вход по аккаунту



код для вставкиСкачать
J. Mass Spectrom. 35, 258–264 (2000)
Mapping of surrogate markers of cellular
components and structures using laser
desorption/ionization mass spectrometry
John M. Koomen,1 Markus Stoeckli2 and Richard M. Caprioli2 *
Laboratory for Biological Mass Spectrometry, Chemistry Department, Texas A&M University, College Station, Texas
77842, USA
2 Vanderbilt University, Mass Spectrometry Research Center, 812 MRBI, Nashville, Tennessee 37232-6400, USA
Laser desorption/ionization mass spectrometry (LDI-MS) has been used to assess the potential of using surrogate
markers, bound to cellular structures containing nucleic acids, to image or map the position of these structures
within biological samples. In this study, organic dyes were used as markers because of their established use in
the histochemical marking of nucleic acids, and also because they are amenable to LDI-MS. Eight cationic dyes
were tested and all could be desorbed from nucleic acid samples without additional matrix after specifically
binding to these molecules. Methylene Blue was the best of these based on its sensitivity to detection by LDI-MS
and the fact that it can be washed from the tissue in areas where it was not specifically bound to provide
low-intensity background signals. Experiments are reported which characterize the MY ion signal obtained
from Methylene Blue with regard to sensitivity, reproducibility and possible use for quantitation. This dye was
used to map (with a lateral resolution of 25 µm) several nucleic acid-containing samples spotted on prepared
surfaces, and to image the location of nucleic acids in two model tissues, retinal vertical sections and thyroid
whole mount sections. Copyright  2000 John Wiley & Sons, Ltd.
KEYWORDS: laser desorption/ionization mass spectrometry; cellular structure; nucleic acids; surrogate markers
The use of dyes as surrogate markers of biological macromolecules, such as those employed in histochemistry
and immunohistochemistry, have been important tools in
assessing the morphology and function of different cell
types. Using light, fluoresence and, most recently, confocal microscopy, these techniques have been refined to
yield high-resolution data concerning the distribution of
cell types in tissues, information on intracellular structure and molecular localization. For the most part, the
compounds used are not natural ligands but rather are synthetic organic compounds having chemical features which
allows them to bind to nucleic acids or proteins. Such
binding of dyes is not structurally specific in that the dyes
tend to bind to most molecules in a specific class to a
more or less similar degree, e.g. Coomassie Brilliant Blue
generally binds to proteins and Methylene Blue to nucleic
acids. Many of these interactions are non-covalent and the
dye can be washed out under a given set of conditions,
albeit in some cases with much difficulty.
Laser desorption/ionization (LDI), particularly matrixassisted LDI (MALDI), mass spectrometry (MS) is a
sensitive method for the analysis of biological molecules
* Correspondence to: R. M. Caprioli, Vanderbilt University, Mass
Spectrometry Research Center, 812 MRBI, Nashville, Tennessee
37232-6400, USA.
E-mail: [email protected]
Copyright  2000 John Wiley & Sons, Ltd.
of wide mass range. Since Karas et al.1 first described
the MALDI technique and instrumentation, the field has
rapidly expanded with the development of protocols for
analyzing oligonucleotides, lipids, oligosaccharides, proteins and peptides. Recent approaches to the analysis
of biological compounds include direct measurements of
samples by molecular desorption and detection of ions
from tissues,2 electrophoretic gels3 and other prepared
surfaces. Furthermore, MALDI-MS has been used to ‘biochemically map’ large areas of tissues (>2 mm2 / for their
molecular contents.4 Other MS techniques, such as secondary ion mass spectrometry, have also been used to
image surfaces of samples, with a considerable body of
information and techniques published for analysis of elements and small molecules.5 – 8 Hercules et al.9 demonstrated that scanning LDI instruments can be used for
the quantitation of dyes (e.g. Gentian Violet or Brilliant
Green) on nitrocellulose membranes.
