Quantitation of folates and their catabolites in blood plasma, erythrocytes, and urine by stable isotope dilution assays

Article abstract of DOI:10.1016/j.ab.2009.11.007

New stable isotope dilution assays were developed for the simultaneous quantitation of the folates 5-methyltetrahydrofolic acid, 5-formyltetrahydrofolic acid, tetrahydrofolic acid, 10-formylfolic acid, and folic acid as well as for their catabolites para-aminobenzoylglutamate (pABG) and acetyl-para-aminobenzoylglutamate (ApABG) in clinical samples. The methods were based on cleanup by strong anion exchange followed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) detection. Deuterated analogues of the folates and [¹³C₅]-labeled isotopologues of the catabolites were applied as the internal standards in stable isotope dilution assays. Extraction in 4-morpholineethanesulfonic acid (MES) buffer at pH 5.0 ensured the optimum stability of folates and, in combination with solid-phase extraction (SPE) based on strong anion exchange, resulted in higher recoveries compared with other combinations of extraction buffers and SPE. The method was sensitive enough to detect pABG in plasma generally and unmetabolized folic acid in the plasma of a volunteer after oral dosage of an aqueous folic acid solution. The sum of folate catabolites increased by a factor of 2 in the urine of the latter volunteer, compared with that resulting when only water was dosed.

Full text of DOI:10.1016/j.ab.2009.11.007

Analytical Biochemistry 398 (2010) 150–160
Contents lists available at ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
Quantitation of folates and their catabolites in blood plasma, erythrocytes,
and urine by stable isotope dilution assays
Sabine Mönch a, Michael Netzel b,1, Gabriele Netzel b,1, Michael Rychlik a,
*
a Lehrstuhl für Lebensmittelchemie, Technische Universität München, D-85748 Garching, Germany
b Department of Human Nutrition, Institute of Nutrition, Friedrich Schiller Universität Jena, D-07743 Jena, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
New stable isotope dilution assays were developed for the simultaneous quantitation of the folates 5-
methyltetrahydrofolic acid, 5-formyltetrahydrofolic acid, tetrahydrofolic acid, 10-formylfolic acid, and
folic acid as well as for their catabolites para-aminobenzoylglutamate (pABG) and acetyl-para-amin-
obenzoylglutamate (ApABG) in clinical samples. The methods were based on cleanup by strong anion
exchange followed by liquid chromatography–tandem mass spectrometry (LC–MS/MS) detection. Deu-
terated analogues of the folates and [13C5]-labeled isotopologues of the catabolites were applied as the
internal standards in stable isotope dilution assays. Extraction in 4-morpholineethanesulfonic acid
(MES) buffer at pH 5.0 ensured the optimum stability of folates and, in combination with solid-phase
extraction (SPE) based on strong anion exchange, resulted in higher recoveries compared with other com-
binations of extraction buffers and SPE. The method was sensitive enough to detect pABG in plasma gen-
erally and unmetabolized folic acid in the plasma of a volunteer after oral dosage of an aqueous folic acid
solution. The sum of folate catabolites increased by a factor of 2 in the urine of the latter volunteer, com-
pared with that resulting when only water was dosed.
Received 17 February 2009
Received in revised form 2 November 2009
Accepted 3 November 2009
Available online 10 November 2009
Keywords:
Folates
Folate catabolites
LC–MS/MS
Plasma
Red blood cells
Stable isotope dilution assay
Urine
Ó 2009 Elsevier Inc. All rights reserved.
In those parts of the world where folate fortification is not
mandatory, intake of this vitamin group is considered to be below
the human dietary recommendations. Because folates play a cru-
cial role as coenzymes in the metabolism of one-carbon groups
[1], folate deficiency is believed to increase the risk of neural tube
defects [2] and is suspected of being associated with the develop-
ment of certain forms of cancer [3], Alzheimer’s disease [4], and
cardiovascular disease [5]. Therefore, accurate methods of folate
analysis, either in foods or in clinical samples, are essential to (i)
monitor folate intake, (ii) assess the human organism’s response
to folate dosage, and (iii) evaluate the individual’s folate status.
For example, short-term studies on folate analysis in blood plasma
after folate consumption can be used to assess folate bioavailabil-
ity [6]. In contrast, folate analysis in erythrocytes reveals the mean
folate status over the past 3 months because red blood cells main-
tain their folate contents during their lifetime of approximately
90 days. Moreover, in recent studies, a correlation between folate
vitamer distribution in erythrocytes and polymorphism of the
enzyme methylenetetrahydrofolate reductase (MTHFR)2 [7,8] has
been found. This enzyme is essential for transferring 5,10-methyl-
enetetrahydrofolate to 5-methyltetrahydrofolate, and the TT geno-
type was shown to express
a thermolabile and less effective
enzyme. For individuals with this mutation, a higher portion of non-
methylated tetrahydrofolate vitamers was found in erythrocytes [9].
Particularly in the case of an individual with inadequate folate sta-
tus, the TT mutation was reportedly associated with a twofold risk
for the birth defect spina bifida [10]. In contrast, less nonmethylated
tetrahydrofolates were present in the blood of individuals with the
CC genotype that codes for the stable MTHFR CC enzyme [9]. There-
fore, the vitamer distribution may be an important indicator for the
genotype and subsequent risk associated with insufficient folate
intake.
Besides erythrocyte folates, the urinary level of the folate
catabolites para-aminobenzoylglutamate (pABG) and acetyl-para-
2
Abbreviations used: MTHFR, methylenetetrahydrofolate reductase; pABG, para-
aminobenzoylglutamate; ApABG, acetyl-para-aminobenzoylglutamate; GC, gas chro-
matography; MS, mass spectrometry; LC, liquid chromatography; UV, ultraviolet; MS/
MS, tandem mass spectrometry; DCC, N,N0-dicyclohexylcarbodiimide; DIPEA, N,N0-
diisopropylethylamine; HOBt, 1-hydroxybenzotriazole hydrate; MES, 4-morpholinee-
thanesulfonic acid; BMI, body mass index; SPE, solid-phase extraction; LOD, limit of
detection; LOQ, limit of quantitation; HPLC, high-performance liquid chromatogra-
phy; SRM, selected reaction monitoring; UV–VIS, UV–visible; ESI, electrospray
ionization; CV, coefficient of variation; H4folate, tetrahydrofolate; PteGlu, folic acid.
* Corresponding author. Present address: Division Bioanalytik Weihenstephan,
Research Center for Nutrition and Food Sciences (ZIEL), Technische Universität
München, D-85350 Freising, Germany. Fax: +49 8161 71 4216.
E-mail address: michael.rychlik@wzw.tum.de (M. Rychlik).
Present address: CSIRO Food and Nutritional Sciences, Cannon Hill, Queensland
1
4170, Australia.
0003-2697/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2009.11.007

Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
151
aminobenzoylglutamate (ApABG) has been suggested as an indica-
Human study
tor for folate status because it was calculated to be proportional to
the total body folate pool [11] and found not to respond to short-
time variations of folate intake [12]. However, until recently, as-
says of the folate’s catabolites could only quantitate the sum of
pABG and ApABG, and they require derivatization for gas chroma-
tography–mass spectrometry (GC–MS) [13] or cleanup and deriva-
tization for liquid chromatography–ultraviolet (LC–UV) [14]. On
the other hand, a recent LC–tandem mass spectrometry (MS/MS)
technology is restricted to the quantitation of the catabolites and
does not include the analysis of folates [15].
