Detection and quantification of lanthanide complexes in cell lysates by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry

Article abstract of DOI:10.1021/ac049162u

Gadolinium (III) complexes are under intense scrutiny as contrast agents for magnetic resonance imaging. Although currently used mainly as extracellular agents, there is a growing interest to exploit their contrast enhancing ability in the intracellular environment To ascertain the preservation of their chemical integrity upon the intracellular entrapment, it is necessary to have a method for their dosage in the cell lysates. Herein, a mass spectrometric method for detection and quantification of gadolinium complexes in cell lysates is reported. The detection by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) was carried out by using a non-acidic matrix (2,4,6-trihydroxyacetophenone), which does not allow any leakage of gadolinium from the complex. Quantification has been possible by using as an internal standard an ytterbium complex with the same ligand of the analyte. Ytterbium was chosen because, among the lanthanides, it is the one with the isotopic distribution pattern the most similar to that of gadolinium. Sensitivity was enough to detect low micromolar quantities of a cationic complex and high micromolar quantities of a neutral complex in cell lysates of rat hepatoma cells. In the case of anionic complexes, sensitivity was too low for quantitative analysis. To the best of our knowledge, this is the first report concerning the quantification of metal complexes by MALDI-TOF-US.

Full text of DOI:10.1021/ac049162u

Anal. Chem. 2004, 76, 6012-6016
Detection and Quantification of Lanthanide
Complexes in Cell Lysates by Matrix-Assisted
Laser Desorption/Ionization Time-of-Flight Mass
Spectrometry
Davide Corpillo,* Claudia Cabella,† Simonetta Geninatti Crich,‡ Alessandro Barge,‡ and Silvio Aime‡
Laboratorio Integrato Metodologie Avanzate, Bioindustry Park Canavese, 10010 Colleretto Giacosa, Via Ribes 5, Italy
be the most useful for this application because they possess a
large magnetic moment and a long electronic relaxation time.4
Highly stable Gd chelates are currently used in clinical settings
as reporters of blood flow and organ perfusion. The next
generation will deal with systems endowed with targeting capabili-
ties, and some applications will involve imaging probes that are
able to be specifically entrapped into cells.
Gadolinium (III) complexes are under intense scrutiny
as contrast agents for magnetic resonance imaging. Al-
though currently used mainly as extracellular agents,
there is a growing interest to exploit their contrast
enhancing ability in the intracellular environment. To
ascertain the preservation of their chemical integrity upon
the intracellular entrapment, it is necessary to have a
method for their dosage in the cell lysates. Herein, a mass
spectrometric method for detection and quantification of
gadolinium complexes in cell lysates is reported. The
detection by matrix-assisted laser desorption/ ionization
time-of-flight mass spectrometry (MALDI-TOF-MS) was
carried out by using a non-acidic matrix (2 ,4 ,6 -trihy-
droxyacetophenone), which does not allow any leakage of
gadolinium from the complex. Quantification has been
possible by using as an internal standard an ytterbium
complex with the same ligand of the analyte. Ytterbium
was chosen because, among the lanthanides, it is the one
with the isotopic distribution pattern the most similar to
that of gadolinium. Sensitivity was enough to detect low
micromolar quantities of a cationic complex and high
micromolar quantities of a neutral complex in cell lysates
of rat hepatoma cells. In the case of anionic complexes,
sensitivity was too low for quantitative analysis. To the best
of our knowledge, this is the first report concerning the
quantification of metal complexes by MALDI-TOF-MS.
Recently, Gd complexes have also been proposed for neutron
capture therapy (NCT).5-7 In the latter case, entrapment into the
cell is mandatory for the success of the therapy because neutron-
activated Gd nuclei emit Auger electrons whose cell damaging
effects are confined to few micrometers.