In the work reported here, we combined the use of dyes
that bind to structural features of cells such as nucleic
acids with LDI-MS to image or map their location in tissue
samples. The signal of the dye molecule itself is measured,
with the dye acting as a surrogate marker of the location
of the nucleic acid-containing structure. An advantage of
the dye for these initial studies is that it allows the optical image to be compared with the MS image. Under the
best conditions, the dye would be efficiently desorbed to
produce an MC or [M C H]C without use of additional
matrix, show little or no fragmentation due to laser excitation, specifically bind non-covalently to a class or subclass
Received 23 August 1999
Accepted 30 November 1999
of cellular molecules or structures, could be efficiently
washed from non-bound sites to achieve low background
and selectively penetrate into cells without disrupting cell
We studied eight cationic dyes which bind to nucleic
acids in vivo and in vitro. Acridine Orange (3,6-bis(dimethylamino)acridine) has been used in vivo to label nucleic
acids by intercalation or binding to the phosphate backbone. Although in vivo Acridine Orange is not able to
intercalate into the chromosomal DNA,10 it differentially
labels other intracellular nucleic acids, such as RNA,
and nucleoli, and can be used for nuclear staining for
fixed tissue. It is also of interest because of its aggregation into crystalline lattices around metal halide clusters
and base pairs in DNA.11,12 Crystal Violet (hexamethylpararosaniline chloride) is used extensively as a bacterial
stain because it selectively stains Gram-positive bacteria, and is an excellent stain for chromatin and consequently nuclei.13,14 DAPI (40 ,6-diamidine-2-phenylindole)
has been commonly used as a fluorescent nuclear stain
in vivo. Larsen et al.15 indicated that DAPI binds to
DNA in the minor groove, while Williamson16 showed
this binding was adjacent to ‘AT’-rich regions; in addition, Tanious et al. showed that DAPI binds to RNA
by intercalation in ‘AU’ sequences.17 Ethidium bromide (2,7-diamino-9-phenyl-10-ethylphenanthridium bromide) has been used to stain nucleic acids in gels
by intercalation (and possibly ionic interactions) with
the phosphate backbone. Waring18 showed that ethidium bromide complexes with nucleic acids at a ratio of
one molecule per 4–5 bases. This dye is impermeant to
living cells19 and is used only for fixed or damaged tissues. Hoechst 33258 (20 -(4-Hydroxyphenyl)-5-(4-methyl1-piperazinyl)-20 ,50 -bi-1H-benzimidazole) (or Bisbenzomide) and Hoechst 33342 (20 -(4-ethoxyphenyl)-5-(4-methyl-1-piperazinyl)-20 ,50 -bi-1H-benzimidazole) have been
used as in vivo histochemical markers for DNA and have
found wide use in flow cytometry and fluorimetric determinations of cellular DNA content or structures20,21 and
H33258 binds to ‘AT’-rich DNA in the minor groove;22
H33342 is presumed to behave in the same manner. Furthermore, H33342 has been reported to be highly cell
permeant and its fluorescence survived fixation in many
tissues.23 Methylene Blue has been used in a number
of histological applications in bacteriology, cytology and
hematology14,24 and stains nuclei very effectively. Nuclear
Yellow (20 -(4-sulfamylphenyl)-7-[6-(4-methylpiperazino)2-benzimidazolyl]benzimidazole) is similar to the two
Hoechst dyes and has been used in retinal research to
label cells selectively by their differential uptake and thus
differential nuclear staining.25
Molecular Probes (Eugene, OR, USA), Zeta Probe quaternary amine membrane from Bio-Rad Laboratories (Hercules, CA, USA) and NA49 carboxymethylcellulose membrane from Schleicher and Schuell (Keene, NH, USA). All
dyes used in our experiments are highly absorbing at the
laser wavelength of 337 nm.
Nano-spotting techniques
Reference stain spots were generated by applying the
dyes to the surface using a nano-spotting technique.
The apparatus was built by fitting Polymicro Technologies (Phoenix, AZ, USA) silica capillary (350 µm o.d.,
20–25 µm i.d.) into Rainin (Woburn, MA, USA) 200-34
Teflon tubing (0.3 mm i.d., 1.5 mm o.d.) and attaching it
to a 5 µl Osge (Ringwood, Australia) GC syringe, 0.3 mm
i.d. The syringe was placed on a WPI (Sarasota, FL,
USA) SP200iw syringe pump. Optimum conditions were
a flow-rate of 1 nl s 1 and the reproducibility was best at
a minimum volume of 15 nl.
Staining procedures
Samples spotted on membrane supports were stained in
two different ways: by immersion or by overspotting.
