| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
General Clinical Chemistry |
Inorganic Toxicology and Nutrition Branch, Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA.
aAddress correspondence to this author at: Centers for Disease Control and Prevention, 4770 Buford Hwy, Atlanta, GA 30341. Fax 770-488-4139; e-mail CPfeiffer{at}cdc.gov.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We evaluated conditions that influence the complete hemolysis and deconjugation of folate polyglutamates to folate monoglutamates (FMGs) from whole blood (WB). WB samples were hemolyzed in 10 g/L ascorbic acid at various temperatures (room temperature, 30 °C, and 37 °C; n = 15) or hemolysate pH values (pH 4.0, 4.7, 5.2; n = 11) and incubated up to 6 h. FMGs and folate diglutamates (FDGs) were analyzed by liquid chromatography–tandem mass spectrometry (LC/MS/MS) and total folate (TF) by microbiologic assay. We investigated delaying hemolysis by freezing WB for 10 days (n = 20).
Results: Hemolysates frozen immediately after preparation contained 22%–27% FDGs, depending on hemolysate pH. The proportion of FDGs decreased to <3% after incubation at pH 4.7/37 °C for 3 h and did not significantly change on extended incubation up to 5 h. Short-term delayed hemolysis of WB produced results indistinguishable from those of immediate hemolysis. TF results obtained by the microbiologic assay were not different across incubation conditions and agreed with the sum of FMGs and FDGs by LC/MS/MS. The difference between the 2 methods was an insignificant 3% for pH 4.7/37 °C for 3 h.
Conclusions: Hemolysate incubation up to 2 h at 37 °C is not adequate for full polyglutamate deconjugation. We obtained the highest yield of FMGs with lowest FDG concentrations at pH 4.7/37 °C for 3 h. Delaying hemolysis of WB for several days had no negative effect on measurable folate for presumed MTHFR C/C genotype samples.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T polymorphism. Erythrocytes of these individuals have been reported to contain significant fractions of other reduced folate forms stored as long-chain polyglutamates (2)(3)(4). Clinical methods for the determination of erythrocyte folate, such as the microbiologic assay and various immunoassays, measure only total folate (TF) and show considerable variability (5)(6)(7). The Lactobacillus casei microorganism grows equally well on folate monoglutamates (FMGs), folate diglutamates (FDGs), and folate triglutamates, and immunoassays show limited variability with glutamyl chain length. Other chromatography-based methods lack sensitivity (8)(9)(10)(11)(12)(13)(14) or require complex multistep sample preparation and provide only total erythrocyte folate via a derivative of p-aminobenzoic acid, as is the case for gas chromatography–mass spectrometry methods (11)(12)(13)(14). Possible causes of erythrocyte folate variability are incubation pH, temperature, and time (15)(16); trapping of folate within the hemoglobin (Hb) molecule (17); and incomplete hemolysis (18). We recently reported the measurement of serum and whole blood (WB) FMGs by liquid chromatography–tandem mass spectrometry (LC/MS/MS) (4)(19). To confirm that measured FMG concentrations accurately represented the TF concentrations in the erythrocytes, we monitored both FMGs and FDGs by LC/MS/MS to investigate conditions that affect WB hemolysis and folate deconjugation. |
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
preparation of wb pools
To prepare WB quality-control (QC) pools, we collected WB samples into dipotassium EDTA Vacutainer Tubes (Becton Dickinson), mixed the tubes on a rotator for 15 min, and diluted WB 1/11 with 10 g/L ascorbic acid. Hemolysates were incubated at room temperature for 2 h with occasional stirring with a glass rod, and 1-mL aliquots were placed into 2-mL cryovials and frozen within 1 h. The QC pools were stored at –70 °C until analysis and measured in each run in 2 replicates. The low-QC pool contained 5CH3THF; 5-formyltetrahydrofolic acid (5CHOTHF); 5,10CH=THF; and tetrahydrofolic acid (THF) at 33.9, 6.70, 25.5, and 68.2 µg/L of WB, respectively (likely MTHFR T/T genotype). The medium-QC pool contained 5CH3THF and 5CHOTHF at 153 and 18.8 µg/L of WB. The high-QC pool contained 5CH3THF and 5CHOTHF at 258 and 28.8 µg/L of WB. Between-run variability (n = 25 days) was 6%–8% for 5CH3THF, 12%–16% for 5CHOTHF, 13% for 5,10CH=THF, and 13% for THF.