Ethical permission for the pilot study described below was ob-
tained from the Ethics Committee of the Friedrich Schiller Univer-
sity of Jena (Faculty of Medicine, code 1415-09/04). Urine and
blood samples were taken from a healthy nonsmoking male volun-
teer (24 years old, body mass index [BMI] = 21 kg/m2). Two weeks
prior to the consumption of the first test solution, the tissues of the
volunteer were saturated with folic acid (800 lg/day) by dietary
supplementation. To allow stabilization of the plasma level of fo-
late, the supplementation was stopped 2 days before each of the
In general, folate analysis is quite demanding due to the high
variety of vitamers, their occurrence in only trace amounts, and
their lability toward light and oxygen. In several applications, sta-
ble isotope dilution assays using stable isotope-labeled folates and
LC–MS/MS detection have proven their superiority over conven-
tional assay methods. Although folate vitamers have been quanti-
fied in several foods [16], blood serum [17], and erythrocytes [18],
a stable isotope dilution assay has not yet been reported for folates
and their catabolites in urine.
In continuation of our stable isotope dilution assay for plasma
folates [6], the aims of the current study were (i) to improve the
sensitivity and precision of our plasma folate assay, (ii) to develop
respective assays for erythrocytes, and (iii) to design stable isotope
dilution assays for quantifying folates and their catabolites pABG
and ApABG in urine.
study days. On the first trial day, the volunteer consumed 400 lg
of synthetic folic acid dissolved in 100 ml of water (solution A).
Solution B, consisting solely of 100 ml of water, was consumed
by the volunteer 3 months later (the supplementation procedure
with folic acid was the same as for solution A). After application,
the volunteer was allowed to eat and drink foods with low folate
contents (i.e., rye bread, butter, applesauce, jam, and tap water)
according to a standardized protocol. Blood samples were taken
predose (t = 0 h) as well as 1, 2, 3, 4, 5, 6, 8, 10, 12, and 24 h after
consumption and were collected in Vacutainers containing K3EDTA
as an anticoagulant (Sarstedt, Nürnbrecht, Germany). Plasma and
red blood cells were obtained by centrifuging the blood for
10 min at 2000g and 4 °C. In addition, the volunteer collected the
complete postdose urine for 24 h.
Sample preparation
Materials and methods
Plasma
Plasma samples (400 l
l) were spiked with [2H4]5-methyltetra-
Chemicals
hydrofolic acid (5 ng), [2H4]5-formyltetrahydrofolic acid (5 ng),
[2H4]tetrahydrofolic acid (10 ng), [2H4]10-formylfolic acid (20 ng),
The following chemicals were obtained commercially from the
sources given in parentheses: rat serum (Biozol, Eching, Germany);
chicken pancreas (Difco, Sparks, MD, USA); acetonitrile, formic acid,
hexane, sodium hydroxide, methanol, Na2SO4, sodium chloride, and
[2H4]folic acid (5 ng), [13C5]4-aminobenzoyl-
L
-glutamate (3 ng),
-glutamate (5 ng) and then
l of extraction buffer and equilibrated for
and [13C5]N-acetyl-4-aminobenzoyl-
overlaid with 600
L
l
tetrahydrofuran (THF, Merck, Darmstadt, Germany);
a-amylase,
30 min at room temperature. Subsequently, the plasma buffer
mixture was subjected to solid-phase extraction (SPE) on SAX
cartridges as described below.
For comparing the developed method with a phenyl SPE meth-
od similar to that described by Pfeiffer and coworkers [17], aliquots
of plasma were also extracted with ammonium formate buffer and
subjected to cleanup on phenyl SPE cartridges (Discovery DSC-ph,
100 mg, 1 ml, Varian, Darmstadt, Germany). For better comparabil-
ity of the methods, folates were eluted from both SPE columns with
0.5 ml of elution solution.
To determine the limits of detection (LODs), limits of quantita-
tion (LOQs), and recoveries, a recombinate of human plasma was
developed. This synthetic plasma consisted of lyophilized egg
white (0.07%) as the protein component and sunflower oil
(0.06%) in a physiological solution of aqueous NaCl (0.9%).
ascorbic acid, N,N0-dicyclohexylcarbodiimide (DCC), N,N0-diisopro-
pylethylamine (DIPEA), [13C5]glutamic acid, 1-hydroxybenzotria-
zole hydrate (HOBt), 4-morpholineethanesulfonic acid (MES),
2-mercaptoethanol, protease type XIV, sodium acetate, thionyl chlo-
ride, and trifluoroacetic anhydride (Sigma, Deisenhofen, Germany).
[2H4]5-Methyltetrahydrofolic acid, [2H4]5-formyltetrahydrofol-
ic acid, [2H4]tetrahydrofolic acid, [2H4]10-formylfolic acid, and
[2H4]folic acid were synthesized as reported recently [19].
Syntheses of unlabeled and [13C5]-labeled 4-aminobenzoyl-
glutamate and N-acetyl-4-aminobenzoyl- -glutamate are de-
L-
L
scribed in detail in the Supplementary material.
Solutions
Extraction buffer consisted of aqueous MES (200 mmol/L) con-
taining ascorbic acid (20 g/L) and 2-mercaptoethanol (200 mmol/
L) adjusted to pH 5.0. The buffer was prepared on the day of use.
Ammonium formate buffer consisted of ammonium formate
(10 g/L) and ascorbic acid (1 g/L) adjusted to pH 3.2.
Phosphate buffer (100 mmol/L) was prepared by dissolving so-
dium dihydrogen phosphate (100 mmol) and adjusting the solu-
tion with dipotassium hydrogen phosphate (100 mmol) in water
(1 L) to pH 7.0.
Red blood cells
Red blood cells were lyophilized. Aliquots (40 mg, equivalent to
ꢀ100
ll) were taken from the resulting powder and spiked with
[2H4]5-methyltetrahydrofolic acid (50 ng), [2H4]5-formyltetrahy-
drofolic acid (25 ng), [2H4]tetrahydrofolic acid (75 ng), [2H4]10-
formylfolic acid (50 ng), and [2H4]folic acid (50 ng). The spiked
powder was then overlaid with 2 ml of extraction buffer and
digested with protease (50 lg/mg lyophilized red blood cells) for
Eluting solution was a mixture of aqueous sodium chloride (5%)
and aqueous sodium acetate (100 mmol/L) containing ascorbic
acid (1%).
4 h at 37 °C while being constantly agitated. After enzyme digestion,
thesampleswere heated at 100 °C for 10 min, cooledon ice, and then
spiked with rat serum (150 ll) and chicken pancreas suspension
Chicken pancreas suspension was prepared by stirring chicken
pancreas (5 mg) in diluted aqueous phosphate buffer solution
(30 ml, 10 mmol/L) containing 1% ascorbic acid and adjusted to
pH 7.0.
(2 ml). After deconjugase treatment of the samples at 37 °C and
constant stirring overnight, the samples were heated at 100 °C for
10 minandthencentrifugedat16,100gfor15 minat 4 °C. Afterpass-
ing the supernatant through a syringe filter (0.45 lm, Millipore,

152
Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
Bedford, MA, USA), the filtrates were subjected to SPE cleanup as de-
scribed below.
