ArticlePDF Available

Three-dimensional molecular reconstruction of rat heart with mass spectrometry imaging


Abstract and Figures

Cardiovascular diseases are the world’s number one cause of death, accounting for 17.1 million deaths a year. New high-resolution molecular and structural imaging strategies are needed to understand underlying pathophysiological mechanism. The aim of our study is (1) to provide a molecular basis of the heart animal model through the local identification of biomolecules by mass spectrometry imaging (MSI) (three-dimensional (3D) molecular reconstruction), (2) to perform a cross-species validation of secondary ion mass spectrometry (SIMS)-based cardiovascular molecular imaging, and (3) to demonstrate potential clinical relevance by the application of this innovative methodology to human heart specimens. We investigated a MSI approach using SIMS on the major areas of a rat and mouse heart: the pericardium, the myocardium, the endocardium, valves, and the great vessels. While several structures of the heart can be observed in individual two-dimensional sections analyzed by metal-assisted SIMS imaging, a full view of these structures in the total heart volume can be achieved only through the construction of the 3D heart model. The images of 3D reconstruction of the rat heart show a highly complementary localization between Na+, K+, and two ions at m/z 145 and 667. Principal component analysis of the MSI data clearly identified different morphology of the heart by their distinct correlated molecular signatures. The results reported here represent the first 3D molecular reconstruction of rat heart by SIMS imaging. Figure Workflow of the 3D reconstruction. A Tissue section, B gold deposition is done by sputter coating, C, C1 SIMS-ToF mass analyzer, C, C2 mass spectral peaks, C, C3 datacube images; D, E Reconstruction of the heart showing 3D-spatial distributions of three different ions 145 m/z (red), 23 m/z (green), and 39 m/z (blue); F coregistration of 40 individual MS imaging
Content may be subject to copyright.
Three-dimensional molecular reconstruction of rat
heart with mass spectrometry imaging
Lara Fornai &Annalisa Angelini &Ivo Klinkert &
Frans Giskes &Andras Kiss &Gert Eijkel &
Erika A. Amstalden-van Hove &Leendert A. Klerk &
Marny Fedrigo &Giuseppe Pieraccini &
Gloriano Moneti &Marialuisa Valente &Gaetano Thiene &
Ron M. A. Heeren
Received: 20 June 2012 /Revised: 17 September 2012 / Accepted: 21 September 2012
#Springer-Verlag Berlin Heidelberg 2012
Abstract Cardiovascular diseases are the worlds number
one cause of death, accounting for 17.1 million deaths a
year. New high-resolution molecular and structural imag-
ing strategies are needed to understand underlying path-
ophysiological mechanism. The aim of our study is (1) to
provide a molecular basis of the heart animal model
through the local identification of biomolecules by mass
spectrometry imaging (MSI) (three-dimensional (3D) mo-
lecular reconstruction), (2) to perform a cross-species
validation of secondary ion mass spectrometry (SIMS)-
based cardiovascular molecular imaging, and (3) to
demonstrate potential clinical relevance by the application
of this innovative methodology to human heart speci-
mens. We investigated a MSI approach using SIMS on
the major areas of a rat and mouse heart: the pericardi-
um, the myocardium, the endocardium, valves, and the
great vessels. While several structures of the heart can be
observed in individual two-dimensional sections analyzed
by metal-assisted SIMS imaging, a full view of these
structures in the total heart volume can be achieved only
through the construction of the 3D heart model. The
images of 3D reconstruction of the rat heart show a
Please address reprint requests to: Prof. Dr. R.M.A. Heeren, FOM-AMOLF,
Science Park 104, 1098 XG Amsterdam, The Netherlands
Electronic supplementary material The online version of this article
(doi:10.1007/s00216-012-6451-3) contains supplementary material,
which is available to authorized users.