investigation of factors influencing wb hemolysis and folate deconjugation
Pilot study.
Pilot testing with a limited number of samples (n 
5) showed increases in FMGs with increasing incubation time beyond 3 h if hemolysates were prepared at pH 4.0 and incubated at room temperature. In hemolysates incubated at 37 °C, FMGs did not appear to increase beyond 3 h. Pilot testing showed higher initial FMG concentrations with hemolysates at pH 4.7 and 5.2 compared with pH 4.0. However, after a 2-h-incubation at room temperature, the pH 4.0 and pH 4.7 hemolysates appeared to produce slightly higher yields of FMGs than the pH 5.2 hemolysates. To verify whether undiluted WB samples could be stored at 4 °C for up to 1 week before hemolysis, we hemolyzed WB at days 0, 2, 4, and 7 and incubated hemolysates at pH 4.0 and room temperature for 2 h. These hemolysates did not produce results different from those of freshly prepared hemolysates incubated under the same conditions. We also performed pilot tests of the effects of oxygenation on generation of FMGs during hemolysate incubation. Neither TF results by microbiologic assay nor FMG and FDG results by LC/MS/MS were affected by deoxygenation of the ascorbic acid diluent with nitrogen, closed incubation of hemolysates, closed or open incubation with continuous stirring, or closed incubation with continuous rotation. Because our pilot testing revealed significant between-subject variability in FMG generation during sample handling, we included a larger number of individuals (n 
15) in the main studies.
Main study 1: Effect of incubation time, temperature, and pH.
Venous blood samples from 15 on-site volunteers were collected into 7-mL dipotassium EDTA Vacutainer Tubes according to standard procedures with an internal review board–approved human subject protocol. The tubes were mixed on a rotator for 15 min before hemolysis. To study the effect of incubation time and temperature on folate deconjugation, we prepared hemolysates in 2-mL cryovials by adding 100 µL of WB to 1 mL of 10 g/L freshly prepared ascorbic acid solution, pH 2.7 (1/11 dilution), producing a hemolysate pH of 4.0. Hemolysates were incubated at room temperature, 30 °C, and 37 °C for 0, 3, 4, 5, and 6 h each. On the basis of our findings in the pilot study, we did not investigate incubation times shorter than 3 h in this part of the main study. When the FDG concentration was minimized and did not decrease further, we considered the endpoint for optimum deconjugation to have been reached. To study the effect of pH on folate deconjugation, we used freshly prepared ascorbic acid at pH 2.7 (unadjusted, conventional diluent), pH 4.0 (adjusted with 5 mol/L sodium hydroxide), and pH 4.25 (adjusted with 5 mol/L sodium hydroxide) to prepare 1/11-diluted hemolysates with pH values of 4.0, 4.7, and 5.2. Hemolysates were incubated at room temperature for 0, 1, 2, and 3 h and at 37 °C for 3 h. We chose these particular incubation conditions because we were interested in whether a higher hemolysate pH (closer to the pH optimum of plasma 
-carboxypeptidase) would allow us to use shorter incubation times (<3 h) and/or a lower incubation temperature (room temperature vs 37 °C). After incubation was complete, we stored all hemolysates at –70 °C until analysis. Two parallel sets of hemolysates were prepared for all experiments: one to be analyzed by LC/MS/MS for FMGs and FDGs, and one to be analyzed by microbiologic assay for TF.
