Coupling capillary electrophoresis with mass spectrometry for the analysis of oxidized phospholipids in human high-density lipoproteins.

The oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (ox-PAPC) products in human high-density lipoproteins (HDLs) were investigated by low-flow capillary electrophoresis-mass spectrometry (low-flow CE-MS). To accelerate the optimization, native PAPC (n-PAPC) standard was first analyzed by a commercial CE instrument with a photodiode array detector. The optimal separation buffer contained 60% (v/v) acetonitrile, 40% (v/v) methanol, 20 mM ammonium acetate, 0.5% (v/v) formic acid, and 0.1% (v/v) water. The selected separation voltage and capillary temperature were 20 kV and 23°C. The optimal CE separation buffer was then used for the low-flow CE-MS analysis. The selected MS conditions contained heated capillary temperature (250°C), capillary voltage (10 V), and injection time (1 s). No sheath gas was used for MS. The linear range for n-PAPC was 2.5-100.0 µg/mL. The coefficient of determination (R2 ) was 0.9918. The concentration limit of detection was 1.52 µg/mL, and the concentration limit of quantitation was 4.60 µg/mL. The optimal low-flow CE-MS method showed good repeatability and sensitivity. The ox-PAPC products in human HDLs were determined based on the in vitro ox-PAPC products of n-PAPC standard. Twenty-one ox-PAPC products have been analyzed in human HDLs. Uremic patients showed significantly higher levels of 15 ox-PAPC products than healthy subjects.

Determination of oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine products in human very low-density lipoproteins by nonaqueous low-flow capillary electrophoresis-mass spectrometry.

A simple and fast low-flow capillary electrophoresis-mass spectrometry (low-flow CE-MS) method has been developed to analyze oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine (ox-PAPC) products in human very low-density lipoproteins (VLDLs). Native PAPC standard was analyzed to optimize the low-flow CE-MS method. The optimal CE conditions included separation buffer (60% (v/v) acetonitrile, 40% (v/v) methanol, 0.1% (v/v) water, 0.5% (v/v) formic acid, 20 mM ammonium acetate), sheath liquid (60% (v/v) acetonitrile, 40% (v/v) methanol, 0.1% (v/v) water, 20 mM ammonium acetate), separation voltage (20 kV), separation capillary internal diameter (i.d.) (75 µm), separation capillary temperature (23˚C) and sample injection time (6 s). The selected MS conditions included heated capillary temperature (250°C), capillary voltage (10 V), and injection time (1 s). Sheath gas was not used in this study. The total ion chromatograms (TICs), extracted ion chromatograms (EICs) and MS spectra of native PAPC standard and its in vitro oxidation products showed good repeatability and sensitivity. To determine the ox-PAPC products in human VLDLs, the EICs and MS spectra of VLDLs were compared with the in vitro oxidation products of native PAPC standard. For native PAPC standard, the measured linear range was 2.5 - 100.0 µg/mL, and the coefficients of determination (R2) was 0.9994. The concentration limit of detection (LOD) was 0.44 µg/mL, and the concentration limit of quantitation (LOQ) was 1.34 µg/mL. A total of 21 ox-PAPC products were analyzed for the VLDLs of healthy and uremic subjects. The levels of 7 short-chain and 5 long-chain ox-PAPC products on uremic VLDLs were significantly higher than healthy VLDLs. This simple low-flow CE-MS method might be a good alternative for LC-MS for the analysis of ox-PAPC products. Furthermore, it might also help scientists to expedite the search for uremic biomarkers.

Cyclodextrin-micellar electrokinetic chromatography of apolipoproteins on human very low-density lipoprotein.