The need for enzymatic proteolytic degradation was deter-
mined by extracting aliquots of lyophilized red blood cells with
Nucleosil C-18 reversed-phase column (250 ꢁ 3 mm, 4
lm, Phe-
nomenex, Aschaffenburg, Germany) connected to a diode array
detector and a triple quadrupole TSQ Quantum Discovery mass
spectrometer (Thermo Electron).
and without protease (50
To determine LODs, LOQs, and recoveries, lyophilized egg white
(30 mg) was used as a blank matrix.
l
g/mg lyophilized red blood cells).
The mobile phase consisted of a variable mixture of 0.1% aque-
ous formic acid (eluent A) and acetonitrile containing 0.1% formic
acid (eluent B) at a flow rate of 0.3 ml/min. The gradient elution
started at 0% B, followed by raising the concentration of B linearly
to 10% within 2 min and to 25% within a further 23 min. Subse-
quently, the mobile phase was programmed to 100% B within a fur-
ther 2 min and held at 100% B for 3 min before equilibrating the
column for 14 min with the initial mixture.
During the first 11 min of the gradient program, the column
effluent was diverted to waste. The spectrometer was operated in
the positive electrospray mode using selected reaction monitoring
(SRM). The spray voltage was set to 3900 V, the capillary tempera-
ture was set to 320 °C, and the capillary voltage was set to 35 V.
Urine
Urine samples (2 g) were spiked with [2H4]5-methyltetrahy-
drofolic acid (5 ng), [2H4]5-formyltetrahydrofolic acid (15 ng),
[2H4]tetrahydrofolic acid (15 ng), [2H4]10-formylfolic acid (60 ng),
[2H4]folic acid (10 ng), [13C5]4-aminobenzoyl-
L
-glutamate (5 ng),
and [13C5]N-acetyl-4-aminobenzoyl-
L-glutamate (200 ng) and then
overlaid with 4 ml of extraction buffer and equilibrated for 30 min
at room temperature. The solutions were subjected to SPE as
described below.
To exclude the presence of folates and catabolites in human ur-
ine bound to proteins, an aliquot of urine was spiked with the iso-
topologues of folates and catabolites and then overlaid with
extraction buffer and extracted for 4 h at 37 °C. Another aliquot
was spiked with the isotopologues of folates and catabolites,
digested with protease (3 mg), overlaid with extraction buffer,
and extracted for 4 h at 37 °C. The solutions were subjected to
SPE as described below.
For determining LODs, LOQs, and recoveries, a surrogate of
human urine was prepared. This synthetic urine consisted of
lyophilized egg white (0.005%) as the protein component, uric acid
(0.08%), creatinine (0.1%), and urea (2.4%) in a physiological solu-
tion of aqueous NaCl (0.9%) and was used as the blank matrix.
UV spectroscopy
For the determination of extinction coefficients, crystalline
ApABG and pABG (ꢀ1 mg) were weighed into 5-ml volumetric
flasks and dissolved in acetonitrile. Respective dilutions of 100,
40, 10, 5, and 3 lg/ml were prepared with phosphate buffer
(100 mM, pH 7.0). The purity of ApABG and pABG was checked
by full-scan UV–visible (UV–VIS) spectra as well as by LC–MS anal-
ysis. The absorptions of the respective dilutions were recorded by a
UV spectrometer (UV-2401 PC, Shimadzu, Kyoto, Japan) at the
absorption maxima for pABG at kmax = 275 nm and for ApABG at
kmax = 266 nm. The extinction coefficients were calculated using
the following equation:
ml/( mol cm)).
e
= (absorption ꢁ 1000)/concentration (in
Sample cleanup by SPE
l
Extracts were purified by SPE using a 12-port vacuum manifold
(Merck) for red blood cells equipped with Discovery SAX cartridges
(quaternary amine, 500 mg, 3 ml, Sigma) and for plasma and urine
samples equipped with Bond-Elut SAX cartridges (quaternary
amine, 100 mg, 1 ml, Varian). The cartridges were successively
activated with 2 volumes of hexane, methanol, and diluted aque-
ous phosphate buffer (10 mmol/L adjusted to pH 7.0 containing
0.2% 2-mercaptoethanol).
After application of plasma and red blood cell sample extracts,
the columns were washed with 2 volumes of diluted aqueous
phosphate buffer. For urine samples, the cartridges were washed
with 5 volumes of diluted aqueous phosphate buffer. Subse-
quently, the cartridges were dried by vacuum suction and the
folates were eluted with either 0.5 ml (plasma and urine samples)
or 2 ml (red blood cell samples) of eluting solution.
Calibration and quantitation
Solutions of deuterated folates and [13C5]-labeled metabolites
as standards (S) were mixed with the respective analytes (A) in
10 molar ratios [n(S)/n(A)] between 0:10 to 10:1 and 0:20 to 5:1
for H4folate, 5-formylH4folate, PteGlu, 10-formyl-PteGlu, 5-meth-
ylH4folate, pABG, and ApABG, respectively, and were diluted with
water to obtain a total concentration of 0.1 lg/ml (sum of labeled
and unlabeled compounds) before LC–MS/MS analysis. All concen-
trations of solutions containing labeled standards or unlabeled
analytes were verified by UV spectroscopy. After mixing, the solu-
tions were measured by LC–MS/MS and peak area ratios [A(S)/A(A)]
were determined. Calibration functions by using all n(S)/n(A) val-
ues for each standard/analyte combination were calculated from
the obtained A(S)/A(A) ratio using simple linear regression. Stabil-
ity of response was regularly checked by measuring a randomly
chosen n(S)/n(A) value in the linear range.
Absolute recoveries of the analytes from SPE without compen-
sation by isotopologic standards were determined as follows. Solu-
tions of unlabeled folates (130 ng each) and catabolites (90 ng
each) in the respective buffers were first mixed with [2H4]folic acid
(10 ng) and measured by LC–MS/MS. The ratios RBCU of analytes to
[2H4]folic acid were set to 100%. In a second assay, the unlabeled
compounds in the respective buffer were purified over SPE, and
[2H4]folic acid (10 ng) was added to wash solutions and eluates
and measured by LC–MS/MS. The resulting ratios RACU of the sec-
ond assay were compared with RBCU, and the absolute SPE recovery
of each analyte from cleanup without compensation by labeled
LODs and LOQs
LODs and LOQs for folates and their metabolites were deter-
mined according to Ref. [20]. Blank matrices for each kind of sam-
ple were prepared as detailed before. LC–MS/MS analysis
confirmed that the blank matrices contained neither folates nor
their metabolites. For determination of LODs and LOQs, the matri-
ces were spiked (each in triplicate) with the analytes at four differ-
ent concentration levels starting slightly above the LOD and
covering 2 orders of concentration magnitude. After the addition
of the respective labeled internal standards, all samples underwent
sample preparation and cleanup as described above and were final-
ly analyzed by LC–MS/MS. LODs and LOQs were derived statisti-
cally from the data according to a published method [20].
standards was equivalent to RACU/RBCU
.
LC–MS/MS
The samples (10 ll) were chromatographed on a Finnigan
Surveyor Plus HPLC (high-performance liquid chromatography)
System (Thermo Electron, Waltham, MA, USA) equipped with a

Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
153
Precision
Interassay precision was determined by analyzing samples of
during the MS/MS transition and showed the same productions as
the unlabeled analogues. In Fig. 2, the MS/MS spectra display for
pABG and [13C5]pABG the precursor ions at m/z 267 and 272, respec-
tively, along with the product ion at m/z 120. For ApABG and
plasma, erythrocytes, and urine three times in triplicate for 4 weeks.
[
13C5]ApABG, the precursor ions were m/z 309 and 314, respectively,
with the identical product ion occurring at m/z 162.