Immersion staining used three 20 ml vials, each filled
with 15 ml of solution. The first vial held the dye solution (2 mg ml 1 aqueous), the second water and the last
aqueous 10% methanol. The edge of the membrane was
gripped with forceps and immersed in these staining and
destaining solutions until nuclear staining was visible. In
the overspotting method, 50 µl of aqueous stain solution
were applied uniformly across the surface of the membrane, including the sample spots. When staining became
visible, the dye solution was removed and the sample
washed several times with 50 µl of water and 50 µl of
aqueous 10% methanol.
Acquisition of dye mass spectra
Single spectra for these dyes were acquired with a
PerSeptive Biosystems (Framingham, MA) Voyager Elite
MALDI time-of-flight (TOF) mass spectrometer with
delayed extraction, a nitrogen laser with excitation wavelength 337 nm and a 1.9 m flight tube. Modifications to
the instrument to obtain a laser spot size of ¾25 mm in
diameter has been described previously.4 Mass calibration
was done by using internal standards. For optimum resolution, spectra were obtained using a 20 kV source voltage,
72% grid voltage and 0.08% guide wire voltage in the
reflectron mode with detection of positive ions. The mass
range was set to a range of m/z 1–1200, with a sampling
rate of 2 ns, and at least 10 spectra were averaged per
image point (3 Hz laser repetition rate).
Chemicals and synthetic surfaces
MS imaging
Methylene Blue (CI 52015), Acridine Orange (CI 46005),
Nuclear Yellow, DAPI and ethidium bromide were purchased from Sigma Chemical (St Louis, MO, USA), Crystal Violet (CI 42555) from Aldrich Chemical (Milwaukee, WI, USA), Hoechst 33258 and Hoechst 33342 from
The MALDI Image Tool software26 enabled automated
data collection over a large sample area. Data were collected as peak area values for 10 averaged spectra at every
point; the distance between both X and Y points was set
to either 18, 24 or 27 µm. Peak areas were recorded at
Copyright  2000 John Wiley & Sons, Ltd.
J. Mass Spectrom. 35, 258–264 (2000)
every laser spot in a 10 m/z window around the calculated
MC or [M C H]C value for the dye and were viewed as a
2-D map in the MS Image Tool program (peak area was
represented by a selectable color gradient). The relatively
large mass window was chosen in order to detect signals
which might shift due to charging effects of the membrane. SigmaPlot 4.0 (Jandel Scientific) was also used for
graphical and numerical analysis.
Each data point in an image is defined as one laser
spot (the sampled area at each of the X and Y steps).
The values of the dye peak area data were first plotted
against the X and Y steps in SigmaPlot’s 3-D mesh format
and the background level was determined. Background
corrections were performed in Microsoft Excel, so that
peak areas corresponding to specifically bound dye could
be determined. The average peak area per laser spot was
obtained by dividing the total peak area of the dye by the
number of laser spots over the sample (equivalent to the
number of peak area values above background).
Light/fluorescence microscopy
After MS mapping, digital images were acquired with a
C5810 Color Chilled CCD camera (Hamamatsu Photonics Deutschland) mounted on an Olympus Vanox AHBT
light microscope fitted with epifluorescence. Fluorescent
images were acquired using the DAPI filter set for spots
of Hoechst 33258 and 33342, DAPI and Nuclear Yellow.
Images were processed and saved through Adobe Photoshop 5.0.
Mapping cationic dye staining of nucleic acids
A single-stranded DNA oligonucleotide (50 -CCG CTC
at a concentration of 500 mM in 10 mM Tris–HCl and
1 mM EDTA (pH 8) at 60 ° C (Genosys Biotechnologies, Woodlands, TX, USA) and dispensed in 7.5 pmol
aliquots. This 30mer ssDNA oligo was used to test all
eight of the cationic DNA dyes. Aliquots were deposited
on Zeta Probe membrane and stained with aqueous dye
solutions of 1 mg ml 1 . The sample aliquots were stained
by immersion as described above. The membranes were
fixed to the MALDI plate with double-sided Scotch tape
and one spot of each dye was imaged at 27 µm resolution
averaging 10 spectra, using the same laser intensity.
The reproducibility of the peak area signal of Methylene
Blue was determined by staining sets of 12.5 pmol spots
of 30mer ssDNA oligomers. The total Methylene Blue
peak area above background and the average Methylene
Blue peak area above the background per laser spot were
calculated and compared for three of the oligomer spots.
