2301700

A Method for the Analvsis of Catecholamines
by Selected Ion Monitoiing and 14C Isotope
Dilution in Adrenal Medullary Cell Culture
J. C. Lhuguenot?' and B. F. Maumee
Centre National de Reference et d'Essai de Spectrometrie de Masse and ERA C.N.R.S. 267, Faculte de SciencesMirande, UniversitC de Dijon, B.P. 138, 21004 Dijon Cbdex, France
Extracts have been made from culture medium of rat medullar adrenal cells developed in tissue culture in this
laboratory. After pentafluorobenzylimine-trimethylsilylether formation the catecholamine derivatives were
characterized by gas-liquid chromatography chemical ionization mass spectrometry. In order to assess the
catecholamine production capabilities of the cells in culture, a mass spectrometricmethod with isotope dilution
has been devised. Chemical ionization selected ion monitoringallows specific detection at the nanogram level in a
higher mass range (400-600 amu) than in the electron impact mode. The isotopic dilution method with 14C
catecholamines gives rise to accurate measurements and linear response in the picomole range. The use of the
[M-15]+ion for monitoring m / z values minimizes errors in selected ion monitoring analysis. The results
obtained are computerized and treated by the data system for fine background subtraction when high sensitivity
and accuracy are required.
INTRODUCTION
Normal adreno medullary cells isolated from newborn
rats and grown in tissue culture have been established in
this laboratory. These cells can be maintained in primary
cell culture for several weeks and retain a differentiated
morphology as seen by electron microscopy: presence of
numerous chromaffin granules and well developed Golgi
apparatus. Furthermore, dopamine and norepinephrine
have been detected in the culture medium of these cell
cultures and characterized by gas chromatography mass
spectrometry (GCMS) in the methane chemical ionization (CI) mode (Fig. 1).
This cellular system represents an appropriate model
for studying the regulation of the catecholamine
biosynthetic pathway. Measurements of the quantitative
changes in this pathway due to hormonal stimulation or
specific inhibition demand an accurate, sensitive and
specific analytical method. In catecholamine analysis,
gas chromatographic and mass spectrometric methods
have been reported.'-6 Selected ion monitoring (SIM)
with a capillary column was used for the study of the
ontogenic variation of dopamine and norepinephrine
biosynthesis in the adrenal of developing rat.7 In the
present work, taking advantage of the high efficiency of
the capillary column and the high sensitivity and
specificity of detection of their effluents by CI SIM, we
have developed a reliable quantitative method at the
14
nanogram level with C labelled catecholamines as
internal standards.
During the past decade, isotopically labelled materials
have been used with increasing frequency as internal
standards for quantitative mass spectrometry or SIM
Abbreviations: DA = dopamine; NA =norepinephrine; PFB =
pentafluorobenzaldehyde;
BSA = bis(trimethylsily1)acetamide;
DMF = dimethylformamide.
ILaboratoire de Biochimie Appliquee, ENS.BANA-Dijon, France.
$ Author to whom correspondence should be addressed.
and several methods of treatment of raw data have been
d e s ~ r i b e d . ~ -In' ~the case of the use of a I4C internal
standard, we propose here a simple method to treat the
isotope ratio obtained by SIM to obtain the true mole
ratio of the molecule analysed to the known amount of
internal standard (x/y ratio). Furthermore, the choice of
a 14C labelled molecule as the analytical internal standard permits the use of stable labelled compounds as
precursors in biogenic pathways.
0
100
150
200
250
Scon number
Figure 1. Total ion current trace obtained using methane CI of an
extract from a medium of rat medullary adrenal cell in culture.
Age of t h e animals: 4 days, ag e of the culture: 10 days. Mass
spectrum no. 127 is identical t o that of dopamine a s PFB-TMS
derivative. Mass spectrum no. 196 corresponds t o PFB-TMS
norepinephrine. Capillary GC ClMS conditions ar e reported in the
Experimental section.
0306-042X/80/0007-0529$02.00
@ Heyden & Son Ltd, 1980
BIOMEDICAL MASS SPECTROMETRY, VOL. 7, NOS. 11 AND 12, 1980 529
J. C. LHUGUENOT AND B. F. MAUME
EXPERIMENTAL
loor
30
Equipment
The combined GCMS was carried out using a Finnigan
Model 3300 gas chromatograph mass spectrometer
coupled to a Model 6100 data system for data acquisition and processing. A 25 m long by 0.2 mm i.d. glass
capillary column with SE-30 (Spiral, Dijon, France) was
employed under isothermal temperature conditions at
205 "C or programmed temperature conditions (170235 "C at 2 "C min-I).
