Analysis of derivatised steroids by matrix-assisted laser desorption/ionisation and post-source decay mass spectrometry

Article abstract of DOI:10.1016/j.steroids.2005.08.002

Neutral steroids are difficult to analyse using desorption ionisation methods coupled with mass spectrometry (MS). However, steroids with an unhindered ketone group can readily be derivatised with the Girard P (GP) reagent to give GP hydrazones. Steroid GP hydrazones contain a quaternary nitrogen atom and are readily desorbed in the matrix-assisted laser desorption/ionisation (MALDI) process, giving an improvement in sensitivity of two orders of magnitude. Steroids without a ketone group, but with a 3β-hydroxy-Δ⁵ function, can be readily converted to 3-oxo-Δ⁴ steroids and subsequently derivatised to GP hydrazones for MALDI analysis. In addition to giving strong [M]⁺ ions upon MALDI, steroid GP hydrazones give informative post-source decay (PSD) spectra. By using the accurate mass of the precursor-ion measured by MALDI-MS, in combination with the structural information encoded in its PSD spectrum, steroid structures can readily be determined.

Full text of DOI:10.1016/j.steroids.2005.08.002

steroids 7 1 ( 2 0 0 6 ) 42–53
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/steroids
Analysis of derivatised steroids by matrix-assisted
laser desorption/ionisation and post-source decay
mass spectrometry
Muhammad Atif Khana, Yuqin Wanga, Sibylle Heidelbergera, Gunvor Alveliusb,
Suya Liub, Jan Sjo¨vallb, William J. Griffithsa,∗
a
Department of Pharmaceutical and Biological Chemistry, The School of Pharmacy, University of London, 29-39 Brunswick Square,
London WC1N 1AX, UK
b
Department of Medical Biochemistry and Biophysics, Karolinska Institutet, SE-17177 Stockholm, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Neutral steroids are difficult to analyse using desorption ionisation methods coupled with
mass spectrometry (MS). However, steroids with an unhindered ketone group can readily
be derivatised with the Girard P (GP) reagent to give GP hydrazones. Steroid GP hydrazones
contain a quaternary nitrogen atom and are readily desorbed in the matrix-assisted laser
desorption/ionisation (MALDI) process, giving an improvement in sensitivity of two orders of
magnitude. Steroids without a ketone group, but with a 3␤-hydroxy-ꢀ5 function, can be read-
ily converted to 3-oxo-ꢀ4 steroids and subsequently derivatised to GP hydrazones for MALDI
analysis. In addition to giving strong [M]+ ions upon MALDI, steroid GP hydrazones give
informative post-source decay (PSD) spectra. By using the accurate mass of the precursor-
ion measured by MALDI-MS, in combination with the structural information encoded in its
PSD spectrum, steroid structures can readily be determined.
Received 12 October 2004
Received in revised form 2 August
2005
Accepted 8 August 2005
Available online 30 September 2005
Keywords:
Derivatisation
Mass spectrometry
Matrix-assisted laser
desorption/ionisation
Post-source decay
© 2005 Elsevier Inc. All rights reserved.
1.
Introduction
et al. described matrix-assisted laser desorption/ionisation
(MALDI)-MS [4]. Now, at the turn of the 20th century, mass
spectrometry is no longer the preserve of the specialist, but
is integrated into most biochemistry/biology institutions, and
many of today’s biological scientists have direct access to
either a MALDI- or ES-MS instrument.
MALDI is most often coupled to time-of-flight (TOF) mass
analysers, which can be used to obtain accurate mass data [5].
However, MALDI-TOF instruments fitted with a reflectron are
also capable of post-source decay (PSD) fragment-ion analy-
sis [6–8]. In PSD laser activated precursor-ions are analysed.
PSD fragmentation occurs after ion acceleration, but before
ions enter the electric field of the reflectron. The precursor-ion
Since the early 1980s, biological mass spectrometry (MS) has
undergone a paradigm shift. Pre-1981 biological mass spec-
trometry was largely concentrated on the analysis of ther-
mally stable volatile molecules, or those that could be made
so by simple derivatisation [1]. However, in 1981, Barber et
al. published their seminal paper on fast atom bombardment
(FAB) ionisation, which demonstrated the desorption ionisa-
tion of non-volatile, thermally labile and ionic biomolecules
[2]. Following Barber’s discoveries, biological mass spectrome-
try has changed out of all recognition. In 1984, Yamashita and
Fenn described electrospray (ES)-MS [3], and in 1987, Karas
∗
Corresponding author. Tel.: +44 20 7753 5876; fax: +44 20 7753 5964.
E-mail address: wiliam.griffiths@ulsop.ac.uk (W.J. Griffiths).
URL: http://www.ulsop.ac.uk/depts/pharmchem/griffiths.html.
0039-128X/$ – see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2005.08.002