Main study 2: Effect of delayed WB hemolysis and confirmation of optimum incubation conditions.
We obtained one 7-mL dipotassium EDTA Vacutainer with WB from each of 20 blood bank donors (Tennessee Blood Bank). Samples were transported on ice packs by courier overnight delivery. On arrival, the tubes were mixed on a rotator for 15 min. For evaluation of the effect of delayed hemolysis, undiluted WB from each donor was either frozen at –70 °C in 500-µL aliquots or hemolysates (pH 4.7 and 37 °C for 3 h) were prepared on the day of WB arrival and stored at –70 °C until analysis. After 10 days, frozen WB was thawed and a hemolysate (pH 4.7 and 37 °C for 3 h) was prepared for each donor sample. At the same time, to confirm the optimum hemolysate incubation conditions determined in main study 1 (pH 4.7 and 37 °C for 3 h), we prepared hemolysates from each donor sample incubated at pH 4.7 and 37 °C for 3, 4, and 5 h. After incubation all samples were extracted and analyzed for FMGs and FDGs by LC/MS/MS. When the FDG concentration was minimized and did not decrease further, we considered the endpoint for optimum deconjugation to have been reached.
Side study: Interconversion of different folates during WB hemolysis.
To verify that folates do not interconvert during WB hemolysis, we added 1 folate form at a time to a series of WB hemolysates generated from 1 donor sample at pH 4.0 or pH 4.7: 10 µg/L 5CH3THF, 2.5 µg/L 5CHOTHF, 5 µg/L THF, and 2.5 µg/L 5,10CH=THF. Incubation was at pH 4.0 and 37 °C for 4 h and pH 4.7 and 37 °C for 3 h. FMGs and FDGs were analyzed by LC/MS/MS.
preparation of folate calibrators
All folate stock solutions and calibrators were prepared as described previously (4)(19). To maintain the stability and avoid interconversion of 5,10CH=THF (m/z 456.2m/z 412.2) to 5CHOTHF (m/z 474.2m/z 327.1), we prepared stock solutions in 1 mol/L HCl containing 10 g/L ascorbic acid and stored them at –70 °C. The concentration of this calibrator was determined by ultraviolet absorbance at 345 nm (20). Aliquots of a 10 mg/L stock solution were prepared monthly in 0.5 mol/L HCl containing 1 g/L ascorbic acid. These were stored at –70 °C and used for daily calibration. 5CH3THF diglutamate was prepared identically to 5CH3THF (19), and its concentration was determined by ultraviolet absorbance at 290 nm with the absorptivity for 5CH3THF (20).
preparation of aqueous calibrators and wb samples for solid-phase extraction
Sample preparation and chromatographic and MS/MS conditions were as described previously (4)(19). The pseudo-internal standard for 5,10CH=THF in our previous method (4) was replaced with 13C5-5,10CH=THF (m/z 461.2m/z 416.2), its stable-isotope analog. We prepared a separate 5CH3THF diglutamate calibration curve analogous to the 5CH3THF calibration curve (19), but used 13C5-5CH3THF as the pseudo-internal standard because of the lack of an isotopically labeled 5CH3THF diglutamate. The 5CH3THF diglutamate calibration samples were extracted in the same run with the FMG calibration samples. We did not use any other FDG calibrators in the diglutamate calibration curve. We evaluated recovery and solid-phase extraction (SPE) efficiency of 5CH3THF diglutamate by adding the calibrator solution at concentrations of 0, 2, 5, and 10 µg/L to a hemolysate that was then extracted and analyzed in triplicate; the results were comparable (102% recovery and 71% SPE efficiency) to data obtained previously for FMGs (4)(19).