Chemical ionization with methane as reagent gas was
used. The source parameters were: emission current,
0.25 mA; electron energy, 118 eV; reagent gas pressure, 0.75 Torr.
300
400
350
500
450
rn /I
Figure 2. CI mass spectrum of unlabelled dopamine as PFB-TMS
derivative. The [M-15]+ ion (mlz460) is more useful than the
[MH]+ ion ( m l z 476) for SIM used as isotopic calculation data (no
presence of the [M-161+ ion).
Materials and reagents
Dopamine (DA) and norepinephrine (NA) were
obtained from Calbiochem, and [7 -14C]DA Pgecific
activity,
54 mCI mMol-')
and
[7- C]NA
(57 mCi mMol-') from CEA (Gif-sur-Yvette, France).
Pentafluorobenzaldehyde
(PFB)
and
bis(trimethylsily1)acetamide (BSA) were purchased from
Pierce Chemical Co.
Solvents (grade A Merck) were used without further
purification except for dimethylformamide (DMF)
which is purified just before use by azeotropic distillation with benzene.
All glassware was cleaned by washing with 50%
HN03 and was rinsed extensively in glass distilled H 2 0
prior to use.
Preparation of samples
Freshly prepared standard solutions of D A and [714
CIDA, or NA and [7-14C]NA were used to generate a
series of five samples ranging from 0 to 100°/~of the
labelled compound. Each tube contained 1 O ~ gof
(DA + [7-14C]DA) and 10 pg of (NA + [7-14C]NA).
These samples were evaporated to dryness with a stream
of dry nitrogen. To each 1OOpl of DMF, 1 0 0 ~ of
1
acetonitrile and 40 pl of BSA (silylating reagent) were
added. The tubes were capped and heated for 2 min at
60 "C. Then 10 p1 of PFB solution (1mg m1-I) was
added to each and the samples were allowed to stand at
room temperature for 15 min. The solution obtained
was diluted to 1 ml with acetonitrile.
Catecholamines were extracted from rat adrenal cell
culture medium by the usual method reported in an
earlier paper for the adrenal t i ~ s u e . ~
a- and &bonds of the lateral chain with the reteqiion of
the positive charge on the catechol fragment. This
earlier work clearly shows that a reliable measurement is
obtained without a carrier. For this reason, labelled
compounds are used only as internal standards.
It is well known that the use of high m/z values for
monitoring ions augments the specificity in SIM.
However, as mass spectra of PFB-TMS derivatives of
catecholamines are very poor in the high amu range
when the E I mode is
[MI+ or [M- 15]+ ions
cannot be used in E I SIM if high sensitivity is required.
The mass spectrum of dopamine as a PF3-TMS derivative with methane CI (Fig. 2) indicates two possible ions
for monitoring: the m / z 476 ([MH]+) or m / z 460
([M- 15]+) ions. In our case, CI gives better specificity
than the E I mode with good sensitivity.
If data is collected at m / z 476 and 478 corresponding
to the [MH]+ ion for unlabelled and 14C labelled
dopamine, it is necessary to estimate the contribution of
the [MI+. ion isotope cluster to [MH]+. The use of the
[M - 15]+ ion at m / z 460 and 462 for dopamine and at
m / z 548 and 550 for norepinephrine gives rise to
simpler and more accurate isotopic calculation.
Furthermore, the sensitivity is higher for norepinephrine determination due to a greater relative intensity of
this ion as shown in Table 1 and Fig. 3, for the case of the
[7-14C]norepinephrine.