The total desorption of Methylene Blue signal at single
laser spots was measured from another 7.5 pmol spot of
the 30mer ssDNA. Averaging 32 laser shots per spectrum,
the same area of the MALDI plate was repeatedly struck
with the laser and the spectra recorded. The data was
acquired until the signal-to-noise ratio was <2 : 1 and the
peak intensity was consistently below 1000 counts. The
peak area values at m/z 284, the MC peak for Methylene
Blue, were recorded for the calculation of the signal from
total desorption.
Copyright  2000 John Wiley & Sons, Ltd.
Six concentrations of Methylene Blue (2, 1, 0.5, 0.1,
0.01 and 0.001 mg ml 1 in 15 nl aliquots) in Milli-Q
water were prepared and spotted non-sequentially on
NA49 CM cellulose membrane. The area including the
spots was imaged at 27 µm resolution and the peak area
was recorded. Four concentrations of the 30mer ssDNA
oligo in Milli-Q water were spotted on the same piece of
Zeta Probe membrane, stained by immersion in 15 ml of
0.02% Methylene Blue and destained twice by immersion
in 15 ml of Milli-Q water. The samples were imaged as
described above.
PBK-CMV-ˇ-gal plasmid was obtained from Stratagene (La Jolla, CA, USA) and dissolved at 2.4 mg ml 1 in
10 mM Tris–HCl and 1 mM EDTA (pH 8). It was spotted
against the 32mer DNA oligomer; three 36 ng samples of
the plasmid were deposited on the Zeta Probe membrane
next to ssDNA 30mer oligos in 7.5 pmol (68 ng) aliquots
on the same membrane and stained simultaneously for
comparison. The samples were stained with Methylene
Blue by overspotting, washed twice with Milli-Q water
and desiccated.
Chromosomal DNA (Saccharomyces cerevisiae and
Moraxella bovis) were purchased from Promega (Madison, WI, USA). Chromosomes were stored in 0.8% lowmelting-point agarose in 50 mM EDTA with Orange G
at 2–4 ° C. The yeast DNA was embedded in a ratio of
1–2 mg per 5 mm agarose string (1.58 mm diameter); the
M. bovis chromosomal DNA was at half the concentration of the yeast. To extract the DNA from the agarose,
the strands were cut so that 4–8 mg of each type of chromosome could be isolated. The strands were incubated in
100 ml of 10 mM Tris–HCl and 1 mM EDTA with two
units of AgarACE (Promega, 0.224 units ml 1 ) for 10 min
at 37 ° C. After further incubation at 65 ° C for 5 min, two
more units of AgarACE were added and the solution was
incubated at 37 ° C for 30 min. The chromosomes were
precipitated by addition of 10 ml of 3 M sodium acetate
(NaOAc) (pH 5.2) and 1 ml of absolute EtOH. The DNA
was spun down at 2–4 ° C using an Eppendorf Model 5415
centrifuge at the maximum speed for 10 min. The pellet was washed with 70% EtOH, spun down again for
2 min, aspirated and placed on a Speedvac. After drying,
the chromosomes were resuspended in 20 ml of 10 mM
Tris–HCl and 1 mM EDTA. Each of these nucleic acid
samples was spotted in 15 nl aliquots on a Zeta Probe
membrane. After air drying, the sample spots were stained
by immersion in 15 ml of 0.02% aqueous cationic dye
and washed twice by immersion in 15 ml of Milli-Q
water. Stained samples were fixed to the MALDI sample plate with double-sided tape and dried by vacuum
desiccation before mapping using the mass spectrometer.