The apolipoproteins (APOs) of human very low-density lipoprotein (VLDL) were investigated by an optimized cyclodextrin-micellar electrokinetic chromatography (CD-MEKC) method. The separation buffer consisted of 20 mM sodium phosphate, 40 mM bile salts (50% sodium cholate and 50% sodium deoxycholate), 25 mM carboxymethyl-ß-cyclodextrin (CM-ß-CD) (pH 7.0). For CD-MEKC separation, a sample injection time of 12 s, a separation voltage of 15 KV, and a capillary temperature of 15°C were chosen. The optimal CD-MEKC method showed good resolution and repeatability for VLDL APOs. Identification and quantitation of VLDL APOs CI, CIII, and E were based on comparison with human APO standards. Good linear relationships with correlation coefficient (R2 ) 0.99 were obtained for APOs CI, CIII, and E standards. For these three APOs, the linear ranges were within 0.01-0.54 mg/mL, and the concentration limits of detection (LODs) were lower than 0.02 mg/mL. Moreover, VLDL APOs from four uremic patients and four healthy subjects were compared. The uremic and healthy CD-MEKC profiles showed dramatic difference. The levels of APO CIII were significantly higher for two patients, and the level of APO E was significantly higher for one patient. This study might be helpful for following the disease development of uremia and cardiovascular disease (CVD) in the future.

Investigating apolipoproteins of human high-density lipoprotein by cyclodextrin-micellar electrokinetic chromatography.

A cyclodextrin-micellar electrokinetic chromatography (CD-MEKC) method has been developed to determine the apolipoproteins (apos) of human high-density lipoprotein (HDL). The optimal CD-MEKC conditions included a separation buffer mixture of 5 mM sodium phosphate, 40 mM bile salts (50% sodium cholate and 50% sodium deoxycholate), 25 mM carboxymethyl-ß-CD (CM-ß-CD) and pH 7.0. The separation voltage was 15 kV, and the capillary temperature was 15℃. The CD-MEKC profiles of human HDL apolipoproteins showed good repeatability and sensitivity. Linear analysis has been performed for human apolipoprotein standards including apos AI, AII, CI, CII, CIII and E. Linear regression lines with coefficients of determination (R2) greater than 0.99 were obtained for apos AI, AII, CI, CII and E. The linear ranges for the six apolipoproteins were within 0.18-0.70 mg/mL, and the concentration limits of detection (LOD) were lower than 0.0617 mg/mL. Apos AI, AII, CI and CIII were identified and quantified in human HDL by comparing with apolipoprotein standards. Furthermore, the CD-MEKC profiles of uremic patients differed significantly from healthy subjects. The concentration ratios of apo AI/apo CIII were significantly lower for uremic patients than healthy subjects. This study demonstrated the feasibility of determining human HDL apolipoproteins by CD-MEKC. In the future, it might help monitor the progression of uremia and cardiovascular disease.

Separation of total lipids on human lipoproteins using surfactant-coated multiwalled carbon nanotubes as pseudostationary phase in capillary electrophoresis.

Surfactant-coated multiwalled carbon nanotubes (MWNTs) were used as pseudostationary phase (PSP) in CE to investigate the total lipids of high-density lipoproteins and low-density lipoproteins. To optimize the CE conditions, several experimental factors including carbon nanotube concentration, bile salt concentration, sodium phosphate (PB) concentration, organic modifier concentration and buffer pH value have been examined. In addition, the CE capillary temperature and applied voltage have also been examined. The optimal separation buffer selected was a mixture of 3.2 mg/L MWNT, 50 mM bile salt, 10 mM PB, 20% 1-propanol, pH 9.5. The optimal capillary temperature and applied voltage selected were 50°C and 20 kV, respectively. Phosphatidyl choline (PC) has been used as a model analyte and investigated by the optimal CE method. The linear range for PC was 0.1-3 mg/mL with a correlation coefficient of 0.9934, and the concentration LOD was 0.055 mg/mL. The optimal CE method has been used to characterize the total lipids of high-density lipoprotein and low-density lipoprotein. At absorbance 200 nm, one major peak and two or three minor peaks showed for the total lipids of lipoproteins within 13 minutes. Resolutions of the total lipids were enhanced using surfactant-coated MWNTs as PSPs in the CE separation buffer. However, resolutions of the total lipids were not enhanced using surfactant-coated single-walled carbon nanotubes as PSPs in the CE separation buffer.

Detection of C-reactive protein based on magnetic nanoparticles and capillary zone electrophoresis with laser-induced fluorescence detection.