A satisfactory mathematical expression of the calibration curve when a labelled standard is used can be
-
Table 1. Relative intensity ratios of the [M 15]+ and [MI"
ions in the EI mode, and [M- 151' and [MH]+ ions
in the CI mode for unlabelled and I4C labelled
dopamine (DA) and norepinephrine (NA) as PFBTMS derivatives
DA[M-151+/[M+HIC
RESULTS AND DISCUSSION
The absolute sensitivity for dopamine as PFB-TMS with
electron impact (EI) SIM has been reported earlier.7 It
has been demonstrated that the limit of detection with
accurate linearity is about 1pg if the m / z 267 ion is
monitored. This ion corresponds to the cleavage of the
530 BIOMEDICAL MASS SPECTROMETRY, VOL. 7, NOS. 11 AND 12, 1980
CI
4601476
4621478
methane
0.63
0.63
DA IM - 15It/[M1'El
4601475
4621477
(30eV)
0.17
0.17
NA[M-l5I+I[M+Hl+
5481564
3.12
5501566
3.12
NA [M- 15l+/[M1'
5481563
1.66
5501565
1.65
~
@ Heyden & Son Ltd, 1980
ANALYSIS OF CATECHOLAMINES
rr
m/z 550.1
I
\
m/z 548.1
rn/z
Figure 3. CI mass spectrum of [7-"k]norepinephrine as PFBTMS derivative. In this case, the IM-15]+ ion gives a better
sensitivity than MH ion in SIM.
m/z 460.0
obtained from the Pickup and McPherson equation"
reported in Fig. 4.
This equation does not correspond to the general
linear relationship y = ax + 6, but to a homographic
function of pattern y = (ax + b ) / ( c x+ d ) . Values of x / y
ratio (unlabelled/labelled) commonly reported for
internal standard use, lead to a non-linear calibration
curve.
A linear calibration curve is obtained with RN
(normalized value of R k l plotted versus x / y according
to Schoeller' or Thenot'
2
R-a
RN=1 -bR
a=-
where R corresponds to R k l or
experimental SIM values
b = -P1
Pk
Qk
Q1
Pk
I
0.07
I
PI
1 0.13
DA
I
100
200
300
Scan number
400
500
Figure 5. Selected ion recordings obtained for [7-l4C]DA and
[7-14C]NArecorded a t mlz 462 and 550 respectively. The mlz 460
and 548 ions give the response for unlabelled DA and NA.
and k = P k / Q l is equal to the slope of the calibration
curve R N= f ( x / y ) .
According to this preliminary theoretical aspect,
characteristic values (a, b, k ) for calibration are determined by SIM analysis of pure unlabelled and labelled
compounds as shown in Fig. 4.
The values obtained are as follows: for dopamine as
PFB-TMS derivative a = 0.123, b = 0.149 and k = 0.98;
for norepinephrine (PFB-TMS) a = 0.063, b = 0.163
and k = 0.915.
E = 460
F 1462
460
462
ml2
0.14
,
NA
E.540
0.06
540
F = 550
10.94
550
m'z
Figure 4. Isotopic distribution for unlabelled and [7-14C]labelled
DA or NA obtained by SIM. The s mbols used are those of the
1lY
Pickup and McPherson equation.
@ Heyden & Son Ltd, 1980
X/ Y
Figure 6. Calibration curve RN = f ( x l y ) for dopamine determination. The curve is obtained by use of the coefficient k equal
to the slope ( k = 0.98). The three points correspond to three
assays at different x l y values.
BIOMEDICAL MASS SPECTROMETRY, VOL. 7, NOS. 11 AND 12, 1980 531
J. C. LHUGUENOT AND B. F. MAUME
Table 2. Values of x/y ratio obtained from selected ion monitoring ratios by RNcalculationfor three samples with
0.33, 1 and 3 volumetric relative concentrations of
~nlabelled/'~C
labelled dopamine (DA)
/
3t
Theoretical
XJY
1
SIM ratio
RN
X/Y=
RN.-
k
value
OA
k = 0.98
3
1
0.33
*
2.13 0.07
0.86*0.06
0.42Zt0.05
2.94* 0.15
0.85i0.08
0.32+0.05
*
2.99 0.15
0.87*0.08
0.32dz0.05
Table 3. Values of x/y ratio obtained from selected ion monitoring ratios by RNcalculationfor three sampleswith
0.33, 1 and 3 volumetric relative concentrations of
~nlabelled/'~C
labelled norepinephrine (NA)
x/Y
Theoretical
Figure 7. Calibration curve RN = f ( x l y ) for norepinephrine
determination. The curve is drawn by use of the coefficient k
(k = 0.915). The three points correspond to assays for samples
with 0.33, 1 and 3 xly values.
The selected ion recording of [7-'4C]dopamine and
[7-14C]norepinephrine is given in Fig. 5 as an example.
The coefficient k calculated from SIM of labelled and
unlabelled compounds is used from drawing the calibration curve R N= f ( x / y ) . The curves for dopamine and
norepinephrine are reported in Figs. 6 and 7.
