Effect of Hemolyzed Plasma on the Batch Measurement of Nitrate by Nitric Oxide Chemiluminescence

Nitric oxide (NO) is a potent vasodilator and regulator of vascular tone, a neurotransmitter, and a cytotoxic agent (1)(2). In aqueous aerobic environments, the primary decomposition product of NO is nitrite (NO2−) (3), with further oxidation to nitrate (NO3−) being dependent on the presence of additional oxidizing species such as oxyhemoproteins (3). Collectively, these NO oxidation products are referred to as NOx−. Several analytical techniques have been used to quantify NOx− species, including spectrophotometric assays based on the Griess reaction (4)(5) and enzymatic reduction of NO3− (6), gas chromatography–mass spectrometry (7), chromatographic flow systems (8), and chemiluminescence (9)(10)(11). NOx− may be important in sepsis (12)(13)(14)(15)(16). Of possible analytical importance is intravascular hemolysis during sepsis (17)(18)(19). The effect of hemolysis on NOx− analysis is unknown. Other potential causes of intravascular hemolysis include drug-induced hemolytic anemia, hemolytic transfusion reactions, and artificial heart valves (20), and hemolysis can also occur during blood collection (21) and inappropriate blood storage. The objectives of our study were (a) to determine whether hemolysis interferes with the determination of plasma NO3− by NO chemiluminescence batch methodology and (b) to determine whether the interference could be eliminated by sample pretreatment. We purchased helium and oxygen from Praxair. Other chemicals were from Sigma-Aldrich. All chemicals were reagent-grade quality. Deionized water was used to prepare all solutions. Blood samples were obtained from healthy human volunteers by venipuncture with heparin as anticoagulant. Aqueous NO3− calibrators (25 μmol/L) were prepared in deionized water. Digitonin in phosphate-buffered saline (PBS), at final concentrations of 0, 20, 45, and 90 μmol/L, was used to control the degree of whole-blood hemolysis (21). Whole blood was also treated by dilution (1:2 by volume) with lysis buffer (155 mmol/L NH4Cl, 10 mmol/L KHCO3, 0.14 mmol/L EDTA, pH 7.2) to obtain complete hemolysis. Whole blood was diluted 1:2 by volume with digitonin solutions (35 °C) and lysis buffer and mixed gently for 5 min. Supernatants were collected by centrifugation (10 min at 1500g and 18 °C) and refrigerated. Hemoglobin (Hb) content was estimated by hemoximetry (OSM2 Hemoximeter; Radiometer) and determined quantitatively (plasma hemoglobin test reagent set; Sigma Diagnostics). The Hb plasma solutions were supplemented with NO3− (NO3−/Hb) to produce a final concentration of 25 μmol/L. Solutions were left at room temperature for 10 min to allow sample equilibration. In a separate series of tests, Hb was removed from the mildly hemolyzed sample (20 μmol/L digitonin) and the completely hemolyzed sample (lysis buffer) by precipitation with ethanol (1:2 by volume). Samples were mixed vigorously for 1 min and left standing at room temperature for 15 min before supernatants were collected by centrifugation (2 min at 5000g and 4 °C). Background NOx− (endogenous NO2−/NO3−) was subtracted from NO chemiluminescence signals. Digitonin at 1000 μmol/L had no effect on the NO3− calibrator signal. A NO chemiluminescence analyzer (NOA) system (Sievers 270B, NO Chemiluminescence Analyzer; Sievers Instruments) was used to detect NO3−. NO3− was chemically reduced under reflux conditions to NOgas in the Radical PurgerTM by hot (92 °C, circulating hot water bath) vanadium(III) (3.5 mL of 0.05 mol/L VCl3 in 0.8 mol/L HCl, pH 0.48) and subsequently stripped and carried to the NOA reaction chamber (under reduced pressure; Edwards Pump) by helium. The internal reaction chamber pressure was adjusted to 800 Pa. Within the reaction vessel, NOgas reacted with ozone to generate oxygen and the excited state NO2 species, which decayed to give a weak infrared (>600 nm) chemiluminescence signal that was detected and amplified by a photon multiplier tube (22)(23). Output signals in mV were recorded