Both for therapeutic and diagnostic applications, the determi-
nation of the amount of internalized intact Gd complex is of
fundamental importance for monitoring the efficacy of the pro-
posed method. Actually, the determination of Gd ions would have
been much easier, but it may be misleading because the biological
environment may lead to partial metal decomplexation. Therefore,
one has to discard very sensitive techniques, such as inductively
coupled plasma mass spectrometry (ICP-MS), X-ray spectromi-
croscopy, and time-of-flight secondary ion mass spectrometry
(TOF-SIMS), because they are not able to discriminate between
Gd complexes and other gadolinium-containing compounds aris-
ing from the speciation of Gd ions released from the complex.
The specific uptake of Gd complexes has already been shown
in several cell cultures.3,8,9 Among the techniques which can be
used for quantification of complexes in cell lysates, nuclear
magnetic resonance (NMR) has an inherently low sensitivity; high-
pressure liquid chromatography (HPLC) needs further work for
optimization of separation procedures; and electrospray ionization
mass spectrometry (ESI-MS) has a low tolerance toward sample
contaminants, especially salts.
The advent of magnetic resonance imaging (MRI)1 for clinical
diagnosis has brought an increasing interest in design and
synthesis of paramagnetic complexes for their potential use as
contrast enhancers.2 Among the paramagnetic ions, some lan-
thanides, and particularly gadolinium (III),3 have been shown to
* To whom correspondence should be addressed. Fax: +39 0125 53028.
E-mail: davide.corpillo@unito.it.
(4) Runge, V. M.; Clanton, J. A.; Price, A. C.; Wehr, C. J.; Herzer, W. A.; Partain,
C. L.; James, A. E. Magn. Reson. Imaging 1 9 8 5 , 3, 43-55.
† Current address: Bracco Imaging SpA, Bioindustry Park Canavese, Coller-
etto Giacosa, Italy.
(5) Shih, J. L.; Brugger, R. M. Med. Phys. 1 9 9 2 , 19, 733-744.
(6) De Stasio, G.; Casalbore, P.; Pallini, R.; Gilbert, B.; Sanita`, F.; Ciotti, M. T.;
Rosi, G.; Festinesi, A.; Larocca, L. M.; Rinelli, A.; Perret, D.; Mogk, D. W.;
Perfetti, P.; Mehta, M. P.; Mercanti, D. Cancer Res. 2 0 0 1 , 61, 4272-4277.
(7) Matsumura, A.; Zhang, T.; Yamamoto, T.; Yoshida, F.; Sakurai, Y.; Shimojo,
N.; Nose, T. Anticancer Res. 2 0 0 3 , 23, 2451-2456.
(8) Reimer, P.; Bader, A.; Weissleder, R. Eur. Radiol. 1 9 9 7 , 7, 527-531.
(9) Schmalbrock, P.; Hines, J. V.; Lee, S. M.; Ammar, G. M.; Kwok, E. W. J.
Magn. Reson. Imaging 2 0 0 1 14, 636-648.
‡ Current address: Dipartimento di Chimica IFM, Universita` di Torino, Torino,
Italy.
(1) Beckmann, N.; Laurent, D.; Tigani, B.; Panizzutti, R.; Rudin, M. Drug
Discovery Today 2 0 0 4 , 9, 35-42.
(2) Wisner, E. R.; Aho-Sharon, K. L.; Bennett, M. J.; Penn, S. G.; Lebrilla, C. B.;
Nants, M. H. J. Med. Chem. 1 9 9 7 , 40, 3992-3996.
(3) Aime, S.; Cabella, C.; Colombatto, S.; Geninatti Crich, S.; Gianolio, E.;
Maggioni, F. J. Magn. Reson. Imaging 2 0 0 2 , 16, 394-406.
6012 Analytical Chemistry, Vol. 76, No. 20, October 15, 2004
10.1021/ac049162u CCC: $27.50 © 2004 American Chemical Society
Published on Web 09/17/2004

MALDI-TOF-MS, instead, is a simple, rapid, and highly specific
technique that requires minimal sample volumes and no sample
pretreatment and has a high sensitivity and a better tolerance
toward salts than ESI-MS.