Wild-type diploid yeast was grown overnight in 100 ml
of yeast extract peptone (YP) medium supplemented with
2% (v/v) EtOH, 3% (v/v) glycerol and 1% (m/v) glucose to approximately 2 o.d. per ml, corresponding to
2 ð 1011 cells. Cell pellets were washed twice with 15 ml
of sterile water, resuspended in 30 ml of 0.1 M Tris–HCl
(pH 9.4) with 25 mM ˇ-mercaptoethanol. After incubation at 30 ° C for 5 min and aspiration of the medium, the
cells were resuspended in YP with 1 M sorbitol, 20 mM
HEPES–KOH(pH 7), 150 mM KOAc, 5 mM MgOAc, 2%
(m/v) glucose, 5 mg/o.d. zymolyase and 1% glusulase to
remove the cell walls to make spheroplasts. After incubation at 30 ° C for 30 min, the spheroplasts were pelleted in
J. Mass Spectrom. 35, 258–264 (2000)
2 min at 1750 g, resuspended in lysis buffer (0.6 M sorbitol, 10 mM HEPES–KOH, 50 mM KOAc and 0.5 mM
MgOAc) and lysed through a polycarbonate filter with
3 µm pores (Nucleopore). After centrifuging for 2 min at
1750 g, the supernatant was decanted and spun in a Beckman J2-21 centrifuge for 20 min at 10 000 g. The pellet
from this spin contains the nuclei. This nuclear fraction
was stored in lysis buffer at 20 ° C until used. Aliquots
(15 nl) of the nuclear pellet were spotted on the ZetaProbe membrane adjacent to 30mer ssDNA oligo spots
for comparison. The membrane, which had been fixed
to the MALDI sample plate, was stained and washed by
overspotting as described previously.
Preparation and staining of vertical sections of rabbit
Inferior rabbit retina close to the visual streak was fixed
in 4% (w/v) paraformaldehyde, sectioned and imbedded
in agar; 30 µm thick vertical sections were cut using a
vibrating blade microtome (VT1000 M; Leica, Deerfield,
II, USA). The sections were stored in 0.1 M phosphate
buffer containing 0.5% Triton X-100 and 0.1% sodium
azide (NaN3 ). For staining, vertical sections were placed
on a glass slide with 25 µl of 0.02% aqueous Methylene
Blue until differential staining was apparent through a
dissecting microscope at 30ð (10–25 s). The Methylene
Blue solution was removed with a pipet and the sections
were washed twice with 50 µl of Milli-Q water and blotted
on a Zeta Probe membrane that was then fixed to the
MALDI sample plate. MS maps were created using a step
size of 18 µm to give better resolution of the anatomical
Preparation and staining of rat thyroid sections
Thyroid glands were excised from adult male rats and
immediately frozen in a Revco 80 ° C freezer. Frozen
sections were cut at 5 µm thickness (Vanderbilt University Medical Center Surgical Pathology Department) and
air dried on glass slides without fixative. Sections were
stained until differentiable in 50 µl of 0.02% Methylene
Blue under a dissecting microscope and destained twice
with the same volume of Milli-Q water. During the second
destain, the sections were blotted on a Zeta Probe membrane, which was then fixed to the MALDI plate. These
samples were also imaged using an 18 µm step size.
All incorporated animal work was performed in accordance with Federal guidelines and protocols approved by
the University animal care committee (UT-HSC Houston).
MS of dyes
The structures and molecular masses of the dyes used in
the studies are shown in Fig. 1. The LDI mass spectra
of the dyes showed abundant molecular cations (Methylene Blue and Crystal Violet were supplied as chloride salts and ethidium bromide as the bromide salt) or
protonated molecular species (Acridine Orange, DAPI,
Copyright  2000 John Wiley & Sons, Ltd.
Figure 1. Cationic dyes structures and measured mass-to-charge
ratio of most intense ion signal.
H33258, H33342 and Nuclear Yellow), with no significant
fragmentation observed under the conditions used. Under
equivalent conditions, Methylene Blue, Crystal Violet and
H33342 gave the most intense signals of the group. Cocrystallization of these dyes with the usual MALDI matrices, such as ˛-cyano-4-hydroxycinnamic acid (˛CHCA)
or 2,5-dihydroxybenzoic acid (DHB), did not improve the
signal for the molecular species of these dyes, but introduced the well-known molecular and fragment ions of
these matrices into the spectrum.
Desorption of the dyes bound to biological
Laser desorption of the non-covalently bound dyes from
samples of 30mer ssDNA oligonucleotide dye complexes
on Zeta Probe membranes could be demonstrated. The first
example, shown in Fig. 2, is a spot containing 7.5 pmol
of the 30mer ssDNA stained with H33342. The spot is
¾700 µm in diameter, as shown in the light micrograph
in (A). Figure 2(B) is the image produced by measuring
the abundance of the [M C H]C ion at m/z 451.5 in a
raster of laser spots (¾25 µm diameter) with laser spot
centers located at 27 µm intervals in both the X and Y
directions. A total of about 1100 laser spots was used
to produce the image shown in Fig. 2(B) in less than
2 h, covering an area of ¾0.8 mm2 . The darker areas in
the spot represent higher relative intensity, with the white
area surrounding the spot representing background, i.e. no
discernible [M C H]C signal above the noise.