A simple and fast method based on magnetic nanoparticles (MNPs) and capillary zone electrophoresis (CZE) with laser-induced fluorescence (LIF) detection was developed for the detection of C-reactive protein (CRP). To optimize the CZE conditions, several factors including buffer compositions, buffer ionic strength, buffer pH, applied voltage and capillary temperature have been examined. The optimal separation buffer selected was a 30 mM sodium phosphate (PB) buffer, pH 8.0. The optimal CE applied voltage and temperature selected were 20 kV and 35°C, respectively. The CZE profile of the MNP-1°Ab-CRP-2°Ab/FITC bioconjugates showed good reproducibility. One major peak was observed for the MNP bioconjugates. The quantitative analysis also showed good results. The coefficient of variation (CV%) for the major peak area was 8.7%, and the CV% for the major peak migration time was 2.5%. The linear range for CRP analysis was 10-150 µg/mL, and the concentration limit of detection (LOD) was 9.2 µg/mL. Non-specific interactions between bovine serum albumin (BSA) and the system can be prevented by including 10% (v/v) of human plasma in the binding buffers. The CE/LIF method might be helpful for analyzing high concentrations of CRP in a patient's plasma after an acute-phase inflammation. This new method demonstrated the possibility of using MNPs and CE/LIF for the detection of proteins, and provided information for the establishment of appropriate CE conditions.

Characterization of in vitro modified human very low-density lipoprotein particles and phospholipids by capillary electrophoresis.

A simple capillary zone electrophoresis (CZE) method was used to characterize human very low-density lipoprotein (VLDL) particles for four healthy donors. One major peak was observed for native, in vitro oxidized and glycated VLDL particles. The effective mobilities and peak areas of the capillary electrophoresis (CE) profiles showed good reproducibility and precision. The mobility of the oxidized VLDL peak was higher than that of the native VLDL. The mobility of the glycated VLDL peak was similar to that of the native VLDL. Phospholipids isolated from VLDL particles were analyzed by our recently developed micellar electrokinetic chromatography (MEKC) with a high-salt stacking method. At absorbance 200 nm, the native VLDL phospholipids showed a major peak and a minor peak for each donor. For oxidized VLDL phospholipids, the area of the major peak reduced for three donors, possibly due to phospholipid decomposition. For glycated VLDL phospholipids, the peak mobilities were more positive than native VLDL phospholipids for two donors, possibly due to phospholipid-linked advanced glycation end products (AGEs). Very interestingly, at absorbance 234 nm, the major peak of oxidized VLDL phospholipids was resolved as two peaks for each donor, possibly due to conjugated dienes formed upon oxidation.

Chemical analysis of C-reactive protein synthesized by human aortic endothelial cells under oxidative stress.

C-reactive protein (CRP) is a clinical biomarker of inflammation, and high levels of CRP correlate with cardiovascular disease. The objectives of this study were to test our hypothesis that oxidized low-density lipoprotein (ox-LDL) induces the release of CRP from human aortic endothelial cells (HAECs) and to optimize several analytical methods to identify CRP released from cultured cells in a model of atherogenic stress. HAECs were incubated with copper-oxidized LDL, and the supernatant was subsequently purified by diethylaminoethyl chromatography and analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). We identified an optimal buffer for the elution of CRP, which contained 0.05 M sodium phosphate and 2.0 M NaCl (pH 4.5). Purified CRP was digested with trypsin and subjected to high-performance LC with an optimal mobile phase of acetonitrile-water containing 0.1% formic acid (50:50, v/v) and an optimal mobile phase flow rate of 0.2 mL/min. We identified optimal parameters for MS/MS analysis of CRP, including sheath gas pressure (80 psi), capillary temperature (275 °C), collision energy (25%), tube lens offset (-5 V), auxiliary gas pressure (0 psi), and isolation width of parent ion (m/z value = 3). Characterization of CRP was based on the extracted ion chromatograms and selected multiple-reaction monitoring spectra of three peptides (peptide-1, -2, and -3) derived from trypsin-digested intact CRP standard. CRP peptide-2 and peptide-3 were identified in the supernatant of ox-LDL-treated HAECs. Confirmation of CRP was based on LC-MS/MS and enzyme-linked immunosorbent assay analysis of CRP in purified HAEC supernatant, as well as real-time PCR analysis of CRP mRNA levels in HAECs.

Analysis of in vitro oxidized human LDL phospholipids by solid-phase extraction and micellar electrokinetic capillary chromatography.