Quantitative analysis in MALDI-TOF-MS experiments has been
carried out so far by adding to the samples a stable isotope-
incorporated internal standard, chemically equivalent to the target
molecule,10-15 to minimize any variability in sample/ matrix co-
crystallization and desorption/ ionization processes. This approach
is, however, very expensive, especially when a mass shift of at
least 10 Da is needed, as it would be in the case of Gd complexes
(whose isotopic pattern is composed of at least six predominant
peaks).
When dealing with complexes, an effect analogous to the
isotope-labeling, but simpler and cheaper, can be pursued by
using, as an internal standard, a complex with the same ligand of
the analyte to be quantified that contains a different metal ion.
For the quantitative analysis of Gd complexes, we found that
ytterbium could be ideal as an internal standard metal because it
is the lanthanide ion with an isotopic distribution the most similar
to gadolinium, and their masses are different enough (16 Da) to
avoid any overlap between standard and analyte peaks.
For an accurate determination of metal complex content in cell
lysates, it is also important to choose experimental conditions
which avoid the leakage of metal ions from the complexes during
both lysis and MALDI experiments. In this context, the use of
the appropriate matrix is crucial.
Synthesis of Gd- and Yb-DTP A-BAG. DTPA (diethyl-
enetriaminepentaacetic acid) was dissolved in pyridine and mixed
with an excess of acetic anhydride. The suspension was heated
at 65 °C for 3 h, then cooled and filtered. The solid (DTPA-
bisanhydride) was washed with acetic anhydride, dichloromethane,
and diethyl ether and lyophilized. Agmatine (4-guanidinobutyl-
amine) sulfate was dissolved in water, and agmatine was extracted
in 1-butanol at pH 13. The solvent was removed under reduced
pressure, and the residue was dissolved in dimethyl sulfoxide
(DMSO) at 50 °C. DTPA-bisanhydride was then dissolved in
DMSO and added to agmatine at 50 °C. The reaction mixture
was stirred at 50 °C for 10 h. An excess of acetone was then added,
and the solid DTPA-BAG (N,N′-(4-guanidinobutyl)diethylenetri-
aminepentacetic acid bisamide) was then filtered, washed with
acetone, dissolved in water at pH 2, and purified onto an Amberlite
XAD 1180 resin, by elution with a gradient of methanol from 0 to
100%. Lyophilized DTPA-BAG was then dissolved in water at pH
6.5 and mixed with an excess of gadolinium (III) or ytterbium
(III) chloride. The reaction mixture was stirred at 70 °C for 8 h.
1
Complex formation was followed by measuring H NMR spectra
up to the complete disappearance of the free ligand signals. The
excess of gadolinium (III) or ytterbium (III) chloride was removed
by formation of Gd(OH)3 or Yb(OH)3 at basic pH, followed by
centrifugation and filtration (on a 0.2-µm syringe filter) of the
resulting suspension. The metal complex containing solution was
then lyophilized.
Cell Culture. HTC (rat hepatoma cells) were grown in 75-
cm2 flasks in a humidified incubator at 37 °C and at CO2/ air (5:
95 v/ v) in DMEM-F12 medium supplemented with 5% FBS, 2 mM
glutamine, 100 U/ mL penicillin and 100 µg/ mL streptomycin.
Cells were seeded in 10-cm Petri dishes at a density of 15 000
cells/ cm2. After 24 h, cells were ready for the uptake experiments.
Uptake Experiments. HTC cells seeded in 10-cm Petri dishes
were used to carry out the uptake experiments with either Gd-
DTPA-BAG or Gd-HPDO3A. After removal of the culture
medium, the cells were washed with 5 mL of phosphate buffered
saline (PBS) and then 5 mL of Earl’s balanced salt solution
(EBSS: CaCl2 0.266 g/ L, KCl 0.4 g/ L, NaCl 6.8 g/ L, glucose 1
g/ L, MgSO4 0.204 g/ L, NaH2PO4 0.144 g/ L, and NaHCO3 1.1 g/ L
pH 7.4) were added. HTC cells were incubated for 6 h at 37 °C
with either 1 mM Gd-DTPA-BAG or 50 mM Gd-HPDO3A. At
the end of the incubation, the medium was removed, and the cells
were washed three times with ice-cold PBS. Then cells were
harvested with a rubber policeman in 250 µL of PBS and
centrifuged at 800g for 5 min. The supernatant was discarded,
and the cell pellet was resuspended in 100 µL of H2O and sonicated
for 10 s for a complete lysis of the cells. Untreated cells were
incubated, and lysates were collected by following the same
procedure.