steroids 7 1 ( 2 0 0 6 ) 42–53
43
(along with its PSD fragments which reach the reflectron with
the same velocity) is selected using a timed ion-selector (Brad-
bury Nielsen Gate) with a precursor-ion resolution (M/ꢀM) of
about 100, and fragment-ion spectra are sequentially recorded
in several segments which are then stitched together.
pregnanolone (5␤-P-3␣-ol-20-one), 7␤-hydroxycholesterol (C5-
3␤,7␤-diol), 24S-hydroxycholesterol (C5-3␤,24S-diol), 7␣,25-
dihydroxycholesterol (C5-3␤,7␣,25-triol), 7␣,27-dihydroxy-
cholesterol (C5-3␤,7␣,27-triol) and [16,16,17(or 20),22,22,23,23-
2H7]7␣,27-dihydroxycholesterol were from previous studies
in the Karolinska laboratory (Table 1).
While the MALDI and ES ionisation methods are well suited
to the analysis of both acidic and basic biomolecules, they
are not well suited to the analysis of neutral biomolecules
[9]. This is a significant drawback in lipid analysis, where
many important species are neutral. For example, steroid
hormones, anabolic steroids, oxysterols and cholesterol itself
are all neutral molecules and are ionised poorly by both ES
and MALDI [10]. Atmospheric pressure chemical ionisation
and atmospheric pressure photoionisation provide alternative
methods of ionisation of neutral steroids, but these ionisation
techniques often lead to the formation of dehydrated proto-
nated molecules with the inherent loss in structural informa-
tion [11,12]. One way of enhancing neutral steroid analysis by
either ES- or MALDI-MS is to derivatise the neutral steroid to
an acidic, basic or charged analogue. In the present communi-
cation, we described the derivatisation of neutral steroids to
their charged Girard P (GP) hydrazone analogues, and the per-
formance of these derivatives upon MALDI and PSD analysis.
2.1.
steroids
Oxidation of 3␤-hydroxy-ꢀ5 steroids to 3-oxo-ꢀ4
The 3␤-hydroxy-ꢀ5 steroids were oxidised with cholesterol
oxidase essentially as described by Brooks et al. [13]. The
3␤-hydroxy-ꢀ5 steroids (1 ␮g) were dissolved in 50 ␮L of iso-
propanol, and 10 ␮L of cholesterol oxidase (1 mg/mL, 20 U/mg
protein) in 1 mL of phosphate buffer (50 mM KH2PO4, pH 7)
added, the mixture was incubated at room temperature (25 ◦C
for up to 12 h). The reaction was stopped by the addition of
1 mL of methanol, and the solution was passed through a Sep-
Pak C18 cartridge (Waters, Milford, MA) prepared in water. The
effluent and 1 mL of water were combined and passed again
through the same Sep-Pak cartridge. After washing with 1 mL
of 20% methanol, the 3-oxo-ꢀ4 steroids were eluted with 1 mL
of methanol.
2.2.
Derivatisation of 3-oxo-ꢀ4 steroids
2.
Experimental
The derivatisation of 3-oxo-ꢀ4 steroids to GP hydrazones was
carried out essentially as described by Wheeler [14]. Ten mil-
ligrams of GP hydrazine (Sigma–Aldrich) was added to a solu-
tion of oxosteroid (1 ␮g) in 70% aqueous methanol (1 mL) con-
taining glacial acetic acid (200 ␮L), and warmed at 70 ◦C for
30 min. After cooling, the steroid GP hydrazone was dried
under a stream of nitrogen gas and reconstituted in a solu-
All solvents were of analytical grade. Water was from a
Milli-Q water system (Millipore, Molsheim, France). The
steroids androstenedione (A4-3,17-dione), dehydroepisandro-
sterone (DHEA, A5-3␤-ol-17-one), testosterone (A4-17␤-ol-
3-one), [19,19,19-2H3]testosterone, norgestrel (18-homo-E4-
17␣-ethynyl-17␤-ol-3-one),
progesterone
(P4-3,20-dione),
Table 1 – Oxysterols analysed by MALDI-PSD
Name Abbreviation
Androstenedione
Oxidation producta
Mass of GP GP hydrazone
A4-3,17-dione
A5-3␤-ol-17-one
–
–
–
–
–
420
422
422
425
446
(I)/(II)
(III)
(IV)
(V)
Dehydroepisandrosterone (DHEA)
Testosterone
[19,19,19-2H3]testosterone
Norgestrel
A4-17␤-ol-3-one
[19,19,19-2H3]A4-17␤-ol-3-one
18-Homo-E4-17␣-ethynyl-17␤-
ol-3-one
(VI)
Progesterone
P4-3,20-dione
–
–
448
452
534
534
550
550
557
(VII)/(IIX)
(IX)
(XI)
(X)
(XII)
Pregnanolone
5␤-P-3␣-ol-20-one
C5-3␤,7␤-diol
7␤-Hydroxycholesterol
24S-Hydroxycholesterol
7␣,25-Dihydroxycholesterol
7␣,27-Dihydroxycholesterol
[16,16,17(or 20),22,22,23,23-2H7]
7␣,27-dihydroxycholesterol
C4-7␤-ol-3-one
C4-24S-ol-3-one
C4-7␣,25-diol-3-one
C4-7␣,27-diol-3-one
[16,16,17(or 20),22,22,23,23-
2H7]C4-7␣,27-diol-3-one
C5-3␤,24S-diol
C5-3␤,7␣,25-triol
C5-3␤,7␣,27-triol
[16,16,17(or 20),22,22,23,23-
2H7]C5-3␤,7␣,27-triol
(XIII)
(XIV)
.
a
Oxidation product upon treatment with cholesterol oxidase.