lc/ms/ms analysis
After SPE, samples were divided into 2 portions and stored at –70 °C until LC/MS/MS analysis of FMGs and FDGs was conducted on 2 consecutive days. As expected, 5CH3THF diglutamate eluted at the same retention time as the monoglutamate. MS/MS experiments using 5CH3THF diglutamate produced a protonated [M + H]+ molecular ion at m/z 589.2 and a predominant fragment ion at m/z 313.2, identical to the fragment ion of 5CH3THF (m/z 460.2m/z 313.2). The product ion scan for 5CH3THF diglutamate is shown in Fig. S1 of the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol51/issue12. We confirmed that no 5CH3THF monoglutamate was recorded in the 5CH3THF diglutamate ion channel and vice versa. Multiple-reaction monitoring (MRM) methods for 5CHOTHF diglutamate and THF diglutamate were created analogously, assuming that both compounds produce fragment ions identical to the ones produced by their respective monoglutamates. The compounds thus lose the diglutamate portion of the molecule as a result of neutral loss (5CHOTHF diglutamate, m/z 603.2m/z 327.2; THF diglutamate, m/z 575.4m/z 299.2). The fragmentation behavior of 5,10CH=THF is different from that of other folate forms in that it loses only a fragment with m/z 44, which corresponds to carbon dioxide. We hypothesized, therefore, that 5,10CH=THF diglutamate will lose 1 glutamate residue in addition to carbon dioxide as a result of neutral loss (5,10CH=THF diglutamate, m/z 585.2m/z 412.2). Because of the lack of stable-isotope–labeled FDGs for use as internal standards, we calculated area ratios, using the respective FMG internal standards, and determined concentrations with the respective FMG calibration curves. The MS/MS method conditions for the FDG analysis are shown in Table 1
. Tandem MRM profiles for various FDGs in the WB low-QC pool (presumably from a T/T individuals) are shown in Fig. S2 of the online Data Supplement. Tandem MRM profiles for various FMGs in the same WB low-QC pool have been shown previously (4).
|
microbiologic assay
The TF content of WB hemolysates subjected to different incubation conditions was determined by microbiologic assay carried out on microtiter plates with the chloramphenicol-resistant strain of L. casei based on the procedures of O’Broin and Kelleher (21) and Molloy and Scott(22). The assay was calibrated using 5CH3THF from Merck Eprova. The variability of this assay over 20 days was 8%–9% for concentrations of 151–330 µg/L of WB.
statistical analysis
The percentage FDG was calculated by dividing the FDG concentration by the sum of the FMG and FDG concentrations. All statistical comparisons were evaluated at a significance level of 
= 0.05.
Main study 1.
The effect of time (0, 3, 4, 5, and 6 h) and temperature (room temperature, 30 °C, and 37 °C) on FMG concentrations and separately on FDG concentrations was studied by a 2-factor ANOVA. We used a Bonferroni t-test of differences to find the shortest time and the lowest temperature that maximized FMG concentrations and minimized FDG concentrations. Analogous analysis was conducted to assess the effect of time (0, 1, 2, and 3 h at room temperature) and pH (4.0, 4.7, and 5.2) on FMG concentrations and separately on FDG concentrations. Complete data sets from 15 individuals were available for the time/temperature analyses. Because of the lack of sufficient blood, complete data sets from only 11 individuals were available for the time/pH analyses. The differences in FMG and FDG concentrations between hemolysates that were prepared at room temperature and 3 h vs 37 °C and 3 h (various pH values) were analyzed by paired t-test. The effect of pH on hemolysates prepared at 37 °C and 3 h was assessed by 1-way ANOVA. Method agreement for FMG + FDG concentrations by LC/MS/MS vs TF by microbiologic assay (pH 4.7 and 37 °C for 3 h) was assessed by Deming regression and Bland–Altman difference analysis. Complete data sets from 15 individuals were available for this analysis, but the data from the 1 presumably T/T individual was excluded.
Main study 2.