In this work, we report the determination of intracellular
amounts of two Gd complexes differing in the residual charge,
Gd-DTPA-BAG being positively charged and Gd-HPDO3A
neutral under lysis conditions.
MATERIALS AND METHODS
Materials. The DMEM-F12 medium and fetal bovine serum
(FBS) were from Cambrex Bioscience (Verviers, Belgium).
Glutamine, penicillin/ streptomycin, and all other chemicals were
from Sigma Chemical Co. (St. Louis, MO).
Gd-HPDO3A (Prohance) is a Bracco Imaging (Milan, Italy)
trademark. Both Gd-HPDO3A complex (white powder) and
HPDO3A ligand were kindly provided by Bracco Imaging.
Synthesis of Yb-HP DO3 A. Yb-HPDO3A was synthesized
by adding an excess of ytterbium (III) oxide to an aqueous solution
of HPDO3A (10-(2-hydroxypropyl)-1,4,7,10-tetraazacyclododecane-
1,4,7-triacetic acid). The reaction mixture was stirred at 70 °C for
8 h. Complex formation was followed by measuring 1H NMR
spectra up to the complete disappearance of the free ligand signals.
The excess of Yb2O3 was removed by formation of Yb(OH)3 at
basic pH, followed by centrifugation and filtration (on a 0.2-µm
syringe filter) of the resulting suspension. The metal complex
containing solution was then lyophilized.
MALDI-TOF-MS Experiments. Each sample was diluted 1:1
with a saturated solution of the matrix (2,4,6-trihydroxyacetophen-
one, THAP, in 50% acetonitrile), and 0.5 µL was spotted directly
on a MALDI target plate. MALDI mass spectra were acquired in
the positive reflectron ion mode with delayed extraction on a
Reflex III time-of-flight instrument (Bruker Daltonics, Bremen,
Germany) equipped with a 337-nm nitrogen laser. Experimental
settings: ion acceleration voltage, 20.00 kV; reflector voltage, 23.00
kV; and first extraction plate, 17.00 kV. Mass spectra were
(10) Duncan, M. W.; Matanovic, G.; Cerpa-Poljak, A. Rapid Commun. Mass
Spectrom. 1 9 9 3 , 7, 1090-1094.
(11) Desiderio, D. M.; Wirth, U.; Lovelace, J. L.; Fridland, G.; Umstot, E. S.;
Nguyen, T. M. D.; Schiller, P. W.; Szeto, H. S.; Clapp, J. F. J. Mass Spectrom.
2 0 0 0 , 35, 725-733.
(12) Gobom, J.; Kraeuter, K. O.; Persson, R.; Stehen, H.; Roepstorff, P.; Ekman,
R. Anal. Chem. 2 0 0 0 , 72, 3320-3326.
(13) Wittmann, C.; Heinzle, E. Biotechnol. Bioeng. 2 0 0 1 , 72, 642-647.
(14) Horak, J.; Werther, W.; Schmid, E. R. Rapid Commun. Mass Spectrom. 2001,
15, 241-248.
(15) Bungert, D.; Heinzle, E.; Tholey, A. Anal. Biochem. 2 0 0 4 , 326, 167-175.
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004 6013

Figure 1. Formula of DTPA-BAG ligand.
obtained by averaging 300 laser shots. Calibration of the spectra
was performed externally with a solution of standard peptides
covering the range 200-2000 m/ z.
Quantification by MALDI-TOF-MS. The standard curve for
Gd-DTPA-BAG quantification was obtained by adding to a lysate
of untreated HTC cells aliquots of the internal standard Yb-
DTPA-BAG (up to a final concentration of 40 µM) and aliquots
of Gd-DTPA-BAG to final concentrations ranging from 4 to 100
µM (six points, each in triplicate). The same amount of internal
standard was then added to unknown samples (deriving from Gd-
DTPA-BAG-treated cells), before mixing them to the matrix
solution, as previously described.