J. Mass Spectrom. 35, 258–264 (2000)
Figure 2. (A) Light micrograph and (B) corresponding MS map
of dye peak area of a 7.5 pmol spot of 30mer ssDNA oligomers
stained with Hoechst 33342. The scale bar (200 mm) applies to
both (A) and (B).
This same experiment was performed with all eight
cationic dyes, keeping the protocol as reproducible as
possible (spot size, laser intensity, dye concentration,
etc.). Figure 3 shows the total signal intensity of the
molecular species for each dye above background, i.e.
the total counts from the signal summed over the entire
spot. Although these results can only be viewed as semiquantitative because of experimental variables, one can
conclude that Methylene Blue and Crystal Violet give the
best response under these conditions. In the remaining
work, we chose Methylene Blue as the dye for study.
Desorption yield of methylene blue bound to a DNA
An estimate of the amount of Methylene Blue desorbed
from ssDNA that had been treated with the dye was made
by the irradiation of a single spot on the sample of ssDNA
for the acquisition of 320 mass spectra, each consisting of
an average of 32 individual laser shots, for a total of 10240
laser shots. These data are plotted in Fig. 4. The ion yield
for the first averaged spectra was ¾7500 counts (arbitrary
units), whereas that for the last few (e.g. spectrum 320)
was ¾950 counts with a signal-to-noise ratio of 2 : 1. From
Figure 4. Methylene Blue peak area values from total desorption
of one laser spot of ssDNA oligomers stained with Methylene
Blue. Every averaged spectrum consists of 32 single-shot
the area under this curve, assuming >90% of the dye in
that spot was desorbed after acquisition of 320 spectra,
it is estimated that about 1% of the dye was desorbed in
each of the first few shots and about 0.12% in the last few.
Effect of amount of ssDNA on methylene blue signal
Duplicate sample spots containing 0.375, 0.75, 1.5 and
7.5 pmol of ssDNA were prepared, treated with a 0.02%
aqueous solution of Methylene Blue, destained and finally
analysed by MS. The result was a linear correlation
(r D 0.999) with an average variance of š5% between
the duplicate measurements. These data indicate that,
in the range tested, the binding of the dye to ssDNA
is linear with the amount of ssDNA and that the signal from the dye is not saturating. The latter is consistent with other measurements in this laboratory indicating that Methylene Blue concentrations in similar spots
containing less than 200 pmol gave a linear desorption
Imaging of nucleic acids in prepared samples
Figure 3. Peak area per laser spot (total peak area above
background/number of laser spots) for each of the eight cationic
DNA dyes investigated. The number over each bar is the amount
of laser spots differentiated above the background (selected
based on signal-to-noise ratio >3 : 1).
Copyright  2000 John Wiley & Sons, Ltd.
Several types of nucleic acids isolated from biological
systems were imaged as prepared spots on target surfaces, including PBK-CMV ˇ-Gal plasmids, fractionated
yeast chromosomes and fractionated yeast nuclei. These
experiments were qualitative since the exact concentration
of the target molecules was not known. Figure 5 shows
the MS imaged spots taken at m/z 284.4 together with
light micrographs of these same spots. Both the plasmid
spots and the prokaryotic and eukaryotic chromosome
spots were clearly seen by MS image analysis of the
dye. However, spots of nuclei fractionated from Saccharomyces cerevisiae spheroplasts, stained with Methylene
Blue and washed, showed no dye signal above the background, although some signal could be detected around
the edges of these spots. Methylene Blue must penetrate
the nuclei and stain the nucleic acids, but a dye signal
could not be obtained from inside of the intact nuclear
J. Mass Spectrom. 35, 258–264 (2000)
Figure 5. (A) Light micrograph (left) and MS map (right) over three spots of Methylene Blue-stained plasmids. (B) Light micrograph
(left) and MS map (right) of Methylene Blue peak area over two spots of eukaryotic chromosomes from Saccharomyces cerevisiae.
(C) Light micrograph (left) and MS map (right) of Methylene Blue peak area over two spots of fractionated yeast nuclei.
