Psoriasis alters HDL composition and cholesterol efflux capacity.

Psoriasis, a chronic inflammatory skin disease, has been linked to increased myocardial infarction and stroke. Functional impairment of HDL may contribute to the excess cardiovascular mortality of psoriatic patients. However, data available regarding the impact of psoriasis on HDL composition and function are limited. HDL from psoriasis patients and healthy controls was isolated by ultracentrifugation and shotgun proteomics, and biochemical methods were used to monitor changed HDL composition. We observed a significant reduction in apoA-I levels of HDL from psoriatic patients, whereas levels of apoA-II and proteins involved in acute-phase response, immune response, and endopeptidase/protease inhibition were increased. Psoriatic HDL contained reduced phospholipid and cholesterol. With regard to function, these compositional alterations impaired the ability of psoriatic HDL to promote cholesterol efflux from macrophages. Importantly, HDL-cholesterol efflux capability negatively correlated with psoriasis area and severity index. We observed that control HDL, as well as psoriatic HDL, inhibited dihydrorhodamine (DHR) oxidation to a similar extent, suggesting that the anti-oxidative activity of psoriatic HDL is not significantly altered. Our observations suggest that the compositional alterations observed in psoriatic HDL reflect a shift to a pro-inflammatory profile that impairs cholesterol efflux capacity of HDL and may provide a link between psoriasis and cardiovascular disease.


Paraoxonase activity assay
Ca 2+ -dependent arylesterase activity was determined with a photometric assay using phenylacetate as the substrate ( 23 ). HDL (2 µg protein) or apoB-depleted serum (3 µl) was added to 200 µl buffer containing 100 mmol/l Tris, 2 mmol/l CaCl 2 (pH 8.0), and phenylacetate (1 mmol/l). The rate of hydrolysis of phenylacetate was monitored by the increase of absorbance at 270 nm, and readings were taken every 30 s at room temperature to generate a kinetic plot. The slope from the kinetic chart was used to determine ⌬ Ab 270nm / min. Enzymatic activity was calculated with the Beer-Lambert Law from the molar extinction coeffi cient of 1,310 L × mol Ϫ 1 × cm Ϫ 1 for phenylacetate.

Lp-PLA2 activity assay
Lp-PLA2 activity was measured with a commercially available photometric assay (Cayman Europe, Talinn, Estonia) using 2-thio PAF as substrate.

Determination of the anti-oxidative capacity of HDL
The anti-oxidative activity of HDL was determined as previously described with modifi cations ( 24 ). Briefl y, dihydrorhodamine (DHR) was suspended in DMSO to a 50 mmol/l stock, which was diluted in HEPES (20 mmol/L HEPES, 150 mmol/l NaCl, pH 7.4) containing 1 mmol/l 2,2'-azobis-2-methyl-propanimidamidedihydrochloride (AAPH) to a 50 µmol/l working reagent. In a 384-well, 7.5 µg HDL protein was placed, 15 µl of DHR working reagent was added, and the volume was completed to 100 µl with HEPES buffer. The increase in fl uorescence due to the oxidation of DHR was measured every 2 min for 1 h at 538 nm. The increase in fl uorescence per minute was determined for samples containing only DHR and for samples containing DHR and individual HDL samples from healthy controls or psoriasis patients.

LC-MS/MS analysis
Proteomic profi ling of HDL was performed as previously described ( 15 ). HDL was digested with trypsin, and the resulting peptides were separated using a nano-HPLC. The samples were ionized in the nanospray source equipped with nanospray tips and analyzed in a LTQ-FT mass spectrometer (Thermo Scientifi c, Waltham, MA). The standard deviation of spectral counts was below 10% between duplicates. Spectral counts were recorded and used for data analysis by searching the human Swis-sProt public database downloaded on May 4, 2011, with Spectrum Mill Rev. A.03.03.084 SR4 (Agilent, Vienna, Austria). Detailed search criteria were as follows: trypsin; maximum missed cleavage sites = 2; fi xed modifi cation = carbamidomethylation at cysteine; variable modifi cation = oxidized methionine; precursor mass tolerance = ±0.05 Da; and product mass tolerance = ±0.7 Da. Protein hits were subjected to automatic validation by Spectrum mill as In this explorative study, quantitative shotgun proteomic profi ling and biochemical analysis were used to investigate whether the systemic infl ammatory state in psoriasis modifi es HDL composition and function.