44
steroids 7 1 ( 2 0 0 6 ) 42–53
Scheme 1 – Structures of the steroid GP hydrazones analysed.

steroids 7 1 ( 2 0 0 6 ) 42–53
45
tion of 10% aqueous methanol (1 mL). The steroid GP hydra-
zone was separated from un-reacted GP reagent by extrac-
tion on a Sep-Pak C18 cartridge. After washing with 10%
aqueous methanol (1 mL), the GP hydrazone was eluted from
the cartridge with methanol (1 mL). The derivatisation pro-
tocol has been applied to mixtures of oxosteroids on the
microgram–nanogram level [15] and is also suitable for the low
level (picogram) derivatisation of neutral steroids extracted
from tissue. For the derivatisation of oxosteroids with a labile
7-hydroxy-ꢀ4-3-oxo structure, the amount of GP hydrazine
reagent was increased from 10 mg to 100 mg and the reaction
was carried out overnight at room temperature.
Preuss) has catalytic activity towards steroids with shorter C-
17 side-chains than cholesterol (data not shown).
3.1.
MALDI mass spectra
Neutral steroids are difficult to analyse by MALDI with out
derivatisation. Taking testosterone as an example of a 3-
oxo-ꢀ4 steroid, derivatisation to the GP hydrazone gave an
enhancement in ion-current of at least 500. Furthermore,
derivatisation of oxosteroids to their GP hydrazone analogues
adds 134 Da to their measured mass and thereby moves the
ion of interest to a region of reduced matrix-associated noise.
Shown in Fig. 1 are MALDI mass spectra of mixtures of testos-
terone (IV) and pregnanolone (IX) GP hydrazones. In Fig. 1a,
both steroids were present in the matrix solution at a con-
centration of 5 ng/␮L, while in Fig. 1b the steroid concen-
tration was 0.5 ng/␮L. GP hydrazones contain a quaternary
nitrogen and exist as ions in solution, and as salts in the crys-
talline form. In MALDI mass spectra GP hydrazone [M]+ ions
are observed which correspond to closed-shell even-electron
species. It can be noted that preganolone GP hydrazone gives
less abundant [M]+ ions than testosterone GP hydrazone, this
2.3.
MALDI spectra
Steroid GP hydrazones (Scheme 1) in methanol were mixed
with an equal volume of matrix solution (␣-cyano-4-
hydroxycinnamic acid, 10 g/L in 50% acetonitrile and 0.1%
trifluoroacetic acid) and 1 ␮L aliquots of the mixture spotted
on to the MALDI target plate. The samples were allowed to
dry in air. All MALDI spectra were recorded on an Applied
Biosystems (Framingham, MA, USA) Voyager TOF mass spec-
trometer equipped with a nitrogen laser (337 nm). All MALDI
mass spectra were recorded in the positive-ion mode and
were averages of 100 individual laser shots. The accelerat-
ing voltage was set at 20 kV, the focusing guide wire was at
0.075% and the extraction delay time was 200 ns. Spectra were
externally, or internally, calibrated using steroid GP hydrazone
standards.
In the PSD experiments, the precursor-ions were selected
using the timed ion-selector. Fragment-ions were refocused
onto the detector by stepping the voltage applied to the
reflectron in appropriate increments to give ∼10 segments.
The individual segments obtained for each mirror ratio were
stitched together. The laser intensity was adjusted for each
segment to maximise the number of structurally significant
fragments.
3.
Results
Previous studies have shown that 3-oxo-ꢀ4 steroids can be
converted to their 3-GP hydrazone analogues in 30 min at 70 ◦C
[15]. In the current study, it was found necessary to reduce the
reaction temperature to 25 ◦C for the derivatisation of labile
7-hydroxy-ꢀ4-3-oxo steroids to avoid thermal elimination of
water. However, when the reaction was left overnight, quanti-
tative conversions to 3-oxo-ꢀ4 GP hydrazones were achieved
(data not shown). Although 17- and 20-oxosteroids are not
quantitatively converted to their GP hydrazones under the
above conditions, when analysed by MALDI MS they were
found to give ion-currents at least 100 times greater than the
non-derivatised steroids.
In the present study, hydroxycholesterol substrates have
been converted quantitatively to 3-oxo-ꢀ4 steroids using
cholesterol oxidase. Using this enzyme, it is possible to
oxidise steroids with a 3␤-hydroxy-ꢀ5 or 5␣-hydrogen 3␤-
hydroxy function to 3-oxosteroids, and furthermore, the spe-
cific enzyme used (cholesterol oxidase, from Brevibacterium,
recombinant, lyophilized, obtained from Roche Diagnostics
GmbH, Mannheim, Germany, through the courtesy of Dr. Ingo
Fig. 1 – MALDI mass spectra of a mixture of GP hydrazones
of testosterone (m/z 422.28) (IV) and pregnanolone (m/z
452.33) (IX). (a) 5 ng and (b) 0.5 ng of each steroid loaded on
the MALDI target plate. Matrix peaks are indicated with an
asterisk.