Recoveries of stable isotope dilution assays
Blank matrices of plasma (400 ll) and urine (2 g) were spiked
Calibration for stable isotope dilution assays
(each in triplicate) with three different amounts of H4folate
(0.3–3.3 pmol), 5-methylH4folate (0.2–1.6 pmol), 5-formylH4folate
(1.5–15 pmol), PteGlu (0.2–2.2 pmol), 10-formyl-PteGlu (0.8–
8 pmol), pABG (0.02–0.15 pmol), and ApABG (0.1–1.2 pmol) and
were analyzed by stable isotope dilution assay. For erythrocytes,
the blank matrix (30 mg) was spiked with three different amounts
of H4folate (1.0–10 pmol), 5-methylH4folate (0.6–7 pmol), 5-form-
ylH4folate (5–50 pmol), PteGlu (2.5–30 pmol), and 10-formyl-Pte-
Glu (1.5–15 pmol). The recovery was calculated as the mean of
the addition experiments.
Calibration curves for folates (MS/MS parameter given in
Table 1) and their catabolites were linear in molar ratios ranging
between 0.1 and 4 (R2 = 0.9968), between 0.1 and 4 (R2 = 0.9955),
between 0.1 and 9 (R2 = 0.9991), and between 0.1 and 9 (R2 = 0.9958).
The response equation
A ðlabeled standardÞ=A ðanalyteÞ
¼ RF ꢂ n ðlabeled standardÞ=n ðanalyteÞ þ b
showed RF values of 0.5386, 1.2980, 0.3978, 0.4629, 1.2467, 1.0898,
and 1.073 and b intercepts of ꢃ0.0407, +0.0056, ꢃ0.0005, ꢃ0.0324,
ꢃ0.0227, ꢃ0.0011, and ꢃ0.0122 for H4folate, 5-methylH4folate, PteGlu,
10-formyl-PteGlu, 5-formylH4folate, pABG, and ApABG, respectively.
Results and discussion
For accurate quantitation of folates in foods and blood plasma
by LC–MS, the use of stable isotopologues proved to be straightfor-
ward in compensating for losses during cleanup due to the lability
of folates and to circumvent ion suppression in electrospray ioniza-
tion (ESI). However, our first stable isotope dilution assay on folate
quantitation in blood plasma lacked sensitivity and precision, with
1 ml of plasma and analysis in triplicates being necessary. In addi-
tion, an assay for red blood cell folates and a method for quantify-
ing folates and their catabolites in urine did not exist at this time.
Analysis of folates and catabolites in plasma
In human studies, the sample amount of blood plasma is lim-
ited. For reducing the impact on the study persons, high precision
and sensitivity are crucial. Although the use of labeled internal
standards compensates for losses during analysis, an improvement
in folate stability could decisively reduce the LODs of the analytes.
For plasma folate analysis, different extraction buffers and SPE
cleanups have been reported. In our previous report on plasma fo-
lates by stable isotope dilution assay, we used a Hepes/Ches buffer
at pH 7.85 followed by SPE cleanup on SAX cartridges [6], which is
a method adapted from food folate analysis. Another stable isotope
dilution assay on serum folates was reported by Pfeiffer and
coworkers [17] using ammonium formate buffer at pH 3.2 and
phenyl SPE cartridges for cleanup. Because preliminary experi-
ments revealed the highest folate stability at pH 5.0, we adjusted
a MES buffer to this pH and used SAX cartridges because the car-
boxyl groups of the folates are dissociated at these conditions.
To test the sensitivity of the new stable isotope dilution assay,
we extracted aliquots of the same plasma with our procedure
and evaluated it by a rough comparison of peak areas with those
of a phenyl SPE sample preparation. In the case of the latter proce-
dure, we observed lower areas for all folates in comparison with
our cleanup. In detail, the peak areas of the phenyl SPE method
made up only 3% H4folate, 14% 5-methylH4folate, 65% PteGlu,
85% 10-formyl-PteGlu, and 96% 5-formylH4folate of our SAX car-
tridge procedure. In contrast to the phenyl SPE method, which ini-
tially was intended to analyze folates only, our SAX cartridge
method also enabled us to quantify the folates along with their
catabolites pABG and ApABG. To enable one cleanup for all ana-
lytes, we needed to compromise with a possible rebinding of 5-
methylH4folate to plasma proteins [23]. However, because the la-
beled 5-methylH4folate was shown to equilibrate quickly with
unlabeled 5-methylH4folate, this effect would not affect the ratio
of the isotopologues and, hence, the result.
Syntheses of folate catabolites
As depicted in Fig. 1,
[
13C5]4-aminobenzoyl-
L
-glutamate
-glutamate
([13C5]pABG, IV) and [13C5]N-acetyl-4-aminobenzoyl-
L
([13C5]ApABG, V) were synthesized by coupling the dimethyl ester
of [13C5]glutamic acid with 4-trifluoracetamidobenzoic acid (II)
and 4-acetamidobenzoic acid (III), respectively, followed by cleav-
age of the protection groups. Therefore, a strategy similar to that of
Maunder and coworkers [21] was applied with extension to ApABG
and some further modifications.
To prepare II and III, 4-aminobenzoic acid was treated with tri-
fluoracetic anhydride and with acetic acid anhydride, respectively.
However, a solvent-free acetylation of 4-aminobenzoic acid as de-
scribed by Mojtahedi and coworkers [22] proved to be unsatisfac-
tory because an equimolar amount of the respective acid
anhydride could not dissolve 4-aminobenzoic acid. Therefore,
dichloromethane was used as an aprotic solvent for this reaction.
The carboxyl groups of [13C5]glutamic acid were reacted with
thionyl chloride and methanol to give the respective dimethyl es-
ter hydrochloride (I), which was then coupled with II and III,
respectively. The latter reaction was accomplished by the use of
HOBt, DCC, and DIPEA. Subsequently, the protection groups were
removed under basic conditions. Because the conditions described
by Maunder and coworkers [21] were not alkaline enough, a more
concentrated sodium hydroxide solution was used and the pH was
maintained at 12.0 on permanent pH control.
The possible drawback of anion exchange cleanup, that elution
buffers containing high salt concentrations might interfere with
LC–MS/MS detection, was circumvented by diverting the LC efflu-
ent into waste at the low retention period when mineral salts are
eluted from the column.
Due to the fact that folates and their catabolites are ubiquitous in
human plasma and can be detected there even in dilutions to a factor
of 50, authenticplasmacould notbe usedasa matrix fordetermining
MS studies on folate catabolites
On positive ESI, the synthesized urine catabolites revealed the
protonated molecule [M+H]+, being the base peak in each one-
dimensional spectrum. In two-dimensional MS spectra, the metabo-
litesshowedsomefragmentationasa resultofthelossoftheglutam-
yl residue; therefore, the labeled compounds lost their labeling

154
Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
Fig. 1. Synthetic route to [13C5]-labeled pABG and ApABG.
the LODs and LOQs. Therefore, we developed a recombinate for
human plasma that consisted of 0.07% protein and 0.06% fat in an
aqueous solution of NaCl and that was devoid of all analytes.