Characteristics of study subjects and blood collection
Blood was sampled from patients with moderate to severe chronic plaque-type psoriasis and healthy volunteers after obtaining written informed consent, according to a protocol approved by the Institutional Review Board of the Medical University of Graz (No. 21-523 ex. 09/10). Blood was collected in Vacuette serum tubes (Greiner, Kremsmünster, Austria). The clinical characteristics of control subjects (n = 15) and psoriatic patients (n = 15) are given in Table 1 . Plasma lipid profi les of psoriasis patients displayed characteristics of metabolic dyslipidemia, with increased plasma triglycerides and total cholesterol similar to that reported in previous studies ( 7,20 ). The patients had a mean psoriasis area and severity index (PASI) of 10.4, indicating a moderate to severe disease status.

Isolation of HDL
Serum density was adjusted with potassium bromide (Sigma, Vienna, Austria) to 1.24 g/ml, and a two-step density gradient was generated in centrifuge tubes (16 × 76 mm, Beckman) by layering the density-adjusted plasma (1.24 g/ml) underneath a NaCl density solution (1.006 g/ml) as described ( 15,21 ). Tubes were sealed and centrifuged at 90,000 rpm for 4 h in a 90Ti fi xed angle rotor (Beckman Instruments, Krefeld, Germany). After centrifugation, the HDL-containing band was collected, desalted via PD10 columns (GE Healthcare, Vienna, Austria), and immediately used for experiments or stored at Ϫ 70°C.

Statistical analysis
Differences in plasma and HDL parameters between control subjects and psoriatic patients were analyzed using 2-tailed Mann-Whitney U-test. Changes in the HDL proteome were evaluated from spectral counts of automatically validated proteins (i.e., the number of MS/MS spectra assigned to a protein). To assess for data normality, the Shapiro-Wilk test (at the level of 10%) was used. Because proteomic data markedly violate the assumption of normality, the Mann-Whitney U-test was employed for analysis of differences.
All correlations between compositional and functional data were determined with the use of Pearson product-moment estimates. Signifi cance was accepted at * P < 0.05 and ** P < 0.01. Statistical analyses were performed with PASW Statistics version 18.

Psoriasis is associated with altered HDL composition
We hypothesized that HDL from psoriatic patients might display altered protein cargo and lipid composition. Therefore, we isolated HDL from psoriatic patients (n = 15) and healthy individuals (n = 15) by ultracentrifugation. To identify HDL-associated proteins, we performed proteomic analysis of purifi ed HDL using LC-MS/MS. Table 1 describes the clinical characteristics of the subjects studied. We identifi ed 38 HDL-associated proteins in control subjects and 55 HDL-associated proteins in psoriatic subjects. Identifi ed HDL-associated proteins were grouped into functional categories ( Table 2 ); statistical analysis identifi ed several proteins to be signifi cantly altered ( Table 2 ). The analysis revealed that apoA-I and apoM were significantly reduced, whereas several acute-phase proteins, including SAA, prothrombin, and ␣ -1-acid glycoprotein 1, were increased. Moreover, we observed an increased content of apoA-II, ␣ -1-antitrypsin, and IgA-1 on HDL isolated from psoriasis patients.
To validate the proteomic result, we performed immunoturbidimetry analysis of major apolipoproteins (apoA-I, apoA-II, apoC-II, apoC-III, and apoE), demonstrating that the proteomic data are valid (supplementary Table I).
In addition, we analyzed the lipid moiety of HDL for its major lipid classes. HDL derived from psoriatic patients displayed a signifi cantly decreased content of total cholesterol, phosphatidylcholine, and sphingomyelin ( Table 3 ). LPC content was not signifi cantly altered.