46
steroids 7 1 ( 2 0 0 6 ) 42–53
Fig. 2 – MALDI mass spectra of 2.5 ng of GP hydrazones of:
(a) C4-24S-ol-3-one (m/z 534.41) (X) and (b)
C4-7␣,27-diol-3-one (m/z 550.40) (XIII). Matrix peaks are
indicated with an asterisk.
may be a consequence of better desorption properties of 3-
oxo-ꢀ4 GP hydrazones than 20-oxo GP hydrazones, and/or
incomplete derivatisation of the 20-oxo group in pregnanolone
to its GP hydrazone. Each of the steroid GP hydrazones anal-
ysed were readily detected at the 2.5 ng/␮L level, this is further
illustrated in Fig. 2a and b which displays MALDI mass spec-
tra of matrix solutions containing 2.5 ng/␮L of GP hydrazones
of C4-24S-ol-3-one (X) and C4-7␣,27-diol-3-one (XIII). It can be
noted in Fig. 2b that the presence of a hydroxyl group on the
steroid ring system at position 7 results in the formation of
[M − 18]+ ions in addition to [M]+ ions.
Scheme 2 – (a and b) Fragmentation of protonated
testosterone. A prime to the right of a fragment-describing
letter indicates the ion contains the added proton, e.g.
bꢀ1-12.
PSD spectrum shown in Fig. 3a was obtained. Fragment-ions
are observed at m/z 97 (b1ꢀ -12) and 109 (b3ꢀ -CH3–CH) [10,18,19]
(Scheme 2). Although these ions are structurally informative,
a sample requirement of at least 100 ng is a considerable draw-
back to the analytical method.
3.2.
MALDI-PSD
While accurate molecular weight information combined with
chemical information can lead to the determination of a
chemical formula [16,17], MALDI mass spectra provide little
additional structural information. However, by application of
the PSD method, a considerable amount of structural informa-
tion can be obtained.
It has been demonstrated above that derivatisation of 3-
oxo-ꢀ4 steroids to GP hydrazones increases their “pseudo-
molecular” ion current by a factor of 500; thus, PSD exper-
iments were subsequently performed on derivatised testos-
terone and related steroids. Shown in Fig. 3b–e are MALDI-
PSD spectra of 5 ng aliquots of GP hydrazones of: (b) testos-
terone (IV), (c) [19,19,19-2H3]testosterone (V), (d) norgestrel
(VI) and (e) DHEA (III). Each of these spectra show fragment-
Initial PSD experiments were performed on the [M + H]+
ion of underivatised testosterone. Using 100 ng of sample, the

steroids 7 1 ( 2 0 0 6 ) 42–53
47
Fig. 3 – MALDI-PSD of the: (a) [M + H]+ ion of testosterone (m/z 289); and the [M]+ ions of GP hydrazones of—(b) testosterone
[M]+ (m/z 422) (IV), (c) [19,19,19-2H3]testosterone [M]+ (m/z 425) (V), (d) norgestrel [M]+ (m/z 446) (VI), (e) DHEA [M]+ (m/z 422)
(III) and (f) pregnanolone [M]+ (m/z 452) (IX). One hundred nanograms of underivatised testosterone or 5 ng of derivatised
steroid was loaded on the MALDI target plate. See Schemes 2–4 for a description of the fragment-ion nomenclature.
ions at 80 (␴ꢀ1), 93 (␴2), 94 (␴1ꢀ ), 120 (ꢀ␴3), 135 (ꢀ␴4) and 137 (␴ꢀ4)
of the GP hydrazone group. These patterns of fragment-
ions allow the simple identification of GP hydrazones. In the
PSD spectrum of testosterone GP hydrazone (IV), a triad of
fragment-ions is observed at m/z 177, 163 and 151 (Fig. 3b).
which are characteristic of the GP hydrazone group (Table 2;
Schemes 3 and 4). In addition, neutral losses of 79 and
107 Da from [M]+ are observed, which are also characteristic