Imaging of methylene blue bound to structures in
tissue sections
Sections of rabbit retina were stained with Methylene
Blue and imaged by MS for the presence of the dye. The
rabbit retina, when cut in vertical sections, has two major
and one minor nuclear layers separated by two plexiform
layers, which are made up of synapses. A dye specific for
nucleic acids, such as Methylene Blue, should be localized
to these nuclear layers as shown in Fig. 6. Image analysis
of m/z 284.4 in this section clearly shows two nuclear
layers, measuring ¾60 µm for the outer nuclear layer and
20 µm for the inner nuclear layer. The ganglion cell layer
is not well visualized in this section, although some dye
signal can be seen. All three layers have been labeled
in the MS map and the light micrograph in Fig. 6. The
absence of a significant signal from the ganglion cell layer
in the MS image is most probably due to the lower cell
density of this layer and because the sectioning may not
have bisected many of the nuclei in this cell layer. Recall
from the experiment with yeast nuclei that no signal from
the dye was observed when the nuclear membrane was
Sections of rat thyroid also were successfully mapped
by MS (data not shown). The nuclei of the cells encircling
the thyroglobulin follicles were clearly distinguishable,
while the dye signal in the proteinaceous follicles was
at background.
Comparison with protein–dye complexes
In contrast, signals from anionic dyes binding to proteins can only be observed in the presence of additional
matrix (protein spots of lysozyme and carbonic anhydrase
were stained and imaged on an NA49 carboxymethylcellulose membrane, Fig. 7). For example, a spot of carbonic
anhydrase of roughly 400 ð 500 µm containing a total of
8.0 pmol of protein has been stained with Coomassie Brilliant Blue R250 (CBR250) and washed. In contrast to the
ssDNA oligomers, protein spots treated in this way did
not yield any signal from the dye unless matrix (˛CHCA)
was added. In this example, bradykinin was added to the
matrix as an internal marker prior to electrospraying a thin
Copyright  2000 John Wiley & Sons, Ltd.
Figure 6. (A) Light micrograph and (B) MS peak area map of a
Methylene Blue-stained vertical section of rabbit retina. Three
layers of cells are mapped by the presence of Methylene Blue
staining in their nuclei: the outer nuclear layer (ONL), the inner
nuclear layer (INL) and the ganglion cell layer (GCL). The scale
bar applies to both (A) and (B).
coating over the protein spot. Figure 7 shows the light
micrograph (A) and the MS image at [M C H]C of the
dye, at m/z 805 (B). The bradykinin signal was used as a
number to qualify the matrix concentration in a given laser
spot and used to normalize the dye signal. Figure 7(C)
shows a three-dimensional plot, with the z-axis representing the relative intensity of the signal from CBR250.
Small molecules which bind non-covalently to macromolecules, or fragments of macromolecules, can be desorbed and analyzed from this bound state by LDI mass
spectrometry. These experiments used dyes commonly
J. Mass Spectrom. 35, 258–264 (2000)
Figure 7. (A) Light micrograph of 8 pmol of carbonic anhydrase
stained with Coomassie Brilliant Blue R250. (B) CBR250 peak
area of a MALDI scan, normalized by des-Arg-bradykinin matrix
marker mapped over the same area. (C) 3-D plot of normalized
peak area.
employed in histochemical studies for marking the position of specific macromolecules and cellular structures
containing these macromolecules. It has been shown that
these dyes can act as surrogate molecular markers and can
be used to image the position of nucleic acids at levels normally found in nuclei, chromosomes and other biological
systems. The goal of this work was not to replace light
microscopy in histochemical studies, but rather provide
a basis for studies involving small molecules binding to
larger molecules, such as in ligand–receptor interactions
where the ligand may be a drug or other synthetic organic
compound, or a natural metabolite whose role is to bind
to a receptor to produce or mediate biological function.
The studies were carried out on a commercial MALDI
instrument, which makes this technique applicable in a
large number of research laboratories. However, in our
experiments, dyes bound to ssDNA fragments could be
desorbed, ionized and detected by MS without the need for
another matrix (LDI). It was found that when an additional
matrix was added, no increase in dye signal was produced,
and in many cases, a decrease was actually observed. In
any case, it is clear that for the cationic compounds that
bind to ssDNA fragments, no additional matrix is needed.