This plasma substitute enabled us to simulate matrix effects
and provided more realistic values for LODs and LOQs than the pro-
cedures using solely aqueous solutions or high dilutions of human

Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
155
119.9
100
80
pABG MS/MS 267 (15 V)
60
40
266.9
20
0
49
119.9
40
30
20
10
[13C5]pABG MS/MS 272 (15 V)
272.1
0
23
161.9
ApABG MS/MS 309 (14 V)
15
10
5
309.1
179.9
267.4
120.2
0
15
162.01
[13C5]ApABG MS/MS 314 (14 V)
10
5
313.9
272.9
180.3
180
0
60
80
100
120
140
160
200
m/z
220
240
260
280
300
320
340
Fig. 2. MS/MS spectra of unlabeled and 13C5-labeled pABG (upper panels) and ApABG (lower panels). Collision energies in parentheses.
plasma. Using the procedure proposed by Vogelgesang and Hädrich
[20], 1, 0.8, 5, 0.6, 4, 0.05, and 0.5 nmol/L were obtained as LODs for
H4folate, 5-methylH4folate, 5-formylH4folate, PteGlu, 10-formyl-
PteGlu, pABG, and ApABG, respectively. The recovery of folates
and their catabolites for the complete stable isotope dilution as-
says was determined by spiking the blank recombinate with
defined amounts of H4folate, 5-methylH4folate, 5-formylH4folate,
PteGlu, 10-formyl-PteGlu, pABG, and ApABG and performing stable
isotope dilution assay with SPE on SAX cartridges as detailed be-
fore. Recoveries were found to be 105, 95, 96, 105, 106, 110, and
107%, respectively.
Thus, we obtained equally good recoveries in comparison with
our previously reported method [6]. However, when considering
the LOD of 1.6 ng/ml of the latter assay for 5-methylH4folate and
the higher sample volume of 1 ml, the new assay increased sensi-
tivity by a factor of 5, which is obviously due to the optimized SPE
cleanup and the better performance of our new LC–MS/MS
equipment.
that equilibration between folate isotopologues was complete even
at the end of the extraction without protease. This might be due to
folates binding to hemoglobin, which Hansen and coworkers [24]
identified as a low-affinity binding protein for folates by using
anion exchange chromatography and 3H-labeled folates.
These results led to the conclusion that working without la-
beled standards essentially requires enzymatic degradation of red
blood cells to obtain accurate data.
Due to the presence of high folate concentrations in human red
blood cells, and because of the high protein contents of lyophilized
erythrocytes (90% hemoglobin), we used lyophilized egg white as
folate-free red blood cell matrix for method validation. As a result
of these studies, we determined 16, 12, 54, 24, and 4 nmol/L as
LODs for H4folate, 5-methylH4folate, 5-formylH4folate, PteGlu,
and 10-formylPteGlu in erythrocytes (fresh weight), respectively.
Because the analytical procedure for red blood cells is quite long
(typically 24 h), the question of analyte stability is obvious. In pre-
liminary studies, we observed absolute SPE recoveries of 37, 48, 32,
88, and 67% for H4folate, PteGlu, 5-methylH4folate, 10-formyl-Pte-
Glu, and 5-formylH4folate, respectively. However, because these
losses were due mainly to SPE cleanup, they were likely to be com-
pensated for by the presence of the labeled standards. Interconver-
sions of folates were not observed in LC–MS/MS screenings for
further folates. Hence, the recoveries of folates in the red blood cell
substitute for the complete stable isotope dilution assays were 105,
105, 95, 105, and 104% for H4folate, 5-methylH4folate, 5-form-
ylH4folate, PteGlu, and 10-formyl-PteGlu, respectively.
Analysis of folates in red blood cells
Our stable isotope dilution assay for red blood cell folates is dif-
ferent from many other assays in that it starts with lyophilization
of the erythrocytes. This procedure not only affected hemolysis but
also enabled optimum homogenization of the sample and avoided
the need for pipetting the viscous red cell solution. Because red
blood cell folates occur mainly as pteroylpolyglutamates, we used
a combined deconjugation by rat serum and chicken pancreas to
generate the monoglutamates. Moreover, during method develop-
ment, we identified an enzymatic proteolytic degradation as an
excellent way to boost sensitivity. Aliquots that were extracted
with protease showed areas that were three times higher than
those extracted without protease. However, because the ratios
between analytes and internal standards were the same, both
extractions resulted in identical folate contents, thereby indicating
Analysis of folates and their catabolites in urine
The newly developed stable isotope dilution assay enables the
quantitation of folates and their catabolites directly from human
urine. Because no conversion of ApABG in pABG is needed, ApABG
and pABG can be measured directly and in the same LC–MS/MS
run.

156
Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
For stable isotope dilution assays, it is important to elucidate
By reflection of these results it can be concluded that at pH 5.0
folates show the best stability and that for this pH value SPE on
SAX cartridge is, in our opinion, the preferred way to recover all fo-
lates and catabolites simultaneously.
the time required for equilibration between unlabeled and labeled
analytes. By variation of the equilibration time, we found 30 min at
room temperature to be sufficient. To test the presence of a folate
binding protein in human urine, we extracted aliquots either with
or without protease. The quantity of pABG increased by protease
treatment even though no effect was observed for the folates and
ApABG. To clarify this effect, we extracted water instead of human
urine with protease and detected 0.03 nmol of pABG in the blank
assay, which revealed that pABG was not protein bound but rather
endogenously present in the protease used.
For quantitation of the catabolites, the concentrations of the
calibration solutions must be adjusted accurately. However, be-
cause ApABG is not commercially available and the synthesized so-
lid standards often lack purity, spectrophotometry was used after
purification of the catabolites by HPLC to verify the concentrations
of solutions. We obtained for pABG at kmax = 275 nm and for
ApABG at kmax = 266 nm in phosphate buffer (100 mmol/L, pH
7.0) molar extinction coefficients of 11.4 0.3 and 17.1 0.6 ml/
Validation data and application of urine stable isotope dilution assay
As found for plasma and red blood cells, a folate- and catabolite-
free matrix could be obtained only by preparing a human urine
recombinate or high dilutions of human urine. We chose synthetic
human urine, which we generated by mixing 2.4% urea, 0.1% creat-
inine, 0.08% uric acid, and 0.005% protein in a physiological solu-
tion of NaCl.
By using this matrix, we determined 0.3, 0.1, 0.9, 0.1, 0.7, 0.01,
and 0.1 nmol/L as LODs for H4folate, 5-methylH4folate, 5-form-
ylH4folate, PteGlu, 10-formylPteGlu, pABG, and ApABG,
respectively.
The recoveries of folates and their catabolites for the complete
stable isotope dilution assays were 105, 95, 96, 110, 103, 110,
and 110% for H4folate, 5-methylH4folate, 5-formylH4folate, PteGlu,
10-formyl-PteGlu, pABG, and ApABG, respectively.
(lmol cm), respectively.
Urine cleanup by SPE
The new stable isotope dilution assay for urinary folates and
their catabolites was applied to samples obtained from a folate bio-
availability study. The contents of catabolites ranged from 14 to
42 nmol/day and from 392 to 790 nmol/day for pABG and ApABG,
respectively. The levels for the former were similar to those re-
ported by Caudill and coworkers [14] and Sokoro and coworkers
[15]. In contrast, the urinary concentrations for ApABG found in
our study exceeded those found by Caudill and coworkers [14]
by a factor of at least 3 but were in line with those reported by Sok-
oro and coworkers [15].
For optimization of urine cleanup prior to LC–MS detection, we
compared the SPE recoveries of all folates and catabolites when
combining MES extraction buffer or the ammonium formate buffer
at pH 3.2 reported by Pfeiffer and coworkers [17] with SAX or phe-
nyl SPE cartridges. In preliminary studies, SPE recoveries were ob-
tained by comparing the initial contents of all analytes quantified
by using [2H4]folic acid as the internal standard with the analytes’
contents in the eluates to which the internal standard was added
after elution.