Psoriatic HDL shows reduced cholesterol effl ux capacity from macrophages
Cholesterol effl ux capacity of apoB-depleted serum shows a strong inverse correlation with the likelihood of coronary artery disease ( 10 ). Strikingly, HDL from psoriatic patients was signifi cantly less effi cient in promoting cholesterol effl ux from macrophages compared with controls ( Fig. 1A ). Importantly, PASI negatively correlated with cholesterol effl ux capacity of isolated HDL ( Fig. 1B ), whereas serum C-reactive protein (CRP) values and gender did not correlate with HDL-cholesterol effl ux capacity (supplementary Table II).

Altered HDL composition is linked to impaired cholesterol effl ux capacity
To assess which compositional alterations are involved in the impairment of cholesterol effl ux capacity, we performed correlation analysis between cholesterol effl ux capability of HDL and compositional data. We observed that HDL-associated phosphatidylcholine ( Fig. 2A ), HDLassociated sphingomyelin ( Fig. 2B ) and HDL-associated apoA-I ( Fig. 2C ) were the strongest positive predictors of HDL-cholesterol effl ux capacity. Interestingly, apoA-II, of which the overall amount was increased in psoriatic patients, was not associated with cholesterol effl ux capability (supplementary Table III).
A full list of correlations of cholesterol effl ux capacity with all HDL-associated proteins can be found in supplementary Table III.

Capability of psoriatic HDL to inhibit dihydrorhodamine oxidation (anti-oxidative activity) is not altered
Besides the important role in lipid metabolism, HDL exhibits unique anti-oxidative activity. Therefore, we assessed the intrinsic ability of HDL to be oxidized by measuring increasing fl uorescence due to dihydrorhodamine 123 oxidation over time ( 24 ). Interestingly, we observed that control HDL and HDL from psoriatic patients inhibited DHR oxidation to a similar extent ( Fig. 3A ), indicating that the ability of psoriatic HDL to inhibit DHR oxidation is not altered. Previous data suggested that HDL-associated paraoxonase (PON) and/or lipoprotein associated phospholipase A2 (Lp-PLA2) contribute to the anti-oxidant activity of HDL ( 25,26 ). In good agreement with proteomic data ( Table 2 ), we observed that PON activity was not altered ( Fig. 3B ), whereas Lp-PLA2 activity was significantly increased in psoriatic HDL ( Fig. 3C ) and correlated with disease severity ( Fig. 3D ). Similar results were obtained when PON and Lp-PLA2 activities were assessed in Our results suggest that a loss of apoA-I, phosphatidylcholine, and sphingomyelin content of psoriatic HDL is a key determinant of the low cholesterol effl ux capability. This assumption is based on the highly signifi cant correlation between apoA-I/phosphatidylcholine content and the cholesterol effl ux capability of HDL. Sphingomyelin is known to bind cholesterol with high affi nity, suggesting that the sphingomyelin content of HDL contributes apoB-depleted serum of controls and psoriatic patients (supplementary Fig. I).