The differences in FMG and FDG concentrations between hemolysates that were prepared fresh and those that were delayed in preparation were analyzed by paired t-test. The effect of extended incubation beyond 3 h for hemolysates prepared from frozen WB and incubated at pH 4.7 and 37 °C was assessed by 1-way ANOVA. Complete data sets from 20 individuals were available for these analyses.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
influence of time, temperature, and ph on folate polyglutamate deconjugation
The FMG concentrations and percentage FDGs in WB hemolysates (pH 4.0) incubated under different time/temperature conditions (n = 15 individuals) are presented in Fig. S3 of the online Data Supplement and in Fig. 1
. There were no interactions (2-way ANOVA) between time and temperature for FMG (P = 0.99) or FDG (P = 0.49) concentrations. There was a significant (P <0.001) effect of time on FMG and FDG concentrations. The effect of temperature was significant (P <0.001) for FDG but not for FMG concentrations (P = 0.50). FMG concentrations at 3 h were significantly higher than at 0 h; however, there were no significant differences between FMG concentrations at 3, 4, 5, and 6 h (Fig. S3 of the online Data Supplement). Although we found no significant differences between the 3 temperatures, FMG concentrations were highest in hemolysates incubated at 37 °C for 3 h. Freshly prepared hemolysates that were immediately frozen without any incubation contained 
25%–30% FDGs (Fig. 1
). FDG concentrations at 3 h were significantly lower than at 0 h, and concentrations at 4 h were significantly lower still. However, we observed no significant differences between FDG concentrations at 4, 5, and 6 h. FDG concentrations in hemolysates incubated at 30 °C for 4 h were significantly lower than concentrations in hemolysates incubated at room temperature for 4 h, and concentrations in hemolysates incubated at 37 °C for 4 h were lower still; however, the difference from concentrations in hemolysates incubated at 30 °C the same length of time was not significant. Thus, incubation at 37 °C for 4 h minimized FDG concentrations without having a negative effect on FMG concentrations, leaving only 2% residual FDGs in the WB hemolysate. This condition was therefore considered an endpoint for optimum deconjugation.
|
FMG concentrations in hemolysates incubated at pH 4.0 and 37 °C for 4 h in 14 of the 15 samples consisted mainly of 5CH3THF (90%) and small amounts of 5CHOTHF (10%). These samples were presumably from donors with an MTHFR C/C or C/T genotype. FMGs found in 1 sample from a presumably T/T donor consisted of 35% 5CH3THF, 11% 5CHOTHF, 19% 5,10CH=THF, and 35% THF.
The FMG concentrations and the percentage FDGs in WB hemolysates incubated under different time/temperature/pH conditions (n = 11 individuals) are presented in Fig. S4a of the online Data Supplement and in Fig. 2A
. There was no interaction (2-way ANOVA) between time and pH for room temperature conditions for FMG (P = 0.30) or FDG (P = 0.88) concentrations. There was a significant (P <0.001) effect of time on FMG and FDG concentrations. The effect of pH was significant (P <0.01) for FMG concentrations, but not for FDGs (P = 0.18). FMG concentrations were highest at 3 h and were significantly higher than at 0 and 1 h, but not significantly higher than at 2 h (Fig. S4a of the online Data Supplement). Freshly prepared hemolysates that were immediately frozen without any incubation contained 20%–25% FDGs, depending on hemolysate pH (Fig. 2A
). FDG concentrations were lowest at 3 h, but we found significant differences only between 2 and 0 h, 3 and 0 h, and 3 and 1 h. FMG concentrations were significantly higher and FDG concentrations were significantly lower in hemolysates incubated at 37 °C for 3 h compared with those incubated at room temperature for 3 h for each pH value (2-way ANOVA for temperature and pH). Although there were no significant differences between the hemolysates incubated at different pHs and 37 °C for 3 h, pH 4.7 hemolysates yielded the highest FMG concentrations (1-way ANOVA). Respective analyses showed significant differences in FDG concentrations in hemolysates incubated at various pH values and 37 °C for 3 h; pH 4.7 hemolysates yielded the lowest FDG concentrations, and these concentrations were significantly different from those from pH 4.0 but not from pH 5.2 hemolysates.
