). TBARS (thiobarbituric acid-reacting substance) formation in LDL was measured as reported in . NaHS did not interfere with TBARS formation as tested using pure MDA.
HUVEC (human umbilical-vein endothelial cells) were isolated and cultured as reported previously .
Isolation of total RNA, cDNA synthesis and quantitative PCR were performed as described in . Primers for real-time PCR were designed using the Primer3 software at http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi, and exon junction primers were selected to avoid amplification of residual DNA contaminations [HO-1 (haem oxygenase-1): 5′-AGACGGCTTCAAGCTGGTGA-3′ and 5′-TCTCCTTGTTGCGCTCAATCTC-3′; acidic ribosomal phosphoprotein P0: 5′-CTCGCCAGGCGTCCTCGT-3′ and 5′-TTGCCCATCAGCACCACAG-3′]. For normalization, mean Ct (threshold cycle values) of all replicate analyses were normalized to RPLP0 within each sample to obtain sample-specific ΔCt values (=Ct HMOX1 – Ct RPLP0). To compare the effect of various treatments with untreated controls, 2−ΔΔCt values were calculated to obtain fold expression levels, where ΔΔCt=(ΔCt treatment – ΔCt control). Real-time analysis using primers specific for cyclophilin b (GenBank® accession no. NM_000942) as the reference standard and normalization for equal amounts of total RNA input (1 μg for each sample) gave comparable results (results not shown).
HPLC separation of standard 9-HODE and 9-HPODE
Separations were preformed by injecting 5 μl of 64 μmol/l HPODE for time effect or 15 μl of 21.3 μmol/l HPODE for concentration-effect experiments into the HPLC (Waters 2695 Alliance system, Nova-Pak C18 3.9 mm×150 mm column, 4 μm) with a mobile phase of 0.3% H3PO4/acetonitrile (40:60, v/v) for 30 min with a flow of 1 ml/min. UV absorbance was monitored at 210 and 234 nm in dual wavelength mode. HPODE and HODE were prepared in PBS (chelexed). Substances were quantified via peak area.
HPLC analysis of oxLDL
Lipid oxidation products were determined by HPLC analysis of saponified lipids extracted from LDL by the method of Thomas and Jackson . In brief, 200 μg of LDL oxidized by the dialysis method (see above) was extracted with 2 ml diethyl ether/ethanol (3:1) and protein was precipitated by centrifugation. Decanted extracts were dried under nitrogen and subsequently lipid was dissolved in 0.5 ml of ethanol/0.05 ml of 10 M NaOH. After saponification (20 min at 60 °C) samples were neutralized and dried under nitrogen. Water was added and lipids were extracted into heptane. Dried heptane phases were dissolved in ethanol and subjected to HPLC analyses as outlined above.
The results are presented as mean values for two to five experiments. As appropriate, specific effects were evaluated by one-way ANOVA with post hoc testing using the Newman–Keul test. A value of P<0.05 was considered statistically significant.
Influence of H2S on LOOHs in oxLDL
As can be seen in Figure 1, when samples of oxLDL were incubated with NaHS for 60 min at 37 °C, a concentration-dependent decrease in LOOHs was observed. Approx. 50% (P<0.01) of LOOHs were destroyed by 100 μmol/l H2S treatment (Figure 1A). TPP (100 μmol/l), which is known to reduce LOOHs [19,20], was run as a control and showed comparable results (P<0.001). Figure 1(B) depicts the time-dependent influence of H2S on LOOHs. Under the conditions employed after 30 min of H2S treatment, approx. 25% of the LOOHs in oxLDL were destroyed (P<0.001 for all time points compared with control).
H2S treatment decreases LOOHs in oxLDL
In addition to the LOOH destroying effect, H2S could directly inhibit copper ion-induced LDL oxidation as monitored by diminished TBARS formation and increased lag phase during conjugated diene kinetics (results not shown).