DISCUSSION
The data presented here describe for the fi rst time marked alterations in the composition and function of psoriatic HDL and provide evidence that psoriatic HDL takes on a pro-infl ammatory profi le. We observed a significantly impaired capacity of psoriatic HDL to mobilize cholesterol from macrophages, the critical fi rst step of reverse cholesterol transport. Importantly, cholesterol effl ux capability negatively correlated with the severity of psoriasis. Our fi ndings raise the possibility that dysfunctional HDL contributes to accelerated atherosclerosis in psoriatic patients.
Shotgun proteomic profi ling and biochemical analyses were applied to investigate psoriasis-induced alterations in the HDL proteome and lipid composition. Our present data suggest that psoriasis shifts the HDL composition into a pro-infl ammatory profi le.
Psoriatic HDL contained less phospholipids and cholesterol and markedly reduced levels of apoA-I, whereas apoA-II and several acute-phase proteins, such as SAA, ␣ -1-antitrypsin, prothrombin, and ␣ -1-acid-glycoprotein 1, were signifi cantly increased. Moreover, proteins involved in complement activation, such as complement C3 and hemoglobin and its scavenger protein haptoglobin, were enriched in psoriatic HDL. Interestingly, recent studies reported signifi cantly increased complement C3, hemoglobin, haptoglobin, and hemopexin in HDL of coronary heart disease patients, which lost its anti-infl ammatory capacity by accumulation of proinfl ammatory SAA1 ( 28,29 ).
Although cholesterol effl ux from macrophages represents only a small fraction of the overall fl ux through reverse cholesterol transport, it is probably the most relevant factor regarding atheroprotection ( 10 ). In the present study, we observed that many of the proteins that are enriched in psoriatic HDL (with the exception of apoA-II) negatively correlated with the cholesterol effl ux capability of HDL. In addition to the changes in the protein composition of HDL, a decrease in the phospholipid and cholesterol content of HDL was observed, which strikingly signifi es that psoriasis induced alterations in the lipid composition of HDL.  In summary, the results of our study indicate that compositional alterations observed in psoriatic HDL refl ect a pro-infl ammatory profi le. Consequently, such compositional alterations may improve our ability to monitor therapeutic responses and may provide a novel basis for the identifi cation of psoriatic patients at increased risk of cardiovascular disease. The abnormal effl ux capacity of HDL in psoriatic patients may provide a link between the association of psoriasis and cardiovascular disease; however, larger studies are needed to validate these fi ndings.
to cholesterol association with HDL, as previously suggested ( 30 ).
In accordance with our fi ndings, previous studies suggested that antiatherogenic activities of HDL are inversely correlated with systemic infl ammation in rheumatoid arthritis patients (31)(32)(33). Moreover, a recent study has provided evidence that autoantibodies against apoA-I contribute to reduced HDL levels in systemic lupus erythematosus patients, independently of hepatic HDL biogenesis. This fi nding indicates that premature clearance of apoA-I/ HDL immune complexes may contribute to compositional alterations and low apoA-I/HDL levels ( 34 ). As psoriatic HDL contains increased levels of (auto)antibodies, immune complex formation may result in accelerated clearance and may render HDL dysfunctional.
Notably, we observed that platelet basic protein content was signifi cantly increased in psoriatic HDL compared with control HDL. This may be of particular interest, as there is evidence for in vivo platelet activation in psoriatic patients contributing to the development of thrombotic events ( 35 ).
Another important finding of our study was that Lp-PLA2 activity, also known as platelet-activating factor acetylhydrolase (PAF-AH), is signifi cantly increased in psoriatic HDL. Previous studies have implicated that the potent phospholipid mediator PAF (1-O -alkyl-2-acetylsn-glycero-3-phosphocholine) plays a role in the pathogenesis of psoriasis ( 36 ). Plasma levels of PAF were found to be increased in patients with psoriasis, and a signifi cant decrease in PAF levels was observed with clinical improvement after treatment ( 37 ). Therefore, increased HDL-associated Lp-PLA2 activity in psoriatic patients may more effectively inactivate PAF and contribute to clinical improvement. On the other hand, a recent meta-analysis reported that Lp-PLA2 levels are positively associated with an increased risk of developing coronary artery disease ( 38 ); hence, Fig. 3. Anti-oxidative capacity of HDL form psoriatic patients is not impaired. (A) The anti-oxidative activity of HDL was determined by inhibition of AAPH-initiated oxidation of DHR. Incubation of DHR in the presence of HDL from healthy subjects (control) or psoriasis patients (Psoriasis) led to a reduction in the oxidation of DHR. Results represent measurements of two independent experiments. (B) Arylesterase activity of HDL-associated PON was measured using phenylacetate as substrate. (C) Lipoprotein-associated phospholipase A2 (Lp-PLA2) activity of HDL was measured using 2-thio PAF as substrate. The arylesterase and Lp-PLA2 activities of HDL were calculated from the slopes of the kinetic chart of three independent experiments. (D) Correlation between PASI and Lp-PLA2 activities of control and psoriatic HDL. The Pearson correlation coeffi cient is noted; * P < 0.05.