Table 2 – Characteristic fragment-ions observed in the PSD spectra of oxosteroid mono-GP hydrazone [M]+ ions
ꢀ
ꢀ
ꢀ
ꢀ
␴
4
ꢀ
Precursor and fragment-ions
M
␴
␴
␴
␴
␴
#b1-12 #b3-CH3–CH *b1-12 *b3-CH3–CH *b2
M − 107 M − 79 Other major ions
2
3
1
2
4
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
A4-3,17-dione GP hydrazone (I)/(II)
A5-3␤-ol-17-one GP hydrazone (III)
A4-17␤-ol-3-one GP hydrazone (IV)
[19,19,19-2H3]A4-17␤-ol-3-one GP hydrazone
(V)
18-Homo-E4-17␣-ethynyl-17␤-ol-3-one GP
hydrazone (VI)
P4-3,20-dione GP hydrazone (VII)/(IIX)
5␤-P-3␣-ol-20-one GP hydrazone (IX)
C4-24S-ol-3-one GP hydrazone (X)
C4-7␤-ol-3-one GP hydrazone (XI)
C4-7␣,25-diol-3-one GP hydrazone (XII)
C4-7␣,27-diol-3-one GP hydrazone (XIII)
[16,16,17(20),22,22,23,23-2H7]C4-7␣,26-diol-3-
one GP hydrazone (XIV)
420
422
422
425
80
80
80
80
93
93
93
93
94
94
94
94
120
120
120
120
135a
137
137
137
137
123
–
123
126
135a
–
135a
135b
151–
–
163–
–
177–
–
177
180
313
315
315
318
341
343
343
346
269h
√
√
√
√
√
√
271k
135
135a
135b
253i
√
√
√
√
√
√
√
√
√
151
154
163
√
√
√
163f
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
446
80
93
94
120
135c
135a
137d
109–
135c
137d
163g
163g
339
367
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
448
452
534
534
550
550
557
80
80
80
80
80
80
80
93
93
93
93
93
93
93
94
94
94
94
94
94
94
120
120
120
120
120
120
120
137
137
137
137
137
137
137
123
–
135a
–
135a
151e
151e
151e
151e
151
–
163
–
163
177–
–
341
–
369
373
455
455
471
471
478
125j
125j
√
√
135
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
135a
123
123
123
123
123
151
151e
151e
151e
151e
177–
177–
177–
177–
177–
427
427
443
443
450
√
√
√
√
√
135
135
135
135
179–
179–
179–
179–
√
√
√
√
See Schemes 2–4 for a description of the fragmentation nomenclature; ‘ ’ signifies the ion to be present, ‘–’ indicates the ion to be absent.
a
135 corresponds to ꢀ␴ and #b3-CH3–CH.
4
b
135 corresponds to ꢀ␴ and #b3-CD3–CH.
4
c
135 corresponds to ꢀ␴ and #b3-H–CH.
4
d
e
f
ꢀ
137 corresponds to ␴ and *b1-12.
4
151 corresponds to #b3-CH3–CH and *b1-12.
163 corresponds *b3-CD3–CH.
163 corresponds to *b3-H–CH and *b2.
269 corresponds to ꢀe.
g
h
i
253 corresponds to ꢀe-18.
j
125 corresponds to *D2.
k
271 corresponds to ꢀe.

steroids 7 1 ( 2 0 0 6 ) 42–53
49
Scheme 3 – Fragmentation of 3-oxo-ꢀ4 steroid GP hydrazones. Fragment-ions derived from the [M − 79]+ intermediate ion
are designated with an asterisk, e.g. *b1-12. Fragment-ions derived from the [M − 107]+ intermediate ion are designated with
a superscript hatch, e.g. #b1-12. A prime to the right of a fragment-describing letter indicates that cleavage proceeds with
H-atom transfer from the neutral leaving-group to the fragment-ion, e.g. ␴ꢀ1. A prime to the left of the fragment-describing
letter indicates that cleavage proceeds with H-atom transfer from the fragment-ion to the neutral leaving group, e.g. ꢀ␴3. The
*b1-12, *b2 and *b3-CH3–CH fragment-ion structures are drawn with charge located on the side-chain, alternative structures
could be envisaged with the side-chain forming part of a bicyclic ring.
These ions correspond to the B-ring fragments *b2, *b3-28 (*b3-
CH3–CH) and *b1-12 (Scheme 3). Incorporation of three deu-
terium atoms on C-19 shifts the triad to m/z 180, 163 and 154
(Fig. 3c), while the absence of C-19 results in the elimination of
ions at m/z 177/180 and 151/154, but the ion at m/z 163 remains
(Fig. 3d). These results indicate that the *b2 ion corresponds
to the A-ring plus C-19 (i.e. the C-10 ␤ substituent) and C-6.
The *b3-28 ion in testosterone GP hydrazone (IV) becomes a
*
*b3-31 ion in its [19,19,19-2H3] analogue (V), and a b3-14 ion
in norgestrel GP hydrazone (VI). This indicates that in testos-

50
steroids 7 1 ( 2 0 0 6 ) 42–53
Scheme 4 – Formation of: (a) ꢀe-18 and ꢀe fragment-ions from DHEA 17-GP hydrazone (III) and (b) *D2 fragment-ions from
pregnanolone 20-GP hydrazone (IX).
*
terone GP hydrazone, the b3-28 fragment-ion consists of the
A-ring plus C-6 and C-7, but has lost the C-19 methyl group
(i.e. the C-10 ␤ substituent), a C-atom possibly from C-10 and
a H-atom, possibly from the A-ring, but or more likely from
group is maintained in this ion which probably corresponds
to A-ring plus C-19 but minus C-5. In norgestrel GP hydrazone,
the absence of the C-19 group results in the b1-12 ion being
shifted to 137 Da. A satellite series of ions are observed in the
spectra of 3-oxo-ꢀ4 steroid GP hydrazones displaced by 28 Da
from the *b-ion series. The satellite series (#b-ion series) differ
from the main series in that a CO group has been additionally
eliminated (Scheme 3). In DHEA GP hydrazone (III) (Fig. 3e), the
*
*
C-6 [10,18]. The b3-28 fragment-ion can thus be described as
*b3-CH3–CH. The *b1-12 ion increases in mass by 3 Da in spec-
trum of [19,19,19-2H3] testosterone GP hydrazone as compared
to the monoisotopic isotopomer, this suggests that the C-19

