Artifacts Introduced by Sample Handling in Chemiluminescence Assays of Nitric Oxide Metabolites

Abstract

We recently developed a combination of four chemiluminescence-based assays for selective detection of different nitric oxide (NO) metabolites, including nitrite, S-nitrosothiols (SNOs), heme-nitrosyl (heme-NO), and dinitrosyl iron complexes (DNICs). However, these NO species (NOx) may be under dynamic equilibria during sample handling, which affects the final determination made from the readout of assays. Using fetal and maternal sheep from low and high altitudes (300 and 3801 m, respectively) as models of different NOx levels and compositions, we tested the hypothesis that sample handling introduces artifacts in chemiluminescence assays of NOx. Here, we demonstrate the following: (1) room temperature placement is associated with an increase and decrease in NOx in plasma and whole blood samples, respectively; (2) snap freezing and thawing lead to the interconversion of different NOx in plasma; (3) snap freezing and homogenization in liquid nitrogen eliminate a significant fraction of NOx in the aorta of stressed animals; (4) A "stop solution" commonly used to preserve nitrite and SNOs leads to the interconversion of different NOx in blood, while deproteinization results in a significant increase in detectable NOx; (5) some reagents widely used in sample pretreatments, such as mercury chloride, acid sulfanilamide, N-ethylmaleimide, ferricyanide, and anticoagulant ethylenediaminetetraacetic acid, have unintended effects that destabilize SNO, DNICs, and/or heme-NO; (6) blood, including the residual blood clot left in the washed purge vessel, quenches the signal of nitrite when using ascorbic acid and acetic acid as the purge vessel reagent; and (7) new limitations to the four chemiluminescence-based assays. This study points out the need for re-evaluation of previous chemiluminescence measurements of NOx, and calls for special attention to be paid to sample handling, as it can introduce significant artifacts into NOx assays.

Keywords: DNIC; NO metabolites; Pitfall; S-nitrosothiol; chemiluminescence methodology; heme-NO; nitrite.

Figure 1

Effects of incubation at room temperature on NOx signals. n ≥ 5. Samples were kept in the dark at room temperature and measured using I₃⁻ as the purge vessel reagent at different time points after collection. (A,B) Plot of the time dependence of NOx levels. Data points from each individual animal are connected by lines. Black lines show the linear regression with slopes significantly different from zero. (C,D) Representative NOx signal traces. Two trimmed traces of injections were clipped and lined up on one continuous x-axis for better comparison. Arrows indicate the injections made at different time points after blood sample collection. (A,C) Adult vein plasma (AVP). (B,D) Adult vein whole blood (AVWB). AVP and AVWB were from healthy low-altitude sheep. Between injections, the purge vessel was washed at least once with DI-water and refilled with fresh reaction reagent.

Figure 2

Effects of snap freezing in liquid nitrogen on the NOx signals of plasma from high- and low-altitude sheep. n ≥ 5. NOx were measured using I₃⁻ as the purge vessel reagent. (A,C) Representative NO signal of UVP (A), UAP (B), and AVP (C). Each plasma sample was divided into two aliquots: the first (red) was intact and measured within 4 min of blood collection; the second (blue) one was snap frozen in liquid nitrogen for 30 min and then thawed in hand for measurement within 4 min. All traces (A,C) were from one healthy high-altitude sheep. Arrows represent injections. (D) Summary of the effects of snap freezing on NOx quantities as determined by measurement of the area under the peak. Despite significant qualitative differences in the shapes of NOx peaks between fresh and frozen samples from both high- and low-altitude ewes, no significant difference was detected in the quantities by paired t-tests. UVP: umbilical vein plasma; UAP: umbilical artery plasma; AVP: adult vein plasma.

Figure 3

Effects of methods of homogenization on NOx levels in aorta homogenates. n ≥ 5. Thoracic aortas were homogenized in either liquid nitrogen (L-N₂) or ice-cold HEPES buffer (Ice). NOx were measured by chemiluminescence NO analyzer with different purge vessel reagents. (A) I₃⁻. (B) K₃[Fe(CN)₆]/AcOH. (C) K₃[Fe(CN)₆]/PBS. (D) VitC/AcOH (see the Materials and Methods section or reference #21 for explanation of which NOx metabolites are detected by each reagent). Measurements were normalized to protein concentrations in the homogenates. N.D.: not detectable. Two-way ANOVA with Sidak’s post hoc tests. * for p < 0.05 and ** for p < 0.01 for L-N₂ vs. Ice. Results of the ice-cold HEPES buffer group were recently reported elsewhere [22].

Figure 4

Effects of stop solution and deproteinization on the NOx signals of whole blood samples from high- and low-altitude sheep. n ≥ 5. Samples were measured using I₃⁻ as the purge vessel reagent. (A,C) Representative NOx signal of UVWB (A), UAWB (B), and AVWB (C). Whole blood was divided into three aliquots: the first one (red) was left intact and measured within 4 min of blood collection; the second one (brown) was measured after mixing sample with stop solution at a volume ratio of 5:1; the third one (black) was prepared by mixing the second one with cold methanol at a volume ratio of 1:1 and then centrifugation (deproteinization) at 10,000 rpm for 30 s for supernatant. All traces (A,C) were from one representative high-altitude sheep. Arrows represent injections. Parallel injections of stop solution and methanol per se did not result in any NO signal in the assay. (D) Summary of the effects of stop solution and deproteinization on NOx quantities after correction for dilution factors. Two-way ANOVA. ** for comparison with control; # for comparison with the stop solution. Single symbol for p < 0.05; double symbols (** and ##) for p < 0.01; triple symbols (###) for p < 0.001. p-value in (D) for paired t-test.

Figure 5

Effects of blood on NOx measurements with VitC/AcOH assay, which detects nitrite and heme-NO [21]. n ≥ 5. Samples were measured using VitC/AcOH as the purge vessel reagent. (A) NOx was not detected in whole blood no matter if it was injected before or after plasma, whereas NOx was detected in plasma only when it was injected before whole blood. Between injections, the purge vessel was washed at least once with DI-water and then filled with fresh reaction reagent. However, trace blood clots could still be observed on the wall of the purge vessel. (B) Plasma did not significantly quench the NOx signal of the spiked 0.5 μM of nitrite. (C,D) Representative traces showing that whole blood significantly quenched the NOx signal when spiked with 0.5 μM (C) and 5 μM (D) of nitrite. Inset in (D) is a representative trace of iron-nitrosyl hemoglobin (HbNO). (E) Recovery of NOx from whole blood spiked with 5 μM of nitrite. Nitrite was spiked into HEPES (100%), plasma, or blood, and measured within 4 min. (F) Proposed diagram for the effects of blood on nitrite measurements with VitC/AcOH assay. *: for p < 0.05, **: for p < 0.01, ns for no significant difference, One-way ANOVA versus HEPES. P = plasma, WB = whole blood, UVP = umbilical venous plasma, UAP = umbilical arterial plasma, AVP = adult venous plasma, UVWB = umbilical venous whole blood, UAWB = umbilical artery whole blood, AVWB = adult venous whole blood.