Figure 2. MALDI-TOF spectrum of 40 µM Gd-DTPA-BAG dis-
solved in water and mixed with R-cyano-4-hydroxycinnamic acid
(CHCA) in 50% acetonitrile as matrix. The two peaks are the
monoprotonated complex (monoisotopic peak at 770 m/z) and the
monoprotonated ligand (monoisotopic peak at 618 m/z), respectively.
Peak integration was executed with the X-Mass software
(Bruker Daltonics, Bremen, Germany). The ratio of the Gd-
DTPA-BAG to Yb-DTPA-BAG signal intensities was measured
as the area of the first six monoisotopic peaks of the Gd complex
divided by the area of the first six monoisotopic peaks of the Yb
complex. This ratio was plotted as a function of the final
concentration of Gd-DTPA-BAG in cell lysates.
The same procedure was applied for Gd-HPDO3A. The
concentration of the internal standard (the Yb complex) was 400
µM, and the covered range of the analyte concentration (the Gd
complex) was 100 µM to 1 mM.
RESULTS AND DISCUSSION
To assess whether it is possible to quantify metal complexes
in cell lysates, we first selected a complex, Gd-DTPA-BAG
(whose ligand formula is presented in Figure 1), which is
positively charged in the lysis conditions (pH around 7-8), as
cationic species are likely to give more intense peaks when
analyzed with MALDI-TOF-MS in the positive ion mode.
First, it was necessary to find proper experimental conditions
which avoid any leakage of the metal during the mass spectro-
metric measurement. The experiment with R-cyano-4-hydroxy-
cinnamic acid (CHCA) as matrix16 yielded a MALDI spectrum
which shows, in addition to the monocharged peak of the complex,
a peak corresponding to the free ligand. This finding clearly
indicates a partial dissociation of the complex (Figure 2).
Then 6-aza-2-thiothymine (6-ATT), a neutral matrix often used
for noncovalent interactions studies,17 was tested. This compound
does not affect the integrity of the complex but displays a high
sensitivity to sample contaminants when the experiment is carried
out in a HTC lysate.
Figure 3. MALDI-TOF spectrum of 40 µM Gd-DTPA-BAG dis-
solved in HTC lysate and mixed with 2,4,6-trihydroxyacetophenone
in 50% acetonitrile as matrix. The monoprotonated complex peak
(monoisotopic peak at 770 m/z) is present, together with various
sodium and potassium adducts, while the monoprotonated ligand
(monoisotopic peak at 618 m/z) is absent.
acetonitrile, gave the best result, because it was possible to find
the intact complex peaks also when Gd-DTPA-BAG was dis-
solved in cell lysates in low micromolar concentrations, while the
ligand peak was completely absent (Figure 3).
As the internal standard, a complex with the same ligand but
a different metal was used. Metal choice was limited to lanthanides
to minimize the differences with the Gd complex for their chemical
and physical properties. Initially, we tried with the europium
complex, but there were problems concerning both the overlap
with Gd-complex peaks (mass difference between the two lan-
thanides is only 4 Da) and the linearity of the response when
quantitative trials were carried out.
Finally, a matrix consisting of a compound often used for
oligosaccharides18 and oligonucleotides19 analysis, THAP in 50%
(16) Zhu, X.; Papayannopoulos, I. A. J. Biomol. Technol. 2 0 0 3 , 14, 298-307.
(17) Glocker, M. O.; Bauer, S. H.; Kast, J.; Volz, J.; Przybylski, M. J. Mass
Spectrom. 1 9 9 6 , 31, 1221-1227.
(18) Papac, D. I.; Wong, A.; Jones, A. J. Anal. Chem. 1 9 9 6 , 68, 3215-3223.
(19) Zhu, Y. F.; Chung, C. N.; Taranenko, N. I.; Allman, S. L.; Martin, S. A.,
Haff, L.; Chen, C. H. Rapid Commun. Mass Spectrom. 1 9 9 6 , 10, 383-388.