Mixing the analytes in our extraction buffer and cleanup over
SAX cartridges resulted in SPE recoveries of 16, 23, 39, 10, 48,
and 48% for 5-methylH4folate, 5-formylH4folate, 10-CHO-PteGlu,
PteGlu, pABG, and ApABG, respectively. In the wash solution, no fo-
lates or folate catabolites were detected. In contrast, the analytes
dissolved in ammonium formate buffer and cleanup with phenyl
cartridges gave SPE recoveries of 50, 42, 49, 82, 4, and 87% for Pte-
Glu, 5-methylH4folate, 5-formylH4folate, 10-formylPteGlu, pABG,
and ApABG, respectively. In the wash solution of this method, we
recovered 11% 5-methylH4folate, 32% pABG, and 18% ApABG and
no other folates.
In summary, the SPE recovery of the folates and ApABG over SAX
cartridges was found to be rather low (10–48%) and to be decisively
higherwhenusingphenylcartridges. However, therecoveryofpABG
over phenyl cartridges was as low as 4% and, therefore, rendered the
current procedure with phenyl cartridges unsuitable for urine sam-
ple cleanup due to insufficient affinity for pABG. Another important
advantage of the SAX procedures was the improved stability of the
analytes at pH 5.0, which resulted in higher peak areas, although
the absolute SPE recoveries were lower. In addition, a potential loss
of sensitivity due to protein binding discussed for 5-methylH4folate
in plasma at pH 5.0 [23] is not likely to occur in urine.
An extraction in MES buffer followed by cleanup with phenyl
cartridges was found to be unsuitable (data not shown). Extraction
with citric acid buffer (0.5 mol/L) at pH 5.0 and cleanup with phe-
nyl cartridges also revealed SPE recoveries of only 50, 16, 36, 26,
and 9% for PteGlu, 5-methylH4folate, 5-formylH4folate, 10-form-
ylPteGlu, and ApABG, respectively, and no SPE recovery for pABG.
The wash solution (0.05 mol/L citric acid buffer, pH 5.0) contained
9, 3, 4, 19, 14, and 18% of PteGlu, 5-methylH4folate, 5-form-
ylH4folate, 10-formylPteGlu, ApABG, and pABG, respectively. These
results allow the conclusion that at pH 5.0 folates and their catabo-
lites are present in their ionic forms and show less retention on the
phenyl cartridges.
To clarify this contradiction, we thoroughly checked the quality
assurance of our method and compared it with that of the concur-
ring reports.
The method developed in our lab includes only one short clean-
up step and is able to quantify both metabolites directly with com-
pensation for losses by
[
13C5]-labeled internal standards. In
contrast, the method of Caudill and coworkers [14] requires re-
moval of folates on affinity chromatography and separation of
pABG from ApABG, which subsequently needs to be deacetylated.
This elaborate procedure further involves separation of the ana-
lytes on cation exchange chromatography and then derivatization
of pABG to an azo-compound, which again is purified on HPLC
and is converted to pABG by reduction and finally analyzed by
HPLC–UV. Calculation of the contents includes the results of recov-
ery studies with the tritium-labeled analogues, which gave recov-
eries of 80 and 60% for pABG and ApABG, respectively. From this
description, it appears that this procedure may be prone to various
errors. The procedure lacks measurement specificity when com-
pared with that of Sokoro and coworkers [15], who used an LC–
MS technology with simple precipitation of proteins by the addi-
tion of acetonitrile and direct injection into LC–MS. The latter pro-
cedure involves the use of deuterated glutamic acid as the internal
standard and gave urine contents for the sum of pABG and ApABG
ranging between 500 and 5200 nmol/L. These data significantly ex-
ceed those of Caudill and coworkers [14] and, in contrast, are well
in line with the results for our volunteer. However, a final assess-
ment is still open because the number of volunteers (n = 1) for this
study is limited, in contrast to Caudill and coworkers and Sokoro
and coworkers, who examined more than 20 volunteers each.
Although our method may be considered as more laborious
than that of Sokoro and coworkers [15], the need for detecting
the folates along with the catabolites essentially requires SPE
cleanup, and due to incomplete SPE recovery for some of the ana-
lytes, the use of labeled isotopologues for all compounds is
essential.

Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
157
Precision of the new assays for folates and their catabolites
ter testing, PteGlu was detectable but below its LOQ. Of the
catabolites, pABG was detected at rather low contents ranging be-
tween 0.9 and 3.2 nmol/L without showing any noticeable peaks in
its level after the control dosage. In contrast, pABG showed a signif-
icant peak 1 h after dosage of folic acid before returning to the base
level. To our knowledge, this is the first report on plasma pABG re-
sponse to folate dosages and deserves further investigation.
Precision of real sample analyses was evaluated in a multiple-
injection assay of identical extracts, in an intraassay study of a
sample prepared several times within 1 day, and in an interassay
study of a sample prepared on several days within 4 weeks. All
analyses were performed in triplicate.
For the multiple-injection assay, coefficients of variation (CVs)
in red blood cells were 2 and 3% for H4folate and 5-methylH4folate,
respectively. In urine, the CVs were 9, 10, 4, 8, and 1% for H4folate,
PteGlu, 5-methylH4folate, pABG, and ApABG, respectively. In plas-
ma, the CVs were 1, 1, and 4 for PteGlu, 5-methylH4folate, and
pABG, respectively.
In the intraassay studies, the CVs in red blood cells were 7 and
5% for H4folate and 5-methylH4folate, respectively. In urine, the
CVs were 9, 7, 6, 2, and 8% for H4folate, PteGlu, 5-methylH4folate,
pABG, and ApABG, respectively. In plasma, the CVs were 5, 3, and
8% for PteGlu, 5-methylH4folate, and pABG, respectively.
For the interassay precision studies, real samples were spiked
with analytes in amounts close to the quantitation limit and ana-
lyzed on several days within 4 weeks. CVs in red blood cells were
10, 6, 11, 5, and 3% for H4folate, PteGlu, 5-formylH4folate, 10-for-
myl-PteGlu, and 5-methylH4folate, respectively. In urine, the CVs
were 7, 6, 5, 3, 9, 4, and 8% and in plasma the CVs were 5, 4, 3,
11, 3, 7, and 5% for H4folate, PteGlu, 5-formylH4folate, 10-formyl-
PteGlu, 5-methylH4folate, pABG, and ApABG, respectively.
The other folate vitamers and the catabolite ApABG were not
found above their LOQs given before. As depicted in Fig. 3, the plas-
ma folates revealed significantly different curves for the two tests.