|
The FMG forms of the samples from presumably T/T donors were dependent on the hemolysis pH (37 °C for 3 h). These forms were nearly identical for pH 4.0 (25% 5CH3THF, 8% 5CHOTHF, 19% 5,10CH=THF, and 48% THF) and pH 4.7 (24% 5CH3THF, 8% 5CHOTHF, 21% 5,10CH=THF, and 47% THF). However, at pH 5.2, the percentage of 5,10CH=THF increased at the expense of THF (28% 5CH3THF, 11% 5CHOTHF, 35% 5,10CH=THF, and 26% THF).
In general, the microbiologic assay showed no difference in response to incubation time, temperature, or pH (see Table S1 of the online Data Supplement). The sum of FMG and FDG concentrations determined by LC/MS/MS agreed well with TF determined by microbiologic assay (Table S1 of the online Data Supplement). The biggest difference (<10%) between the 2 assays was at 0 h. This result could be attributable to residual folate triglutamates that were measured by the microbiologic assay but not by the LC/MS/MS method. We should note that although we handled all samples in an expeditious and systematic way, it took 15 min (for sample pipetting/boxing) before hemolysates were placed at –70 °C, and it took 15 min for the temperature in the hemolysate to drop below a reasonable enzyme activity. Therefore, the 0-h samples were subjected to some deconjugation by the endogenous plasma -carboxypeptidase. We compared the 2 methods by use of either FMG concentrations or the sum of FMG and FDG concentrations detected by LC/MS/MS for hemolysates incubated at pH 4.7 and 37 °C for 3 h. The 2 methods were highly correlated and in agreement [Pearson correlation, r = 0.91 for FMGs and r = 0.92 for FMGs + FDGs; Deming regression, slope (95% confidence interval), 1.13 (0.79–1.47) for FMGs and 1.08 (0.79–1.38) for FMGs + FDGs; intercept, –17.00 (–95.28 to 61.28) µg/L for FMGs and –10.92 (–80.51 to 58.67) µg/L for FMGs + FDGs; difference, 13.26 (–0.52 to 27.05) µg/L for FMGs and 8.21 (–4.50 to 20.93) µg/L for FMGs + FDGs]. The LC/MS/MS FMG results were, on average, an insignificant 5% lower than microbiologic assay results. LC/MS/MS FMG + FDG results were only 3% lower.
effect of delayed hemolysis of wb and confirmation of optimum incubation conditions
We found only 5CH3TH F and 5CHOTHF in the samples from all 20 donors from main study 2. We therefore presumed that the MTHFR genotype of these donors was C/C or C/T. If undiluted WB was stored at –70 °C (n = 20 donor samples) for 10 days before hemolysis (pH 4.7 and 37 °C for 3 h), paired t-test revealed that FMG concentrations were indistinguishable from those in freshly prepared hemolysates [mean (95% confidence interval), 168 (137–199) vs 171 (138–203) µg/L; P = 0.15]. Although we found a significant difference in FDG concentrations between delayed vs freshly prepared hemolysates [mean (95% confidence interval), 1 (0.4–1.6) vs 6 (3.4–8.6) µg/L; P <0.001], the concentrations were so low that they did not significantly alter the sum of FMG and FDG concentrations. Delayed hemolysis produced hemolysates that contained, on average, only 1% FDGs after incubation at 37 °C for 3h, even less than the 3% found in freshly prepared hemolysates. Extended incubation of delayed hemolysates beyond 3 h at pH 4.7 and 37 °C (1-way ANOVA) did not further increase the yield of FMGs (Fig. S4b of the online Data Supplement), nor did it decrease the FDG concentrations any further (Fig. 2b
of the online Data Supplement). We therefore confirmed that the incubation conditions pH 4.7 and 37 °C for 3 h were an endpoint for optimum deconjugation for WB that contains 5CH3THF and 5CHOTHF and potentially also for WB that contains other folate forms. However, these endpoints will have to be confirmed by use of WB from several presumably T/T individuals.