steroids 7 1 ( 2 0 0 6 ) 42–53
51
absence of a 3-oxo-ꢀ4 GP function results in the exclusion of
the 177, 163 and 151 triad of ions from the PSD spectrum. How-
ever, additional ions are observed at m/z 271 and 253 which
correspond to the steroid- and dehydrated steroid ring sys-
tem, respectively (Scheme 4).
between C-6 and C-7 which will inhibit *b3-CH3–CH ion forma-
tion.
In Fig. 3, it is evident that 17-oxosteroid GP hydrazones
fragment in a different manner to 3-oxo-ꢀ4 GP hydrazones.
In the PSD spectrum of DHEA (A5-3␤-ol-17-one) 17-GP hydra-
zone (III) (Fig. 3e) intense fragment-ions are observed at m/z
271 (ꢀe) and 253 (ꢀe-18) and these ions correspond to the steroid
ring system, and the dehydrated steroid ring system, respec-
tively (Scheme 4a; Table 2). The PSD spectrum of androstene-
dione (A4-3,17-dione) mono-GP hydrazone (I/II) also shows an
intense ion corresponding to the steroid ring system, but in
this case at m/z 269 (ꢀe) (Table 2). When the GP hydrazone is
at C-20, as in pregnanolone (5␤-P-3␣-ol-20-one) GP hydrazone
(IX), the PSD spectrum shows an intense fragment-ion at m/z
125, this corresponds to the *D2 fragment and is characteristic
of C-20 oxosteroid GP hydrazones (Fig. 3f; Table 2; Scheme 4b).
The value of the derivatisation approach in a biological
application is illustrated in Fig. 4d, which shows the PSD spec-
trum of a hydroxycholesterol, isolated from the equivalent
of 0.1 mg of rat brain, following treatment with cholesterol
Elongation of the C-17 side-chain, as in the oxysterol
derivative C4-24S-ol-3-one GP hydrazone (X), does not cause
a major change in the pattern of fragment-ions characteristic
of 3-oxo-ꢀ4 steroid GP hydrazones, other than attenuation of
the *b2 fragment-ion (Fig. 4a; Table 2). However, incorporation
of a hydroxyl group at C-7 changes the pattern of fragment-
ions observed (cf. Fig. 4a with Fig. 4b and c; Table 2). As in the
PSD spectra of other 3-oxo-ꢀ4 steroid GP hydrazones (with a
C-19 methyl group), the *b1-12 ion is observed at m/z 151 in
the spectra of 7-hydroxy-ꢀ4-3-oxo steroid GP hydrazones, but
*
fragment-ions corresponding to b3-CH3–CH which would be
expected to be shifted in m/z from 163 to 179 due to the incor-
poration of a hydroxyl group on C-7 are not observed. This
can be explained by the propensity of the C-7 hydroxyl group
to be eliminated as water, and thereby generate a double bond
Fig. 4 – MALDI-PSD spectra of GP hydrazones of: (a) C4-24S-ol-3-one [M]+ (m/z 534) (X), (b) C4-7␤-ol-3-one [M]+ (m/z 534) (XI),
(c) C4-7␣,25-diol-3-one [M]+ (m/z 550) (XII) (5 ng of each steroid was loaded on the MALDI target plate) and (d)
24S-hydroxycholesterol isolated from rat brain, oxidised with cholesterol oxidase and derivatised with GP reagent. See
Scheme 3 for a description of the fragment-ion nomenclature.