on a strip recorder (Chromatopak C-R1A; Shimadzu) with the areas of the NO peaks (mV · s) being electronically integrated from baseline to baseline. To test the effects of hemolysis on NO3− detection, a batch protocol was used: a 25 μmol/L aqueous NO3− calibrator was injected five times to establish a baseline NO signal (Pre-Hb) and was immediately followed by 10 repeated injections of a NO3−-supplemented hemolyzed plasma sample (NO3−/Hb) and then by 5 injections of the aqueous NO3− calibrator (Post-Hb). Ten-microliter injections (gas-tight syringe; Hamilton Company) were used for all calibrators and samples. The influence of Hb accumulation in the purge vessel on NO3− determination was calculated by the difference in NO signal between the Pre-Hb and Post-Hb aqueous NO3− calibrators. Duplicate batch runs of each treatment were performed. All values are reported as mean ± SE unless otherwise stated. For all tests of significance, P <0.05 was considered statistically significant. A t-test (Sigma Stat 2.0; Jandel Scientific) was used to assess the influence of the Hb on the difference between Pre-Hb and Post-Hb NO3− calibrators. CVs were based on the combined results of aqueous NO3− calibrators from duplicate batch experiments (n = 10). The effect of different degrees of hemolysis on NO3− determination is shown in Fig. 1 , A–E and G. As Hb accumulated in the purge vessel to concentrations >1000 μg (Fig. 1 , C, D, and G), the NO chemiluminescence response of the hemolyzed plasma samples decreased sequentially with repeated injections. The effect of injecting hemolyzed plasma samples on the subsequent determination of aqueous NO3− calibrators was assessed by comparing the NO responses of the calibrators before (Pre-Hb) and after (Post-Hb) injections of hemolyzed plasma. The results are summarized in Table 1 . As the degree of hemolysis increased from 2.9% to 30%, the difference between the values obtained for the Pre-Hb and Post-Hb aqueous NO3− calibrators increased from 5.2% (P = 0.047) to 19% (P <0.001), respectively, showing that sample hemolysis inhibited not only the plasma NO3− signal, but also that of the calibrator by a carryover effect. NO chemiluminescence batch profiles. Repeated injections (10 μL) of Pre-Hb and Post-Hb aqueous NO3− calibrators (• and ▴) were followed by repeated injections of the NO3−-supplemented hemolyzed plasma samples (NO3−/Hb; ○ and ▵) and a second series of aqueous NO3− calibrators. Whole-blood hemolysis was induced by digitonin (0, 20, 45, and 90 μmol/L) and lysis buffer. Panels A–D show the effect of Hb accumulation on NO3− determination. Panels E–H compare the effect of mildly hemolyzed untreated (20 μmol/L digitonin) and completely hemolyzed (lysis buffer) samples with ethanol-treated samples. ▪, accumulated Hb. Values on x axis indicate number of injections. Vertical dashed lines demarcate NO3− calibrators from hemolyzed samples. Effect of sample hemolysis and ethanol treatment on measurement of NO3− calibrators.1 To assess the effect of sample hemolysis on NO3− detection, chemiluminescence values (mV · s) from injections of the 25 μmol/L NO3− calibrators Pre-Hb were compared with chemiluminescence values from injections Post-Hb. Digitonin was used to induce hemolysis, and PBS was used as control. Ethanol was used to remove Hb from both mildly and completely hemolyzed samples. Duplicate batch analyses were performed. EtOH, ethanol; NS, not significant. CV, %. Hemolysis relative to whole-blood Hb content. Effect of sample hemolysis and ethanol treatment on measurement of NO3− calibrators.1 To assess the effect of sample hemolysis on NO3− detection, chemiluminescence values (mV · s) from injections of the 25 μmol/L NO3− calibrators Pre-Hb were compared with chemiluminescence values from injections Post-Hb. Digitonin was used to induce hemolysis, and PBS was used as control. Ethanol was used to remove Hb from both mildly and completely hemolyzed samples. Duplicate batch analyses were performed. EtOH, ethanol; NS, not significant. CV, %. Hemolysis relative to whole-blood Hb