The current experiments are the first steps in imaging
ligand–macromolecule complexes, and much is yet to be
done. Quantitation perhaps represents the most formidable
task in terms of establishing rigorous protocols for relating the signal intensity with the amount of ligand bound
to the macromolecule, and also from the wide diversity
of structures and complexes which make up these interactions. Sensitivity in terms of the absolute amount of complex needed to record a signal must also be determined,
although the complexity of macromolecular structures will
present special challenges.
J.M.K. gratefully acknowledges support from the UT-HSC Graduate
School of Biomedical Sciences for stipend and for a University
of Texas Presidential Scholarship. The authors thank Dr Stephen
C. Massey for the use of his fluorescence microscope and W. Sunny
Liu (Department of Ophthalmology and Visual Science at University
of Texas Medical School) for the preparation of the vertical sections
of rabbit retina.
1. Karas M, Bachmann D, Bahr U, Hillenkamp F. Int. J. Mass
Spectrom. Ion Processes 1987; 78: 53.
2. Dreisewerd K, Kingston R, Geraerts WPM, Li KW. Int. J.
Mass Spectrom. 1997; 169: 291.
3. Binz P-A, Mueller M, Walther D, Bienvenut WV, Gras R,
Hoogland C, Bouchet G, Gasteiger E, Fabbretti R, Gay S,
Palagi P, Wilkins MR, Rouge V, Tonella L, Paesano S, Rossellat G, Karmine A, Bairoch A, Sanchez J-C, Appel RD, Hochstrasser DF. Anal. Chem. 1999; 71: 4981.
4. Caprioli RM, Farmer TB, Gile J. Anal. Chem. 1997; 69: 4751.
5. Fragu P, Briancon C, Fourre C, Clerc J, Casiraghi O, Jeusset J, Omri F, Halpern S. Biol. Cell 1992; 74: 5 8.
6. Fragu P. Clerc J, Briancon C, Fourre C, Jeusset J, Halpern S.
Micron 1994; 25: 361.
7. Valaskovic GA, Morrison GH. Scanning Microsc. 1992; 6:
8. Tolstogouzov AB, Kitaeva TI, Volkov SS. Mikrochim. Acta
1994; 114: 505.
9. Hercules DM, Novak FP, Viswanadham SK, Wilk ZA. Anal.
Chim. Acta 1987; 195: 61.
10. Delic J, Coppey J, Magdelenat H, Coppey-Moisan M. Exp.
Cell Res. 1991; 194: 147.
11. Obendorf SK. Bioinorg. Chem. 1976; 6: 29.
12. Reddy BS, Seshadri TP, Sakore TD, Sobell HM. J. Mol. Biol.
1979; 135: 787.
Copyright  2000 John Wiley & Sons, Ltd.
13. Lillie RD. H. J. Conn’s Biological Stains (9th edn). Williams
and Wilkins: Baltimore, 1977.
14. Clark G. Staining Procedures (4th edn). Williams and Wilkins:
Baltimore, 1981.
15. Larsen TA, Goodsell DS, Cascio D, Grzeskowiak K, Dickerson RE. J. Biomol. Struct. Dyn. 1989; 7: 477.
16. Williamson DH. Methods Cell Biol. 1975; 12: 335.
17. Tanious FA, Veal JM, Buczak H, Ratmeyer LS, Wilson WD.
Biochemistry 1992; 31: 3103.
18. Waring MJ. J. Mol. Biol. 1965; 13: 269.
19. Haugland RP. Handbook of Fluorescent Probes and Research
Chemicals (6th edn). Molecular Probes: Eugene, OR,
20. Sterzel W, Bedford P, Eisenbrand G. Anal. Biochem. 1985;
147: 462.
21. Otto F, Tsou KC. Stain Technol. 1985; 60: 7.
22. Loontiens FG, Regenfuss P, Zechel A, Dumortier L, Clegg
RM. Biochemistry 1990; 29: 9029.
23. Sawicki W, Mystkowska ET. Anat. Rec. 1990; 227: 3592.
24. Green FJ. The Sigma Aldrich Handbook of Stains, Dyes,
and Indicators. Aldrich Chemical: Milwaukee, WI, 1991.
25. Vaney DI, Gynther IC, Young HM. J. Comp. Neurol. 1991;
310: 154.
26. Stoeckli M, Farmer TB, Caprioli RM. J. Am. Soc. Mass Spectrom. 1999; 10: 67.
J. Mass Spectrom. 35, 258–264 (2000)
Без категории
Размер файла
306 Кб
Пожаловаться на содержимое документа