H2S treatment of oxLDL decreases its ability to induce HO-1
LOOHs in oxLDL can induce HO-1 in vascular cells [21–23]. As seen in Figure 2, oxLDL (100 μg/ml for 18 h) resulted in a 5-fold induction of HO-1 in HUVEC. Pretreatment of oxLDL with 1 mmol/l H2S for 30 min at 37 °C, which completely destroyed LOOHs (results not shown), abolished the ability of the LDL preparation to induce HO-1.
H2S treatment decreases oxLDL-induced HO-1 in HUVEC
Reduction of 9-HPODE to 9-HODE by H2S
9-HPODE is a fatty acid peroxidation product found in oxLDL . Reduction of LOOHs to LOHs by thiol residues on PON (paraoxonase) has been reported by Aviram et al. . Thus we analysed the reaction product(s) of H2S with 9-HPODE by HPLC. When 9-HPODE was incubated for 30 min at 37 °C with H2S this treatment resulted in the formation of 9-HODE. The H2S concentration-dependent reduction (expressed as the ratio HPODE/HODE) is shown in Figure 3(A). Figure 3(B) depicts the time-dependent conversion of 9-HPODE into 9-HODE in the presence of H2S. Approx. 50% of 9-HPODE was reduced after 30 min.
H2S treatment reduces 9-HPODE to 9-HODE
Reduction of 9-HPODE in oxLDL by H2S
Figure 4 shows the influence of H2S treatment on 9-HPODE present in oxLDL. H2S treatment caused a significant decrease (as quantified via peak area) in 9-HPODE from 842±118 (oxLDL) to 544±18 (P<0.001). Ebselen/GSH treatment run as a positive control of LOOH reduction  resulted in a decrease of approx. 85% (121±18).
Reduction of 9-HPODE in oxLDL by H2S or ebselen treatment analysed by HPLC
The third gaseous transmitter H2S has been shown to protect cells and proteins from oxidative modifications by peroxynitrite and HOCl [7,26]. H2O2 is also known to be destroyed by H2S [6–8]. Thus one may assume that H2S may also destroy organic hydroperoxides of pathobiological importance, such as fatty acid hydroperoxides (LOOHs). Focusing on LOOHs in oxLDL the present results show that LOOHs were destroyed in the presence of H2S. This resulted in the inability of oxLDL to induce HO-1 in endothelial cells. Both results indicate that H2S could abrogate the pathophysiological activity of oxLDL. Destroying LOOHs in oxLDL could limit its atherogenic potential as hydroperoxides can (i) break down into lipid alkoxyl (LO•) as well as peroxyl radicals (LOO•), which both can attack and modify other biological molecules  are (ii) cytotoxic and (iii) break down to protein modifying aldehydes like MDA, 4-hydroxynonenal and 4-hydroxyhexenal [28,29].
PON associated with HDL (high-density lipoprotein) has been implicated as an important factor in destroying LOOHs on LDL and may contribute to the beneficial effect of HDL in atherogenesis and cytoprotection of vascular cells . The ability of PON to protect against LDL oxidation involves its free thiol group and not its arylesterase/PON activity  and the results revealed that the decrease in the levels of CE-LOOHs (cholesteryl ester hydroperoxides) by PON was due to the formation of CE-LOHs and the authors suggested a peroxidase-like action of PON on LOOHs . Using 9-HPODE, an LOOH compound found in oxLDL , our results showed that in the presence of H2S at concentrations that have been reported for in vivo (50–150 μmol/l) [32,33], 9-HPODE was reduced to 9-HODE assuming the reaction:
However, recent work has suggested that levels of free H2S in vivo are much lower than previously thought . On the other hand, as H2S can be stored as bound sulfur and released , one may speculate that the overall amount could be adequate to modulate pathophysiological reactions.
In summary, our results indicate that H2S besides PON may be an additional endogenous factor in reducing LOOHs in LDL. An early event in atherogenesis is the trapping and oxidative modification of LDL in the subendothelial space  and H2S (which is produced nearby and released from smooth-muscle cells ) may prevent or diminish the further atherogenic modifications of the lipoprotein.
- FOX assay
Xylenol Orange assay
human umbilical-vein endothelial cells
This work was supported by the Oesterreichische Nationalbank Jubiläumsfonds [grant number 10537 (to S. M. S.)].