52
steroids 7 1 ( 2 0 0 6 ) 42–53
oxidase and GP hydrazine. Accurate mass measurement per-
formed in an initial MALDI spectrum of the GP hydrazone
gave an elemental formula C34H52N3O2, this and the PSD spec-
trum (Fig. 4d) are consistent with the brain steroid being pre-
dominantly 24-hydroxycholesterol. 24S-Hydroxycholesterol is
known to be the major oxysterol in brain [20].
sists of the A-ring plus C-6 and C-7 but minus the C-19 methyl
group, the C-10 carbon and a H-atom from C-6 (Scheme 3).
This structure is analogous to that proposed by Williams et
al. for the fragment-ion at m/z 109 in the CID spectrum of
protonated testosterone (Scheme 2) (cf. Fig. 3a) [18]. We have
previously described the ion at m/z 163 as *b2-CH3–H + 2H, and
proposed it to consist of the A-ring plus C-6, minus the C-
19 methyl group and the 6␤-hydrogen atom, and with two
hydrogen atoms transferred to the fragment-ion from the neu-
tral leaving group [15]. With data from the present study, it
is impossible to differentiate between *b3-CH3–CH and *b2-
CH3–H + 2H fragment-ions, and either could be the structure
of the ion at m/z 163. It should be noted that for 7-hydroxy-ꢀ4-
3-oxo steroid GP hydrazones, a *b3-CH3–CH ion should appear
at m/z 179, while a *b2-CH3–H + 2H ion should appear at m/z
163; unfortunately, intense peaks were not observed at either
of these m/z values in the PSD spectra of the 7-hydroxy-ꢀ4-3-
oxo steroid GP hydrazones studied.
4.
Discussion
3-Oxo-ꢀ4, 17-oxo and 20-oxo steroids can be derivatised to GP
hydrazones. The derivatisation reaction is simple, and even if
the reaction does not go to completion, gives an enhancement
in MALDI precursor-ion current of at least a factor of 100. By
performing derivatisation at room temperature, the reaction
is applicable to labile structures such as 7-hydroxy-ꢀ4-3-oxo.
Further, conversion of 3␤-hydroxy-ꢀ5 steroids to their 3-oxo-
ꢀ4 analogues makes them amenable to derivatisation with the
GP reagent, and subsequent MALDI analysis. In this way, many
steroids of biological interest such as oxysterols, and choles-
terol itself, can be analysed by MALDI at high sensitivity to
give molecular weight information, and by MALDI-PSD to give
structural information. This is in contrast to the untreated
steroids which are poorly ionised by MALDI and often do not
give [M + H]+ ions. This is illustrated in Fig. 4d which shows
the MALDI-PSD spectrum of 24S-hydroxycholesterol isolated
from rat brain, oxidised with cholesterol oxidase and deriva-
tised with GP reagent.
In conclusion, with the growing interest in metabolomics
and lipidomics [30], methods to improve the ionisation prop-
erties of neutral thermally labile molecules are desirable.
Derivatisation of ketone or aldehyde groups to Girard hydra-
zones [31] will provide one such solution; in the current study,
we have demonstrated the utility of such a derivatisation
method for the MALDI analysis of oxosteroids.
Acknowledgements
MALDI-MS has previously been used in lipid analysis, par-
ticularly in the area of phospholipids [21–24]. However, few
previous studies have been made using MALDI for steroid
analysis. Schiller et al. [25,26] have analysed cholesterol
in bronchoalveolar lavage fluid, spermatozoa and seminal
plasma using MALDI-TOF, while Rujoi et al. [27] have anal-
ysed cholesterol in lipid rafts. Cholesterol was found to give
an ion at m/z 369 corresponding to the [M + H−H2O]+ moiety
rather than the [M + H]+ ion. In previous studies, Griffiths et
al. [15,28,29] have shown that cholesterol can be converted
to its 3-oxo-ꢀ4 analogue, derivatised to its 3-GP hydrazone
and then ionised by MALDI. Griffiths’ studies were performed
on a MALDI-quadrupole-TOF instrument and their success
inspired the present work.
This work was supported by The School of Pharmacy, Univer-
sity of London; the UK Biotechnology and Biological Sciences
Research Council (BBSRC grant no. BB/C515771/1); the Swedish
Research Council (grant no. 03X-12551).
r e f e r e n c e s
[1] Waller GR, editor. Biochemical applications of mass
spectrometry. New York: Wiley; 1972.
[2] Barber M, Bordoli RS, Sedgwick RD, Tyler AN. Fast atom
bombardment of solids (FAB): a new ion source for mass
spectrometry. J Chem Soc Chem Commun 1981:325–7.
[3] Yamashita M, Fenn JB. Electrospray ion source. Another
variation on the free-jet theme. J Phys Chem
1984;88:4451–9.
[4] Karas M, Bachmann D, Bahr U, Hillenkamp F.
Matrix-assisted ultraviolet laser desorption of non-volatile
compounds. Int J Mass Spectrom Ion Proc 1987;78:53–
68.
[5] Clauser KR, Baker P, Burlingame AL. Role of accurate mass
measurements ( 10 ppm) in protein identification
strategies employing MS or MS/MS and database
searching. Anal Chem 1999;71:2871–82.
[6] Kaufmann R, Kirsch D, Spengler B. Sequencing of peptides
in a time-of-flight mass spectrometer: evaluation of
postsource decay following matrix-assisted laser
desorption ionisation (MALDI). Int J Mass Spectrom Ion
Processes 1994;131:355–85.
In the current study, we have shown that MALDI mass spec-
tra of steroid GP hydrazones can be obtained from 500 pg of
steroid loaded onto the MALDI target plate (Fig. 1b), and that
structurally informative MALDI-PSD spectra can be obtained
from 5 ng loaded onto the plate (Figs. 3 and 4; Table 2). 3-Oxo-
ꢀ4 steroids give a characteristic pair of fragment-ions b1-12
*
*
(m/z 151) and b3-CH3–CH (m/z 163), but the presence of a 7-
hydroxy group results in attenuation of the *b3-CH3–CH ion
(now shifted to m/z 179), and relative enhancement of the *b1-
ꢀ
12 ion. 17-Oxo steroid GP hydrazones give a characteristic e
ion corresponding to the steroid ring system (Scheme 4a), e.g.
at m/z 271 in A5-3␤-ol-17-one 17-GP hydrazone (III), and 269 in
A4-3,17-dione mono-GP hydrazone (I/II), while 20-oxo steroid
GP hydrazones give a characteristic fragment-ion correspond-
ing to *D2 at m/z 125 (Scheme 4b).
There is some debate as to the structure of the ion at m/z
163 in the spectra of 3-oxo-ꢀ4 steroid GP hydrazones. Here,
we describe the ion as a *b3-CH3–CH fragment, which con-
[7] Kaufmann R, Chaurand P, Kirsch D, Spengler B.
Post-source decay and delayed extraction in
matrix-assisted laser desorption/ionization-reflectron
time-of-flight mass spectrometry. Are there trade-offs?
Rapid Commun Mass Spectrom 1996;10:1199–208.