L. Fornai (*):A. Angelini :M. Fedrigo :M. Valente :G. Thiene
Cardiovascular Pathology, Department of Cardiac,
Thoracic and Vascular Sciences, University of Padua,
Via Gabelli 61,
35121 Padua, Italy
L. Fornai :I. Klinkert :F. Giskes :A. Kiss :G. Eijkel :
E. A. A.-v. Hove :L. A. Klerk :R. M. A. Heeren
Science Park 104,
1098 XG Amsterdam, The Netherlands
L. Fornai
R. M. A. Heeren
G. Pieraccini :G. Moneti
CISM Mass Spectrometry Centre,
Viale Pieraccini 6 University of Florence,
50139 Florence, Italy
R. M. A. Heeren
The Netherlands Proteomics Centre, Utrecht University,
H.R. Kruytgebouw, Padualaan 8,
3584 CH Utrecht, The Netherlands
Anal Bioanal Chem
DOI 10.1007/s00216-012-6451-3
highly complementary localization between Na
, and
two ions at m/z 145 and 667. Principal component analysis of
the MSI data clearly identified different morphology of the
heart by their distinct correlated molecular signatures. The
results reported here represent the first 3D molecular recon-
struction of rat heart by SIMS imaging.
Keywords 3D molecular reconstruction .Secondary ion
mass spectrometry .Heart .Imaging
Non-standard abbreviations and acronyms
DAG Diacylglycerols
FFA Free fatty acids
Matrix-assisted laser desorption/ionization-
mass spectrometry imaging
Metal-assisted-secondary ion mass
MS Mass spectrometry
MSI Mass spectrometry imaging
PCA Principal component analysis
SIMS Secondary ion mass spectrometry
TAG Triacylglycerols
ToF Time-of-flight
VARIMAX VARIance MAXimization
Heart failure is among the leading causes of morbidity
and mortality and can result from either primary or
secondary heart muscle disease [1]. The causes of car-
diac dysfunction in most heart diseases are still largely
unknown but are likely to result from underlying alter-
ations in gene and protein expressions or downstream
metabolic processes. The functional complexity of an
organism far exceeds that indicated by its genome se-
quence alone, and this is dependent on the products of
gene expression, including transcriptomics, proteomics,
lipidomics, and metabolomics [25]. Proteome databases
containing two-dimensional (2D) gel electrophoresis, 2D
gel images, and protein spot identifications have been
compiled for canine and rat myocardial tissues [6,7].
An animal model is crucial to evaluate new basic mo-
lecular insights prior to their application in human stud-
ies. Rats exhibit physiological characteristics similar to
those of humans and have been a key experimental
model in biomedicine for over a century [8]. However,
to date, there is no comprehensive molecular image
database for the rat heart. The construction of such
databases in animal model is important for the identifi-
cation of the molecular basis of pathological substrates
caused by a cardiovascular disease. This requires a
molecular imaging method that provides detailed insight
into the spatial distribution of a broad range of elements
and molecules. A mass spectrometer is described as the
smallest weighing scale ever used in the world [9].
Mass spectrometry (MS) is an analytical technique that
is used to determine the molecular weight of a variety
of chemical compounds. Because the mass of a chemi-
cal compound is dependent on its elemental composi-
tion, it is an important determinant of its identity. Mass
spectrometry imaging (MSI) is a technique that enables
the localization of molecules directly from biological
surfaces. The advantage of MSI is its ability to detect
the distribution of hundreds of unknown compounds in
a single measurement without the use of chemical or
immunological labels [10,11]: It is a true label-free
molecular imaging technique. The aims of the present
study were: (a) the establishment of a molecular atlas of
the heart through a direct and large-scale, local analysis
of lipid and elemental ions in a healthy rat heart tissue;
(b) to perform a cross-species validation through the
analysis of mouse and human heart tissues and compare
and contrast our molecular findings; (c) to test this
methodology on clinical human samples as the last
element of our translational research.
Ethics statement
The animals used in this study were purchased from Harlan
Laboratories (Boxmeer, The Netherlands). The Center for
Cardiovascular Pathology of the University of Padova spe-
cifically approved the use of human tissues. Written in-
formed consent was acquired from human subject involved
in the study.
Methods and materials
Rat heart from adult rats (type WU) and mouse heart (type
9CFW-1) were frozen and stored at 80 °C until sectioning.