interconversion of different folates during wb hemolysis
We confirmed that measurable folates generally do not interconvert during WB hemolysis at pH 4.0 and 37 °C for 4 h or pH 4.7 and 37 °C for 3 h. When we added 1 folate form at a time to the WB hemolysate, we found an increased concentration for the added folate only, with the exceptions of 5CHOTHF and 5,10CH=THF. We have shown previously (4) that there is some interconversion between these 2 folate forms. It is also known that there are further interconversions among different folate formyl forms.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the current report we have expanded on the investigation of factors influencing folate extraction from erythrocytes, and we used a larger number of donor samples than previous reports. Furthermore, we incorporated the measurement of FDGs by LC/MS/MS to monitor the progression of folate deconjugation, an entirely new approach that has not been previously reported.
Our studies showed that incubation of WB hemolysates at pH 4.0 (using the conventional diluent of 10 g/L ascorbic acid without pH adjustment) and 37 °C for 3 h does not provide complete deconjugation to FMGs (5% residual FDGs). Thus, previous recommendations to incubate WB hemolysates (pH 4.0) at 37 °C for 90 min (16) are not adequate if chromatography-based assays are used to measure FMGs. Less specific assays, such as the microbiologic assay and various immunoassays that measure short-chain folate polyglutamates in addition to FMGs, are not necessarily expected to produce different results for various incubation conditions. We confirmed this with the microbiologic assay in the present study.
We obtained optimum folate extraction for WB hemolysates prepared at pH 4.0 and 37 °C for 4 h; this produced the highest yield of FMGs with the smallest residual amount of FDGs (<3%). The same results were obtained for hemolysates prepared at pH 4.7 and 37 °C for 3 h. Because the latter condition takes less time, it is more efficient and is therefore preferable. Under both conditions, the sum of FMGs and FDGs detected by LC/MS/MS and the TF measured by microbiologic assay were in agreement. Hemolysate pH values of 4.0 and 4.7 led to only minimal interconversion of folates during WB hemolysis. At a hemolysate pH of 5.2, however, we found a different folate pattern in the presumably T/T donor than at the other 2 pH values (higher amounts of 5,10CH=THF and 5CHOTHF compared with lower amounts of THF). This finding indicates that more significant folate interconversions may take place at this higher pH. Because some interconversions are unavoidable during the process of folate extraction, it will be important to group folate results into methylated and nonmethylated folate as suggested previously (2)(3).
To facilitate coordination between field and laboratory work, we investigated whether delayed hemolysis of WB is acceptable. We found no negative effect on extractable folate if undiluted WB was stored for up to 1 week at 4 °C (pilot study) before hemolysis. Furthermore, we obtained the same results for undiluted WB that was frozen at –70 °C and hemolyzed 10 days later (main study 2). It is interesting to note that at 0 h we found only 14% residual FDGs in those hemolysates and that after incubation at 37 °C for 3 h the residual amount of FDGs was reduced to only 1% compared with 3% in freshly hemolyzed samples. This might be attributable to some folate deconjugation occurring as a result of freezing/thawing. It thus appears feasible and preferable to freeze WB in the field and perform hemolysis at a later time when the laboratory is ready for analysis.
Full deconjugation of FDGs to FMGs is not only necessary to obtain accurate estimates by chromatography-based assays, but it will also help minimize assay variability. We observed slightly higher assay variabilities for our WB bench QC pools than what we typically reported for serum (19). This difference could be attributable to daily fluctuations of the residual amount of FDGs (12%). We observed higher between-subject variation during the first few hours of hemolysis compared with incubation at 37 °C for 3 h. This could be indicative of individual differences in plasma -carboxypeptidase activity.