The R N values of the three samples for x / y values
corresponding to 0.3,1 and 3 (volumetric ratio) are used
for estimating the accuracy of this method. Tables 2 and
3 give the results obtained. Ea& selected ion recording
S ( 1 / J n )for three experiments.
ratio corresponds to
The accuracy of the method decreases when xly
deviates from a value of about one. For this reason, it is
necessary to have the best sensitivity in SIM when
labelled and unlabelled compounds are checked for the
determination of the coefficient k. The use of the [M151' ion gives a convenient precision due to its high m l z
value obtained with methane CI and to the absence of
[M- 161' ion contribution. Good agreement with the
study of Colby and McCaman'* is observed for
x*
XI v
value
NA
k=0.315
3
1
0.33
SIM ratio
RN
1.92Zt0.07
0.88* 0.01
0.34Zt0.03
2.70*0.11
0 . 9 5 i 0.01
0.29i0.03
x l y = R;-
1
k
2.95k0.15
1.04* 0.03
0.32k0.04
measurement precision. Accurate values of SIM ratios
are obtained at ng level when CI is used.
In conclusion, compared with previously reported
methods, this assay, based on CI SIM offers significant
advantages in terms of specificity and accuracy. Sensitivity at the picogram level can be reached when E I
ionization is used on the same sample, but lower m / z
ions are monitored. Consequently, the CI SIM method
meets the requirements for a correct quantitative
analysis of dopamine and norepinephrine biosynthesis
in adrenal medullary cell culture.
Acknowledgements
This work was supported by the Ministtre des Universites for the
heavy equipment in mass spectrometry, by the Centre National de la
Recherche Scientifique for the computer and through the ERA 267,
and by the Institut National de la SantC et de la Recherche Mkdicale
(FRA 9).
REFERENCES
1. S. H. Koslow, F. Cattabeni and E. Costa, Science 176, 177
(1972).
2. F. Karoum, F. Cattabeni, E. Costa, C. R. J. Ruthven and M.
Sandler, Anal. Biochem. 47,550 (1972).
3. B. F. Maume, P. Bournot, J. C. Lhuguenot, C. Baron, F.
Barbier, G. Maume, M. Prost and P. Padieu, Anal. Chem. 45,
1073 (1973).
4. J. C. Lhuguenot and B. F. Maume, J. Chromatogr. Sci. 14,411
(1974).
5. F. Karoum, J. C. Fillin, R. J. Wyattand E. Costa, Biomed. Mass
Spectrom. 2, 183 (1975).
6. K. P. Wong, C. R. J. Ruthven and M. Sandler, Clin. Chim. Acta
47, 215 (1973).
7. J. C. Lhuguenot, B. F. Maume and P. Padieu, in Recent
Developments in Mass Spectrometry and Medicine, ed. by A.
Frigerio, p. 269. Plenum Press, New York (1978).
8. H.R. Sullivan, P. G. Wood and R. E. McMahon, Biomed. Mass
Spectrom. 3,212 (1976).
9. D. A. Schoeller, Biomed. Mass Spectrom. 3 , 265 (1976).
10. J. F. Pickup and K. McPherson, Anal. Chem. 48, 1885 (1976).
532 BIOMEDICAL MASS SPECTROMETRY, VOL. 7,
NOS. 11 AND 12, 1980
11. B. N. Colbyand M. W. McCaman, Biomed. MassSpectrom. 5,
215 (1978).
12. B. N. Colby and M. W. McCaman, Biomed. MassSpectrom. 6,
225 (1979).
13. D. Picart, J. Jacolot, F. Berthou and H. H. Floch, in Quantitative Mass Spectrometry in Life Sciences 11, ed. by A. P. de
Leenheer, p. 105. Elsevier, Amsterdam (1978).
14. D. J. Jenden, Proceedings of the 2nd International Conference on Stable Isotopes, Oak Brook, Illinois, ed. by E. R.
Klein and P. Klein, p. 420 (1975).
15. J. P. Thenot, Thesis, p. 85, Dijon (1980).
16. A. C. Moffat and E. C. Horning, Biochem. Biophys. Acta 222,
248 (1970).
Received 13 June 1980
0Heyden & Son Ltd, 1980
Paper presented at the Third International Symposium on Quantitative Mass Spectrometry i n Life Sciences, Gent, Belgium, June
1980.
@ Heyden & Son Ltd, 1980