The tissue sections were stained with hematoxylin (Sigma
Diagnostics, Zwijndrecht, The Netherlands) and eosin
(Merck Diagnostica, Darmstadt, Germany) after secondary
ion mass spectrometry (SIMS) analysis to correlate the
observed molecular profiles with morphological features.
Several successive heart sections from three different rats
and one mouse heart were analyzed using standardized and
optimized workflows.
Rat and mouse hearts were sectioned at 12 μmthick-
ness and sliced into sagittal sections at 20 °C on a
cryomicrotome (Microm International, Waldorf, Ger-
many) and then mounted on indiumtinoxide-coated
glass slides (ITO, 48Ωresistance; Delta Technologies,
Stillwater, MN, USA). All samples slices were stored at
L. Fornai et al.
80 °C prior to use and dried in a vacuum desiccator
for 30 min prior to MS analysis. The 12 μm thickness
ensures that enough analyte molecules are available for
ionization and no problems occur with the insulating
properties of tissue [12]. Sample metallization as a
means to improve molecular ion yields was performed
by sputter deposition of a 1-nm layer of gold [13].
Samples were analyzed by high-resolution time-of-
flight secondary ion mass spectrometry (ToF-SIMS) im-
aging. This procedure is often referred to as metal-
assisted-secondary ion mass spectrometry (MetA-SIMS)
when sample metallization is used. ToF-SIMS analysis
of the cross-sectioned heart was done on a Physical
Electronics (Eden Prairie, MN) TRIFT-II (triple focusing
time-of-flight) ToF-SIMS system, equipped with a
22 keV gold liquid metal ion source. The analysis was
performed in positive ion mode. The instrument was
calibrated in positive ion mode on high occurrence
elements and fragments such as H
. Since the sample size is approximately 2 × 2 cm,
the resulting image fidelity of such ToF-SIMS experi-
mentsiscloseto0.6μm per pixel. The raster size (or
fieldofview(FOV))usedwas150μm per tile, with
resolution of 256×256 pixel per tile. The FOV is auto-
matically calculated by WinCadence software
(ULVAC-PHI Inc., Kanagawa, Japan) based on the max-
imum raster size defined by the user and the total
measurement area. The heart tissue was analyzed using
a mosaic mode of 128×128 tiles with a spectral reso-
lution of 22-bit mass channels, for a total of about one
billion pixels. The acquisition time was 3 s for each
tile. Both FOV and acquisition time were constant
throughout the experiment, which covered the entire
sample surface. SIMS is an extremely surface sensitive
technique in which ions are exclusively generated from
a depth of no more than 50 nm from the tissue surface.
We used the static SIMS mode, where the primary ion
dose is so low that each incoming ion hits a unique
spot on the surface, less than 1 % of the surface area is
analyzed [14,15]. In our work, a list is created from
the recorded data containing the position as a two-
dimensional coordinate, the channel number, c(which
is linearly related to the time-of-flight and hence the m/z
value), and the number of counts (n) for that respective
ion. This dataset, which represents a so-called datacube,
can subsequently be converted into an x×yby cunfold-
ed datacube containing the number of counts for each
spectral and spatial combination. These datacubes can
be visualized with the DataCube Explorer (http:// developed at the FOM Institute
AMOLF, a lightweight visualization tool providing a
platform to share and explore MSI datasets. Several
useful features are available to dynamically scroll
through the data, analyze selected regions, and process
and classify the image [16]. Here, we will focus on a
description of the high-resolution metA-SIMS molecular
atlas that was recently completed from us.
PCA analysis and variance maximization (VARIMAX)
Principal component analysis (PCA) was employed to
extract meaningful information out of the complex and
extremely large datasets generated by high-resolution
imaging mass spectrometry (IMS). PCA is a statistical
technique to find patterns in data of high dimensional-
ity. The data are described in such a way that corre-
lated similarities and differences are highlighted. This
is realized by transforming a number of data features
(variables) into a smaller number of orthogonal varia-
bles called principal components (PCs), which de facto
consist of correlated spectral features. The first princi-
pal component accounts for the largest amount of var-
iance in the data. Each successive component describes
a smaller part of the remaining variance. PCA is there-
fore well-suited to be applied on hyperspectral datasets
as used to construct the molecular atlas described in
this paper. The hyperspectral SIMS data have both a
large spatial as well as a large spectral dimension.