Wright et al. (18) suggested the optimum pH for hemolysates to be 5.0, because this acidic pH does not induce Hb denaturation (which occurs irreversibly at pH <4.7), but maximizes both plasma -carboxypeptidase activity and Hb dimerization. Our own findings by LC/MS/MS showed that indeed pH 4.7 gave optimum FMG generation (faster than at pH 4.0). However, at pH 5.2 we obtained slightly lower results for hemolysates incubated at 37 °C for 3 h (6%) than with the microbiologic assay. This result could indicate that the hemolysates contained small residual amounts of folate triglutamates that are captured by the microbiologic assay but not by the LC/MS/MS method. Because we obtained the same assayable folate for hemolysates produced at pH 4.0 and 4.7 as long as we extended the incubation time at the lower pH from 3 h to 4 h, Hb denaturation for hemolysates at pH 4.0 seems unlikely.
Wright et al. (17) hypothesized that folate bound to deoxy-Hb might get trapped as deoxy-Hb picks up oxygen and switches quaternary structure to oxy-Hb. This hypothesis suggests that erythrocyte folate concentrations might be significantly underassayed by current extraction methods. The authors suggested that strategies would need to be developed that allow erythrocyte folate to be deconjugated while all Hb is in the deoxy-Hb state. Although our main study did not address the question of folate trapping, we evaluated in a pilot study whether use of an ascorbic acid diluent previously degassed with nitrogen or stirring the hemolysate for 1 h in an open container exposed to air would affect assayable folate by either LC/MS/MS or the microbiologic assay. We did not find any effects compared with our comparison condition, in which we used a diluent that was not degassed and in which we kept hemolysates incubated for 1 h in a closed container. Our results therefore suggest that these small variations in laboratory practice do not alter final results.
Although this report presents optimum conditions for folate deconjugation in WB hemolysates, certain limitations apply. Because only 1 presumably T/T donor was part of the current investigation, the suitability of the recommended deconjugation conditions for WB containing significant concentrations of nonmethylated folates (5CHOTHF, 5,10CH=THF, and THF) is preliminary until more analyses have been conducted. In addition, we cannot comment on the agreement between the LC/MS/MS method and the microbiologic assay for T/T donor samples. Initial results seem to indicate that the LC/MS/MS method might produce slightly higher results for T/T donor samples. Delayed hemolysis for samples from persons with the T/T genotype will have to be investigated to confirm whether the same deconjugation conditions can be applied as for samples from persons with the C/C genotype. Although the current investigation used a considerably larger sample size than any previous report, it was not large enough to provide sufficient power to detect very small differences. Therefore, some of the analyses that did not reach significance in this report might reach significance with a larger sample.
In conclusion, this is the first report that describes in detail the influence of pH, temperature, and time on folate deconjugation, as determined by LC/MS/MS for the simultaneous analysis of FMGs and FDGs and by comparing the results with TF values obtained by microbiologic assay. We found that longer incubation times than previously proposed are needed to obtain full deconjugation to FMGs. Similar studies would have to be conducted to validate the measurement of FMGs by chromatography-based assays in matrices other than WB.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  Z. Fazili, C. M. Pfeiffer, M. Zhang, R. B. Jain, and D. Koontz Influence of 5,10-Methylenetetrahydrofolate Reductase Polymorphism on Whole-Blood Folate Concentrations Measured by LC-MS/MS, Microbiologic Assay, and Bio-Rad Radioassay Clin. Chem., January 1, 2008; 54(1): 197 - 201. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Welzel, H. A. Katki, L. C. Sakoda, A. A. Evans, W. T. London, G. Chen, S. O'Broin, F.-M. Shen, W.-Y. Lin, and K. A. McGlynn Blood Folate Levels and Risk of Liver Damage and Hepatocellular Carcinoma in a Prospective High-Risk Cohort Cancer Epidemiol. Biomarkers Prev., June 1, 2007; 16(6): 1279 - 1282. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, room temperature; 
, 30 °C; 
, 37 °C.