Phospholipids of in vitro oxidized human low-density lipoproteins (LDL) were separated by two different solid-phase extraction (SPE) methods. One of the two methods was designed to test the effects of gradient elution. This SPE method isolated more phospholipids from in vitro oxidized LDL than the other one according to the results of liquid chromatography and electrospray ionization mass spectrometry (LC ESI-MS) analysis. A micellar electrokinetic capillary chromatography (MEKC) method was also used to analyze phospholipids separated by SPE. The results of MEKC and LC ESI-MS were consistent for the major phospholipid classes, including PC, lysoPC, PE, PI and PS. The MEKC profiles showed significant differences for native and oxidized LDL phospholipids. Therefore, the unique combination of SPE and MEKC methods showed dramatic distinctions between native and in vitro oxidized human LDL phospholipids. The combination also shows great potential for rapid analysis of in vivo oxidized human LDL phospholipids in the future.

Micellar electrokinetic chromatography profiles of human high-density lipoprotein phospholipids.

A simple and fast micellar electrokinetic chromatography (MEKC) method was developed to investigate phospholipids isolated from human high-density lipoproteins (HDL). To optimize the MEKC conditions, several factors including bile salt concentration and organic modifier concentration in the separation buffer as well as temperature have been examined. The optimal separation buffer chosen was a mixture of 50 mM bile salts, 30% v/v 1-propanol and 10 mM sodium phosphate (pH 8.5). The applied voltage and temperature selected were 25 kV and 40°C, respectively. Meanwhile, high-salt stacking has been performed for sample pre-concentration to enhance peak sensitivity. Several factors including organic modifier concentration and salt concentration in the sample matrix as well as sample injection time have been optimized. The optimal sample buffer selected was a mixture of 100 mM NaCl and 20% 1-propanol, and the optimal sample injection time selected was 32 s under a pressure of 0.5 psi. Several phospholipid standards including lysophosphatidyl choline, phosphatidyl choline (PC), sphingomyelin, phosphatidyl ethanolamine, phosphatidyl inositol, phosphatidyl serine and phosphatidic acid have been studied using the optimal MEKC method. The MEKC profile of the mixed phospholipid standards showed good separation and reproducibility. The linear ranges for PC and sphingomyelin were 0.025-1.2 and 0.025-2.0 mg/mL, respectively. The concentration limits of detection of PC and sphingomyelin were 0.0156 and 0.0199 mg/mL, respectively. Using phosphatidic acid as an internal standard, precision and accuracy have been measured for PC and sphingomyelin. The intraday and interday quantitative analysis showed good results. The new MEKC method has been used to characterize native, in vitro oxidized and glycated human HDL phospholipids within 16 min. At absorbance 200 nm, two similar peaks were observed for native and oxidized HDL phospholipids, but three peaks were observed for glycated HDL phospholipids. Interestingly, at absorbance 234 nm, distinctively different MEKC profiles were observed for the three HDL phospholipids.

Characterization of in vitro modified human high-density lipoprotein particles and phospholipids by capillary zone electrophoresis and LC ESI-MS.

A simple capillary zone electrophoresis (CZE) method was used to characterize native, in vitro oxidized and glycated human high-density lipoprotein (HDL) particles. Both native and in vitro oxidized HDL capillary electrophoresis (CE) profiles showed a major peak, but the oxidized HDL particles had higher effective mobilities. The in vitro glycated HDL particles showed a major peak and one or two minor peaks. The effective mobility of the major peak of glycated HDL was similar to that of the major peak of native HDL, whereas the effective mobilities of the two minor peaks were much lower. For the analysis of HDL phospholipids, a solid phase extraction procedure was optimized and a LC ESI-MS method was developed. Several possible HDL phospholipid molecular species including phosphatidylcholine (PC 16:0/18:2, 16:0/18:1, 18:0/18:2 and 18:0/18:1), sphingomyelin (SM 16:0) and lyso-phosphatidylcholine (lysoPC 16:0 and 18:0) were found. It appeared that the ion intensity ratios of hydroperoxy-PC or epoxyhydroxy-PC (16:0/hydroperoxy-18:2 or 16:0/epoxyhydroxy-18:2, m/z 790.4) and trihydroxy-PC (16:0/trihydroxy-18:2, m/z 808.3) relative to PC (C16:0/C18:2, m/z 758.5) were higher for oxidized HDL than for native and glycated HDL. It should be helpful to use both CZE and LC ESI-MS methods for analyzing high-density lipoproteins from patients of cardiovascular disease. Their combination may be also useful for further studies concerning the role of oxidized and glycated HDLs in the development of atherosclerosis.