content. When mildly and completely hemolyzed samples were precipitated with ethanol (Fig. 1, F and H ), there was no evidence of sequential decreases in NO3−/Hb signal, nor was there a significant difference in the NO signal of the aqueous calibrator before and after exposure to precipitated samples (Table 1 ). Recovery of the NO signal was nearly quantitative from the hemolyzed plasma samples, 112% and 115% with 20 μmol/L digitonin and lysis buffer, respectively. The NO chemiluminescence technique, based on the chemical reduction of NOx− species to NO and subsequent reaction with ozone, has several advantages over other methods. These advantages include increased accuracy and precision, and reduced susceptibility to sample interference from colored species, suspended materials (24), and nitro-organic compounds (10). Braman and Hendrix (10) showed that reduction of NO2− and NO3− by vanadium(III) to NO was both temperature and pH dependent. Yang et al. (25) confirmed the suitability of vanadium(III) as the reductant of choice and indicated that a temperature of 80–90 °C should be used to reduce both NO2− and NO3− to NO. Although NO3− was the focus of this study, the conditions used would also detect NO2−. It has been reported that plasma protein causes extensive foaming in the purge vessel and that this interferes with the chemical reduction of NO2− and NO3− (25). We found that simply diluting the plasma fraction of whole blood threefold in PBS (Fig. 1B ) eliminated the requirement for deproteinization before batch processing recommended by Yang et al. (25). Although plasma protein may interfere with NOx− determinations, we found that the NO chemiluminescent response was compromised by the presence of Hb in the absence of extensive foaming. In fact, foaming was encountered only with the highest accumulations of Hb, i.e., >1000 μg, corresponding to samples with 19% and 30% hemolysis, whereas measurement of aqueous NO3− calibrators was affected by much lower degrees of hemolysis (2.9–7.0%) in the absence of foaming (Table 1 ). This suggests that the Hb itself is capable of interfering with the NOx− reduction; additional tests showed that neither free iron nor glutathione produced the effect. When Hb was removed from hemolyzed samples by ethanol precipitation, no deterioration of plasma NO3− signal or decrease between Pre-Hb and Post-Hb aqueous NO3− calibrators was detected. We wondered whether sample dilution accounted for the improvement in Post-Hb aqueous NO3− signal. With respect to the completely hemolyzed sample, a threefold dilution would have been equivalent to a sample with ∼10% hemolysis. This degree of hemolysis, according to Table 1 , would still have produced a significant decrease in the signal produced by the Post-Hb aqueous NO3− calibrators. Because this was not the case, we believe that simple dilution of Hb alone cannot account for the improvement in the NO chemiluminescent signal. According to Braman and Hendrix (10), the chemiluminescent method of detecting NO using vanadium(III) is superior to other methods because many samples can be analyzed sequentially without the necessity of sample prereduction or other preparation. Our results suggest that sample hemolysis ranging from 2.9% to 30% (note that 1–2% hemolysis gives a pinkish color to plasma) may compromise the accurate detection of NOx− species because Hb accumulates in the purge vessel. Ethanol precipitation can be used to clean up the sample with near-quantitative recovery of NOx− species. Other precipitants were not evaluated in the chemiluminescent system. As NOx− measurements gain increasing clinical acceptance in patient monitoring, it will be important to recognize which patient samples are hemolyzed so that corrective measures can be taken to ensure accurate analysis. This research was supported by Medical Research Council Grant MA-13941 (to C.G. Ellis). R.M. Bateman was supported by the Spoerel Research Fellowship, Department of Anesthesia, University of Western Ontario.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.