Better results were obtained with ytterbium, a lanthanide ion
with an isotopic distribution quite similar to that of gadolinium.
Their mass difference (16 Da) is enough to avoid overlaps between
6014 Analytical Chemistry, Vol. 76, No. 20, October 15, 2004

Figure 5. Formula of HPDO3A ligand.
For these reasons, the calibration curve for Gd-HPDO3A
covers the range 100 µM to 1mM, with Yb-HPDO3A 400 µM as
the internal standard.
In the experiments with HPDO3A as the ligand, no peak
corresponding to the monoprotonated complex was detectable,
whereas the typical gadolinium and ytterbium isotopic distribution
pattern was detected for sodium and potassium adducts of the
two complexes. Unfortunately, the mass difference between
ytterbium and gadolinium (16 Da) corresponds to the mass
difference between potassium and sodium, thus yielding to a
perfect overlap of potassium adducts of Gd complex with sodium
adducts of Yb-one. Nevertheless, choosing the dipotassium adduct
of Yb-HPDO3A (which cannot overlap with any Gd-HPDO3A
sodium adduct) as internal standard for the calibration curve, we
were able to quantify Gd-HPDO3A as its dipotassium adduct. In
fact, even if this adduct has the same mass of the sodium-
potassium Yb-HPDO3A adduct, the contribution of the latter can
be considered constant, because the good linear correlation found
demonstrates that in fixed experimental conditions (especially for
what concerning cell line and culture medium choice), the
partition of the complexes in the various salt adducts is always
the same. Interestingly, the disodium, sodium-potassium, and
dipotassium adducts appear in the spectrum as monopositively
charged peaks, even if they should be doubly charged. This
observation can be explained by the (partial) dissociation of the
alcohol group present in HPDO3A complexes. In solution, this
dissociation occurs only at basic pH20 (pKa ) 11.36 for Ln-
HPDO3A);21 however, it is known that practically, only singly
charged ions are detected in MALDI spectra, which are “irrespec-
tive of acid-base chemistry in the solution and the charge state
of the preformed ion within the matrix-analyte solid”.22
The best fit for the calibration curve for Gd-HPDO3A
corresponds to the equation y ) 0.0069x + 1.8533 (with a
correlation coefficient R2 ) 0.9953), where y corresponds to the
Gd-/ Yb-HPDO3A ratio and x is the concentration of Gd-
HPDO3A in cell lysates.
Figure 4. MALDI-TOF spectrum of a lysate from HTC cells treated
with 1 mM Gd-DTPA-BAG. The integrations of the monoprotonated
peaks of both the analyte (monoisotopic peak at 770 m/z) and the
added internal standard (40 µM Yb-DTPA-BAG, monoisotopic peak
at 786 m/z) are shown. From the calibration curve, the resulting
concentration of the Gd-DTPA-BAG in the HTC lysate was 9 µM.
standard and analyte peaks (except a negligible overlap between
Yb complex and the sodium adduct of the Gd complex).
On this basis, a calibration curve for Gd-DTPA-BAG in HTC
lysate was set up by MALDI-TOF-MS. It is usually recommended11
that a ratio of internal standard to analyte range between 4:1 and
1:4. For a Gd-DTPA-BAG concentration of 4-100 µM, we used
Yb-DTPA-BAG with a concentration of 40 µM.
To get a stable ratio among the peaks,12 300 shots were
accumulated for every spectrum. The integration was carried out
on the first six monoisotopic peaks of each compound because
they are the most intense. Peaks corresponding to sodium and
potassium adducts appeared to be a constant percentage if
compared to the corresponding monoprotonated peaks and were
not considered in the integration procedure.
The best fit line of the calibration curve yields to the equation
y ) 0.0298x - 0.0211 (with a correlation coefficient R2 ) 0.9967),
where y corresponds to the Gd-/ Yb-DTPA-BAG ratio and x is
the concentration of Gd-DTPA-BAG in cell lysates. The equation
was then used to quantify this complex in unknown samples after
treatment of HTC cells with Gd-DTPA-BAG (an example is
given in Figure 4, in which the resulting complex concentration
was 9 µM). Because the experimental workup yields a dilution of
about 1:20 of the cell content during lysis procedures, the complex
concentrations in the cells can be assessed as 20-fold higher than
those calculated from the quantification procedure performed on
the lysates.