Investigation of C-reactive protein binding to phosphatidyl choline by CZE and ESI-mass analysis.

A simple CZE has been developed to investigate the binding reactions between C-reactive protein (CRP) and native as well as Cu(2+) oxidized phosphatidyl choline (PC) in this study. An aqueous-organic solvent system containing 20% sodium phosphate, 70% methanol and 10% acetonitrile was selected as the optimal CZE separation buffer. Native PC showed a broad, higher mobility major peak and a sharp, lower mobility minor peak. Oxidized PC (ox-PC) appeared as two adjacent peaks. PC bound to CRP in the presence of metal ions Ca(2+) and Mg(2+). The absorbance at 214 nm for PC+CRP was about five times of that for PC. Surprisingly, ox-PC interacted much more weakly with CRP. The absorbance at 214 nm for ox-PC+CRP was close to that of ox-PC. Meanwhile, ESI-Mass analysis also suggested that four PC molecular species (C16:0/C18:1, C16:0/C18:2, C18:0/C18:2, C18:0/C18:1) bound with CRP since the ion intensities of PC+CRP were lower than that of PC alone. This study uses both CZE and ESI-Mass analysis to characterize the interaction between CRP and native as well as Cu(2+) oxidized PC, and provides new analytical methods for determining the binding reactions. In addition, it also gives new insight into the molecular interactions between CRP and native as well as oxidized PC.

In vitro oxidized and glycated human low-density lipoprotein particles characterized by capillary zone electrophoresis.

A simple capillary zone electrophoresis (CZE) method was used to determine native, in vitro Cu(2+) and glucose modified low-density lipoprotein (LDL) particles for four healthy subjects. The LDL electropherograms are highly reproducible with good precisions of effective mobility and peak area. The native LDL capillary electrophoresis (CE) profile shows a major peak with lower mobility and two minor peaks with higher mobilities. For three-hour Cu(2+) oxidation, one major peak with mobility close to that of the native major peak, and one minor peak with mobility extending to -47 x 10(-5)cm(2)V(-1)s(-1) appear. For eighteen-hour Cu(2+) oxidation, one major peak with mobility much higher than that of the native major peak appears. As the reaction time for LDL and Cu(2+) increases from 0 to 24h, effective mobility of the LDL major peak increases, suggesting that LDL particles become more negatively charged and oxidized as the time increases. The in vitro glycated LDL particles are characterized by a major peak and two minor peaks. Mobility of the major peak is close to that of native major peak, but the second minor peak is much more negatively charged with mobility extending to -53 x 10(-5)cm(2)V(-1)s(-1). Native, oxidized and glycated LDL particles show distinctive differences in their CZE profiles. Agarose electrophoresis shows that the charge to mass ratios of native, three-hour Cu(2+) and glucose modified LDL particles are similar, but that of eighteen-hour Cu(2+) oxidized LDL particles is higher.

Separation and determination of in vitro oxidized phospholipids by capillary zone electrophoresis.

A simple capillary zone electrophoresis (CZE) method was used to determine in vitro oxidized phosphatidyl choline (ox-PC). To optimize the capillary electrophoresis (CE) conditions, organic buffer additives, buffer ionic strength, buffer pH, and applied voltage were examined. The optimal CE separation buffer chosen was an aqueous-organic solvent system containing 10% sodium phosphate buffer (5 mM, pH 7.40), 80% methanol, and 10% acetonitrile. One major peak with a small shoulder was found for phosphatidyl choline (PC), whereas one major peak and a complex region containing several lower-mobility peaks were found for ox-PC. The lower-mobility species of ox-PC has high levels of conjugated dienes characterized by strong absorbance at 234 nm. The electropherograms of PC and ox-PC were significantly different and highly reproducible. The intensities of lower-mobility species decreased significantly when the antioxidant vitamin C concentration was increased from 6 to 600 microM. This study provides a simple CZE method to differentiate in vitro oxidized from nonoxidized PC molecular species.