Whereas after the control dose of unfortified water the plasma
folate level showed a nearly constant baseline at 20 to 27 nmol/L,
a significant folate peak was observed after dosage of the folic acid
solution. Plasma folate rose to 34 nmol/L at 3 h postdose and fell to
the base level at 28 nmol/L thereafter. The detection of unmetabo-
lized folic acid in the samples after folic acid dosage is in agree-
ment with other studies on humans consuming foods fortified
with folic acid [19]. In this regard, it is interesting that unmetabo-
lized PteGlu peaks before plasma 5-methylH4folate, indicating a
delay of 2 h before the bulk of PteGlu was converted to 5-meth-
ylH4folate. In contrast to the latter report, 5-formylH4folate was
not detectable in our samples, possibly due to its high LOD of
5 nmol/L in our assay. From these results, it can be concluded that
the study design and plasma analysis are suitable to monitor the
plasma response to oral folate dosages (e.g., in bioavailability stud-
ies). The value of this approach already has been shown in previous
studies [6,25]. With the plasma folate peaking 3 h postdose, we
Application of stable isotope dilution assays to human samples in a
pilot study
To test the suitability of the developed stable isotope dilution
assays, a pilot short-term study with one healthy nonsmoking
male volunteer was performed. The protocol consisted of two test
days with an intertest period of 3 months. To obtain a significant
response to the folate dosages, the volunteer’s folate body pools
were saturated by a dosage of 800
lg of folic acid/day. On the
morning of the first test day, a single dose of 400
lg of folic acid
dissolved in water was consumed by the volunteer, and on the sec-
ond test day, only water was consumed as a control dose.
After predose blood drawings at t = 0 min, both dosages were
followed by 10 blood drawings during a period of 24 h. Plasma fo-
late was determined for each sampling time and red blood cell fo-
lates in the blood samples 24 h after dosage. Urine was assessed for
folates and folate catabolites.
Plasma folates
In all plasma samples, stable isotope dilution assay proved
5-methylH4folate to be the most abundant folate species. Of the
other folates, only PteGlu was quantifiable in the 1-h postdose sam-
pleafterfolicaciddosage. Inthe2-to5-hpostdosesamplesofthelat-
Fig. 3. Levels of plasma folates and catabolites pABG and ApABG after consumption
of 400 lg of folic acid.
Table 1
SRM transitions of folate vitamers.
Compound
Precursor ion (m/z)
Collision energy (V)
Product ion (m/z)
PteGlu
442
446
446
450
460
464
470
474
474
478
456
19
19
19
19
21
21
19
19
25
25
27
295
299
299
303
313
317
295
299
327
331
412
[2H4]PteGlu
H4folate
[2H4]H4folate
5-CH3-H4folate
[2H4]5-CH3-H4folate
10-CHO-PteGlu
[2H4]10-CHO-PteGlu
5-CHO-H4folate
[2H4]5-CHO-H4folate
5,10-Methenyl-H4folate

158
Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
Table 2
Content of folates and their catabolites pABG and ApABG in red blood cells and the urine of one volunteer after oral dosage of water (control) and of a solution of folic acid
(400 g).
l
Dosage
Red blood cells
Urine
Control
Folic acid
Control
Folic acid
Concentration (in nmol/L)
Amount (in nmol/24 h)
H4folate
61
78
n.q.
2
3
18
8
5-MethylH4folate
5-FormylH4folate
Folic acid
10-Formylfolic acid
5,10-MethyleneH4folate
pABG
327
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
388
281
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
359
n.d.
n.d.
n.d.
n.d.
14
4
n.d.
n.d.
42
790
865
ApABG
392
407
Total sum of folates and catabolites
Note. pABG, para-aminobenzoylglutamate; ApABG, acetyl-para-aminobenzoylglutamate; n.d., not detectable; n.q., not quantifiable.
RT:
100
PteGlu
SRM 442/295
50
0
100
[2H4]PteGlu
50
SRM 446/299
SRM 450/303
SRM 460/313
SRM 464/317
SRM 474/327
H4folate
0
100
[2H4]H4folate
50
0
100
5-Me-H4folate
50
0
100
[2H4]5-Me-H4folate
5-CHO-H4folate
50
0
100
50
0
100
[2H4]5-CHO-H4folate
pABG
50
SRM 478/331
SRM 267/120
0
100
50
0
100
[13C5]pABG
50
SRM 272/120
SRM 309/162
0
100
50
ApABG
0
100
[13C5]ApABG
SRM 314/162
50
0
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
Time (min)
Fig. 4. LC–MS/MS chromatogram of human urine after dosage of a folic acid solution. SRM traces: m/z precursor ion/m/z product ion.

Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
159
could confirm earlier findings that reported plasma folate peaks
Acknowledgment
after folic acid dosages ranging between 1 and 3 h [26,27].
This study was supported by a grant from the Deutsche
Forschungsgemeinschaft (RY 19/7-1, NE 1188/1-1).
Blood cell folates
Similarly to plasma folates, the 5-methylH4folate vitamer was
most abundant in erythrocytes. However, the volunteer’s samples
also revealed H4folate ranging between 16 and 22% of total eryth-
rocyte folates. When considering the results of Huang and cowork-
ers [8], the volunteer can be classified to red blood cell phenotype
II and corresponding to the CC or CT genotype of MTHFR reductase,
which is characterized by H4folate ranging between 5 and 20% and
5,10-methenylH4folate below 5% of total erythrocyte folates. For
this volunteer, 5,10-methenylH4folate was not detectable in our
samples.
Total erythrocyte folates after the test dosages were not found
to be significantly different. In contrast to plasma folate level, red
blood cell folate is considered as a long-term marker of folate sta-
tus because it reflects the mean folate status of the last 3 months,
which is the lifetime of erythrocytes. Therefore, the single dose
application of folates is not likely to significantly change the
long-term folate status, in accordance with our findings (Table 2).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ab.2009.11.007.
References
[1] J. Selhub, Folate, vitamin B12, and vitamin B6 and one carbon metabolism, J.
Nutr. Health Aging 6 (2002) 39–42.
[2] A.F. Czeizal, J. Dudas, Prevention of the first occurrence of neural-tube defects
by periconceptional vitamin supplementation, N. Engl. J. Med. 327 (1992)
1832–1835.
[3] M. Caudill, The role of folate in reducing chronic and developmental disease
risk: an overview, J. Food Sci. 69 (2003) SNQ55–SNQ59.
[4] D.A. Snowdon, C.L. Tully, C.D. Smith, K.P. Riley, W.R. Markesbery, Serum folate
and the severity of atrophy of the neocortex in Alzheimer disease: findings
from the Nun Study, Am. J. Clin. Nutr. 71 (2000) 993–998.
[5] K. Robinson, Homocystein, B vitamins, and risk of cardiovascular disease, Heart
83 (2000) 127–130.
[6] M. Rychlik, M. Netzel, I. Pfannebecker, T. Frank, I. Bitsch, Application of stable
isotope dilution assays based on liquid chromatography–tandem mass
spectrometry for the assessment of folate bioavailability, J. Chromatogr. B
792 (2003) 167–176.
Urinary folates and catabolites
[7] Z. Fazili, C.M. Pfeiffer, M. Zhang, R.B. Jain, 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. 54 (2008) 197–201.
[8] Y. Huang, S. Khartulyari, M.E. Morales, A. Stanislawska-Sachadyn, J.M. Von
Feldt, A.S. Whitehead, I.A. Blair, Quantification of key red blood cell folates
from subjects with defined MTHFR 677C>T genotypes using stable isotope
dilution liquid chromatography/mass spectrometry, Rapid Commun. Mass
Spectrom. 22 (2008) 2403–2412.
[9] D.E.C. Smith, R.M. Kok, T. Teerlink, C. Jakobs, Y.M. Smulders, Quantitative
determination of erythrocyte folate vitamer distribution by liquid
chromatography–tandem mass spectrometry, Clin. Chem. Lab. Med. 44
(2006) 450–459.
[10] L.D. Botto, Q. Yang, 5,10-Methylenetetrahydrofolate reductase gene variants
and congenital anomalies: a HuGE review, Am. J. Epidemiol. 151 (2000) 862–
877.