Even with lower sensitivity for the HPDO3A complexes, the
calibration curve covers a concentration range that includes the
quantities of the complex that were found in the lysates of cells
treated with this complex (Figure 6, in which the resulting
complex concentration was 160 µM).
The same procedure was applied for the quantitative determi-
nation of Gd-HPDO3A (whose ligand formula is shown in Figure
5), a complex which is neutral in the lysis conditions. The net
charge of the complex is expected to affect the quantification
assay. In fact, the sensitivity resulted significantly lower than that
found with Gd-DTPA-BAG experiments. The lower ionization
efficiency is not evident when the complex is analyzed in water
(where it is detectable to at least 40 µM final concentration), but
it decreases by 1-2 orders of magnitude when the experiments
are performed in cell lysates. No improvement was obtained when
negative ions were analyzed.
With complexes that are negatively charged under lysis
conditions, sensitivity was found to be too low for quantification
purposes (data not shown) when both positive and negative ions
were analyzed. It has already been reported22 that signals of
negative ion MALDI spectra are markedly lower if compared to
positive ones. This fact was more evident in reflectron than in
(20) Aime, S.; Botta, M.; Fasano, M.; Terreno, E. Acc. Chem. Res. 1 9 9 9 , 32,
941-949.
(21) Toth, E.; Kiraly, R.; Platzek, J.; Raduchel, B.; Brucher, E. Inorg. Chim. Acta
1 9 9 6 , 249, 191-199.
(22) Karas, M.; Glu¨ ckmann, M.; Scha¨fer, J. J. Mass Spectrom. 2 0 0 0 , 35, 1-12.
Analytical Chemistry, Vol. 76, No. 20, October 15, 2004 6015

especially concerning the matrix, for detection in cell lysates
without loss of the ion from the complex. 2,4,6-Trihydroxyaceto-
phenone in 50% acetonitrile was shown to be the best matrix for
this purpose. With this approach, no pretreatment is needed, and
the specificity of the technique allows discrimination between Gd
complexes and other gadolinium-containing compounds.
In cell lysates, neutral complexes showed a lower ionization
efficiency when compared to cationic ones. The lower limit of
detection of both cationic and neutral complexes (4 and 100 µM,
respectively) was shown, however, to be enough for the quanti-
fication of these complexes in treated cells, as proved by uptake
experiments. Sensitivity was still lower for anionic compounds,
preventing in this case the setup of a calibration curve in a
reasonable range of concentrations. The good linear correlation
found for the calibration curves demonstrates that a metal complex
with the same ligand of the analyte can be considered as good as
a stable isotope-labeled compound, even considering the variability
in sample/ matrix cocrystallization and desorption/ ionization pro-
cesses. We think that this approach for the quantification can be
extended to many kinds of metal complexes, at least to cationic
or neutral ones. Upon choosing the chemical equivalent complex
as an internal standard, it is particularly important to first inspect
the similarity of chemical properties, physical properties, and
isotopic distribution with respect to the analyte.
Figure 6. MALDI-TOF spectrum of a lysate from HTC cells treated
with 50 mM Gd-HPDO3A. The integrations of the dipotassium adduct
peaks of both the analyte (monoisotopic peak at 633 m/z) and the
added internal standard (400 µM Yb-HPDO3A, monoisotopic peak
at 649 m/z) are shown. From the calibration curve, the resulting
concentration of the Gd-HPDO3A in the HTC lysate was 160 µM.
linear mode, but in the latter case, the quantification was prevented
by the worse resolution.
Received for review June 8, 2004. Accepted August 11,
2004.
CONCLUSIONS
For the first time, a quantification of metal complexes has been
performed by MALDI-TOF-MS. Conditions have been optimized,
AC049162U
6016 Analytical Chemistry, Vol. 76, No. 20, October 15, 2004