steroids 7 1 ( 2 0 0 6 ) 42–53
53
[8] Kaufmann R, Wingerath T, Kirsch D, Stahl W, Sies H.
Analysis of carotenoids and carotenol fatty acid esters by
matrix-assisted laser desorption ionization (MALDI) and
MALDI-post-source-decay mass spectrometry. Anal
Biochem 1996;238:117–28.
[9] Griffiths WJ, Jonsson AP, Liu S, Rai DK, Wang Y.
Electrospray and tandem mass spectrometry in
biochemistry. Biochem J 2001;355:545–61.
[20] Smith LL, Ray DR, Moody JA, Wells JD, Van Lier JE.
24-Hydroxycholesterol levels in human brain. J Neurochem
1972;19:899–904.
[21] Harvey DJ. Matrix-assisted laser desorption/ionization
mass spectrometry of phospholipids. J Mass Spectrom
1995;30:1333–46.
[22] Schiller J, Arnhold J, Benard S, Mu¨ ller M, Reichl S, Arnold
K. Lipid analysis by matrix-assisted laser desorption mass
spectrometry: a methodological approach. Anal Biochem
1999;267:45–56.
[23] Al-Saad KA, Zabrouskov V, Siems WF, Knowles NR,
Hannan RM, Hill Jr HH. Matrix-assisted laser
[10] Griffiths WJ. Tandem mass spectrometry in the study of
fatty acids, bile acids and steroids. Mass Spectrom Rev
2003;22:81–152.
[11] Leinonen A, Kuuranne T, Kostiainen R. Liquid
chromatography/mass spectrometry in anabolic steroid
analysis—optimization and comparison of three ionization
techniques: electrospray ionisation, atmospheric pressure
chemical ionization and atmospheric pressure
photoionization. J Mass Spectrom 2002;37:693–8.
[12] Burkard I, Rentsch KM, von Eckardstein A. Determination
of 24S- and 27-hydroxycholesterol in plasma by
high-performance liquid chromatography–mass
spectrometry. J Lipid Res 2004;45:776–81.
[13] Brooks CJW, Cole WJ, Lawrie TDV, MacLachlan J, Borthwick
JH, Barrett GM. Selective reactions in the analytical
characterisation of steroids by gas chromatography–mass
spectrometry. J Steroid Biochem 1983;19:189–201.
[14] Wheeler OH. The Girard reagents. J Chem Educ
1968;45:435–7.
desorption/ionization time-of-flight mass spectrometry of
lipids: ionization and prompt fragmentation patterns.
Rapid Commun Mass Spectrom 2003;17:87–96.
[24] Al-Saad KA, Siems WF, Hill HH, Zabrouskov V, Knowles
NR. Structural analysis of phosphatidylcholines by
post-source decay matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry. J
Mass Spectrom 2003;14:373–82.
[25] Schiller J, Hammerschmidt S, Wirtz H, Arnhold K. Lipid
analysis of bronchoalveolar lavage fluid (BAL) by
MALDI-TOF mass spectrometry and 31P NMR spectroscopy.
Chem Phys Lipids 2001;112:67–79.
[26] Schiller J, Arnhold J, Glander HJ, Arnold K. Lipid Analysis of
human spermatozoa and seminal plasma by MALDI-TOF
mass spectrometry and NMR spectroscopy: effects of
freezing and thawing. Chem Phys Lipids 2000;106:145–56.
[27] Rujoi M, Jin J, Borchman D, Tang D, Yappert MC. Invest
Ophthalmol Vis Sci 2003;44:1634–42.
[28] Griffiths WJ, Liu S, Alvelius G, Sjo¨vall J. Derivatives
revisited: analysis of neutral steroids by ESMS. In:
Proceedings of the 50th ASMS conference on mass
spectrometry and allied topics. 2002 [WOD am Slot: 11:55].
[29] Alvelius G, Sjo¨vall J, Liu S, Griffiths WJ. Oxysterols in brain:
purification, derivatisation and analysis. In: Ashcroft AE,
Brenton G, Monaghan JJ, editors. Advances in mass
spectrometry, vol. 16. Elsevier; 2004 [CD-ROM Suppl.].
[30] Han X, Gross RW. Shotgun lipidomics: electrospray
ionization mass spectrometric analysis and quantitation
of cellular lipidomes directly from crude extracts of
biological samples. Mass Spectrom Rev 2005;24(3):367–412.
[31] Naven TJP, Harvey DJ. Cationic derivatization of
oligisaccharides with Girard’s T reagent for improved
performance in matrix-assisted laser desorption/ionization
and electrospray mass spectrometry. Rapid Commun Mass
Spectrom 1996;10:829–34.
[15] Griffiths WJ, Liu S, Alvelius G, Sjo¨vall J. Derivatisation for
the characterisation of neutral oxo-steroids by electrospray
and matrix-assisted laser desorption/ionisation tandem
mass spectrometry: the Girard P derivative. Rapid
Commun Mass Spectrom 2003;17:924–35.
[16] Beynon JH. Qualitative analysis of organic compounds by
mass spectrometry. Nature 1954;174:735–7.
[17] Russell DH, Edmondson RD. High-resolution mass
spectrometry and accurate mass measurements with
emphasis on the characterization of peptides and proteins
by matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry. J Mass Spectrom
1997;32:263–76.
[18] Williams TM, Kind AJ, Houghton E, Hill DW. Electrospray
collision induced dissociation of testosterone and
testosterone hydroxy analogs. J Mass Spectrom
1999;34:206–16.
[19] Shackleton CHL, Chuang H, Kim J, de la Torre X, Segura J.
Electrospray mass spectrometry of testosterone esters:
potential for use in doping control. Steroids 1997;62:523–9.