In vitro regulation of single collagen fibril length by buffer compositions and temperature.

An understanding of collagen ultrastructure is very important for designing biopolymers mimicking collagen functions in tissue engineering, or for diagnosing abnormal collagen structure in clinical study. The present study examined formation of a large population of type I collagen single fibrils under different buffer compositions and temperatures. Fibril structures were investigated by dark-field microscopy and atomic force microscopy (AFM). In the phosphate buffered saline (PBS) buffer, the average lengths of single fibrils were 4.8+/-2.2, 5.0+/-1.9 and 9.2+/-5.0 microm for 37 degrees C, 33 degrees C and 29 degrees C, respectively. The differences were significant (P < 0.05) between 37 degrees C and 29 degrees C and between 33 degrees C and 29 degrees C. In the sodium phosphate (SP) buffer, the average lengths of single fibrils were 10.6+/-5.4, 11.1+/-4.5 and 19.6+/-11.7 microm for 37 degrees C, 33 degrees C and 29 degrees C, respectively. Similarly, the differences were significant (P < 0.05) between 37 degrees C and 29 degrees C and between 33 degrees C and 29 degrees C. While at the same temperature, the average lengths of single fibrils differed significantly (P < 0.05) between PBS and SP buffers. Single fibrils formed in SP buffer were found to have greater average length than those formed in PBS buffer.

Charge density profiling of circulating human low-density lipoprotein particles by capillary zone electrophoresis.

Capillary zone electrophoresis (CZE) has been utilized to profile the low-density (LDL) particles in human blood serum in this study. A 5 mM sodium phosphate buffer, pH 7.40, was chosen as the most suitable CE buffer and an extensive ultrafiltration (UF) procedure was applied to purify the LDL sample. Two LDL particle species, LDL with lower mobility and LDL- with higher mobility were observed. The electropherograms were highly reproducible with good precision of effective mobilities, corrected peak areas (CPAs) and CPA ratio of LDL-/LDL. LDL particles shown on the electropherogram were also characterized by several procedures. The applications of Sigma HDL cholesterol reagent and CE on-line 2-propanol precipitation indicated that the two particle species shown in the electropherogram belong to LDL. The LDL particles were found to associate with the buoyant LDL fraction and the LDL- particles associate with the dense LDL fraction. This study utilizes CZE for the profiling of LDL isoforms and provides a new analytical method for the resolution of LDL subspecies. It demonstrates a high-mobility LDL particle which circulates in healthy subjects and diminishes in atherosclerotic patients. Diminution of the high-mobility LDL subspecies may be linked to minimal formation of arterial plaque in atherosclerotic patients.

Adiponectin, inflammation, and the expression of the metabolic syndrome in obese individuals: the impact of rapid weight loss through caloric restriction.

Severe obesity increases the prevalence of the metabolic syndrome, and moderate acute weight loss with a very low-calorie diet in obese subjects with the metabolic syndrome leads to significant metabolic benefits. Adiponectin has been implicated in both the pathogenesis of obesity-related insulin resistance and increased inflammation. We analyzed the relationship of the adipocyte-derived hormone adiponectin with indices of inflammation, adiposity, and insulin resistance in obese subjects with (MS+, n = 40) and without (MS-, n = 40) the metabolic syndrome and examined the acute effects of rapid weight loss. MS+ subjects had significantly lower adiponectin (7.6 +/- 0.6 vs. 10.4 +/- 0.6 microg/ml; P = 0.003) and significantly higher TNF-alpha (3.3 +/- 0.2 vs. 2.8 +/- 0.3 pg/ml; P = 0.004) levels compared with MS- subjects matched for age and body mass index. Plasma adiponectin and TNF-alpha levels were inversely related to the number of metabolic syndrome factors in a stepwise manner. After 4-6 wk of weight loss, there was marked improvement in glucose, insulin, leptin, and triglycerides, whereas adiponectin and TNF-alpha concentrations did not change. Thus, increases in plasma levels of adiponectin or reductions in TNF-alpha are not required for marked improvements in glucose/insulin and lipid metabolism with acute weight loss.