Intact folates are considered to occur only in trace amounts in
urine and only when high oral doses have been applied. This is in
agreement with our results given that H4folate, 5-methylH4folate,
5-formylH4folate, and folic acid were found only after the dosage
of folic acid and were not detectable after the control testing
(Fig. 4). Unexpectedly, these folates were found in urine but not
in blood plasma. This result leads to the conclusion that the plasma
method is less sensitive due to a smaller sample size.
The application of our new stable isotope dilution assay to the
folate catabolites pABG and ApABG revealed the latter to be the
most abundant one in urine and confirmed earlier studies
[14,15]. Besides the folate catabolites, minor folate contents were
detected in the volunteer’s urine after the folic acid and control
dosage, as detailed in Table 2. After the control dose, 5-meth-
ylH4folate was detectable in a low content, confirming earlier re-
ports on detection of this vitamer in urine [28]. In contrast,
H4folate, 5-formylH4folate, and unmetabolized PteGlu were
detectable in the urine after the folic acid dosage. These findings
confirm an earlier report of Pfeiffer and coworkers [29], who found
small amounts of unmetabolized folic acid in urine after the dosage
of fortified cereals, and highlight the high sensitivity of the new as-
say. Moreover, we found a significant increase of the sum of
catabolites after the folic acid dosage. This is contradictory to the
report of Caudill and coworkers [13], who concluded that catabo-
lites do not respond to short-term dosage and, in contrast, reflect
the total body folate turnover and are proportional to the whole
body pool.
[11] J.F. Gregory, J. Williamson, J.F. Liao, L.B. Bailey, J.P. Toth, Kinetic model of folate
metabolism in nonpregnant women consuming [2H2]folic acid: isotopic
labeling
of
urinary
folate
and
the
catabolite
para-
acetamidobenzoylglutamate indicates slow, intake-dependent, turnover of
folate pools, J. Nutr. 128 (1998) 1896–1906.
[12] J.M. Wolfe, L.B. Bailey, K. Herrlinger-Garcia, D.W. Theriaque, J.F. Gregory, G.P.A.
Kauwell, Folate catabolite excretion is responsive to changes in dietary folate
intake in elderly women, Am. J. Clin. Nutr. 77 (2003) 919–923.
[13] M.A. Caudill, L.B. Bailey, J.F. Gregory, Consumption of the folate breakdown
product para-aminobenzoylglutamate contributes minimally to urinary folate
catabolite excretion in humans: investigation using
aminobenzoylglutamate, J. Nutr. 132 (2002) 2613–2616.
[
13C5]para-
[14] M.A. Caudill, J.F. Gregory, A.D. Hutson, L.B. Bailey, Folate catabolism in
pregnant and nonpregnant women with controlled folate intakes, J. Nutr. 128
(1998) 204–208.
[15] A.A. Sokoro, M.L. Etter, J. Lepage, B. Weist, J. Eichhorst, D.C. Lehotay, Simple
method for the quantitative analysis of endogenous folate catabolites p-
aminobenzoylglutamate (pABG) and its acetamido (apABG) derivative in
human serum and urine by liquid chromatography–tandem mass
spectrometry, J. Chromatogr. B 832 (2006) 9–16.
[16] M. Rychlik, K. Englert, S. Kapfer, E. Kirchhoff, Folate contents of legumes
determined by optimized enzyme treatment and stable isotope dilution
assays, J. Food Comp. Anal. 20 (2007) 411–419.
Conclusions
[17] C.M. Pfeiffer, Z. Fazili, L. McCoy, M. Zhang, E.W. Gunter, Determination of folate
vitamers in human serum by stable-isotope-dilution tandem mass
spectrometry and comparison with radioassay and microbiologic assay, Clin.
Chem. 50 (2004) 423–432.
[18] Z. Fazili, C.M. Pfeiffer, Measurement of folates in serum and conventionally
prepared whole blood lysates: application of an automated 96-well plate
isotope-dilution tandem mass spectrometry method, Clin. Chem. 50 (2004)
2378–2381.
[19] A. Freisleben, P. Schieberle, M. Rychlik, Syntheses of labeled vitamers of folic
acid to be used as internal standards in stable isotope dilution assays, J. Agric.
Food Chem. 50 (2002) 4760–4768.
[20] J. Vogelgesang, J. Hädrich, Limits of detection, identification, and
determination: a statistical approach for practitioners, Accredit. Qual. Assur.
3 (1998) 242–255.
The stable isotope dilution assays presented here are the first
reported methods to quantify single folates and their catabolites
in diverse clinical samples simultaneously due to enhanced folate
stability and a modified cleanup procedure. The results of this pre-
liminary human study point to the possibility of mirroring the
plasma folate response by measuring the urine catabolite contents,
which would render short-term bioavailability studies much easier
and compliable to the volunteers. A clinical trial with different fo-
late dosages and an appropriate number of subjects is currently
under way to assess this assumption.

160
Quantitation of folates and their catabolites / S. Mönch et al. / Anal. Biochem. 398 (2010) 150–160
[21] P. Maunder, P.M. Finglas, A.I. Mallet, F.A. Mellon, M.A. Razzaque, B. Ridge, L.
Vahteristo, C. Witthöft, The synthesis of folic acid, multiply labeled with stable
isotopes, for bio-availability studies in human nutrition, J. Chem. Soc. Perkin
Trans. 1 1999 (1999) 1311–1323.
[22] M.M. Mojtahedi, A.M. Saeed, M.M. Heravi, F.K. Behbahani, Additive-free
chemoselective acylation of amines and thiols, Monatshefte Chem. 138
(2007) 95–99.
[23] B.C. Nelson, C.M. Pfeiffer, S.A. Margolis, C.D. Nelson, Solid-phase extraction–
electrospray ionization mass spectrometry for the quantification of folate in
human plasma or serum, Anal. Biochem. 325 (2004) 41–51.
[24] S.I. Hansen, J. Holm, J. Lyngbye, Evidence that the low-affinity folate-binding
protein in erythrocyte hemolysate is identical to hemoglobin, Clin. Chem. 27
(1981) 1247–1249.
[26] L.B. Bailey, L.E. Barton, S.E. Hillier, J.J. Cerda, Bioavailability of mono and
polyglutamyl folate in human subjects, Nutr. Rep. Int. 38 (1988) 509–518.
[27] L.M. Rogers, C.M. Pfeiffer, L.B. Bailey, J.F. Gregory, A dual-label stable-isotopic
protocol is suitable for determination of folate bioavailability in humans:
evaluation of urinary excretion and plasma folate kinetics of intravenous and
oral doses of [13C5] and [2H2]folic acid, J. Nutr. 127 (1997) 2321–2327.
[28] D.J. Hart, P.M. Finglas, C.A. Wolfe, F. Mellon, A.J.A. Wright, S. Southon,
Determination of 5-methyltetrahydrofolate (13C-labeled and unlabeled) in
human plasma and urine by combined liquid chromatography mass
spectrometry, Anal. Biochem. 305 (2002) 206–213.
[29] C.M. Pfeiffer, L.M. Rogers, L.B. Bailey, J.F. Gregory, Absorption of folate from
fortified cereal-grain products and of supplemental folate consumed with or
without food determined by using a dual-label stable-isotope protocol, Am. J.
Clin. Nutr. 66 (1997) 1388–1397.
[25] R. Prinz-Langenohl, A. Bronstrup, B. Thorand, M. Hages, K. Pietrzik, Availability
of food folate in humans, J. Nutr. 129 (1999) 913–916.