Additional optimizations can be done after completion
of the PCA to enhance the spectral contrast in the data.
One method is an additional fitting of the principal
components to maximize the variance expressed in
each component. There is a number of maximization
criteria, but the VARIance MAXimization (VARIMAX)
is the most common. It can be used as a post-
processing step after PCA, as previously described
[17]. By rotation of the orthogonal axis, components
with a higher contrast are created. To highlight the
spectral correlations in the three-dimensional (3D) at-
las, PCA was used with VARIMAX optimization.
Rat heart high-resolution imaging mass spectrometry data
Imaging mass spectrometry data are often displayed as
a total-ion-count image; the output of a mass spectrom-
eter consists of a set of mass-to-charge ratios (m/z)of
detected ions. IMS produces such a set of mass-to-
charge ratios for each pixel in a 2D grid, allowing
analysis of the chemical structure of the sample. The
resulting grid of m/z histograms is commonly visual-
ized in the form of a total-ion-count image. For each
pixel, a total-ion-count image maps the number of
items in the corresponding set of m/z values to an
intensity value. Other methods of visualization are
based on various forms of multi-variate analysis on
Three-dimensional molecular reconstruction of rat heart with MSI
the m/z data, producing maps of chemically similar
regions as show from Smit et al. [18].
Human heart Human failing left-ventricular free-wall
heart explants were obtained from the heart transplant
collection at Padua University. Samples (left ventricle)
from explanted heart (heart failure) were frozen and
stored at 80 °C until sectioning. Heart tissue transverse
sections (12 μm thickness) were cut using a cryomicro-
tome and thaw-mounted onto an ITO slide. The same
SIMS-ToF method and statistical analysis described for
rat and mouse heart were applied to the human sample.
A sample of the left ventricle was further divided into
manageable blocks for paraffin embedding. From each
block, three serial sections of 4 μm thickness were
taken. Adjacent (serial) sections were then stained with
one of the three following techniques to correlate hearts
molecular profile with morphological features: (a) hema-
toxylin and eosin (H&E) staining, (b) Sirius red-
collagen-staining for connective tissue, and (c) Heiden-
hain's Azan stained for connective tissue. Comparison
was made with a serial staining section to show the
anatomical structure and confirm the distribution of
connective tissue in that location.
2D MSI-based molecular imaging of rat heart sections
Surface rastering of heart tissue sections generated a
plethora of secondary ions with a molecular weight up
to m/z 1,500 (Fig. 1AC). Several distinctive MS peaks
and correlated image patterns were observed in the
positive-ion mode for the heart sections analyzed. The
selected ion images are highly sensitive to the specific
anatomical tissue types within the sections (Fig. 1A, B).
Several peaks are often visible within a single m/z
range, and a single peak can be used to create descrip-
tive molecular ion images. The resulting intensity map
is usually visualized using a pseudo-color map. Because
of the lack of an efficient MS/MS method that could be
associated with ToF-SIMS imaging, structure attributions
or assignments of ion peaks were made according to the
instrument resolution and accuracy, the valence rule, and
the biological characteristics of the tissue. All the mass
assignments have been done from data taken from the
literature [1922].
The high-spatial-resolution images shown in Fig. 2dem-
onstrate that the spatial resolution obtained in these SIMS
experiments is more than sufficient to reveal in detail all
major anatomical substructures in the rat heart. Note that
conventional matrix-assisted laser desorption/ionization-
mass spectrometry imaging (MALDI-MSI) with a pixel size
of 100 μm is not capable of revealing the detailed structures
shown in Fig. 3where the distribution of a wide variety of
secondary ions imaged in the heart is observed. The ions at
m/z 667 and m/z 840 localized very precisely within the
aorta wall (Fig. 3c, d). For instance, m/z 83 is localized in
aorta wall, semilunar valve, and endocardium, whereas it is
hardly visible in both ventricles and atria (Fig. 3a). The
image in Fig. 3f shows the ion at 145 m/z as very much
localized in pulmonary artery, right atrium, and atrioventric-
ular valve. Interestingly, m/z 175 and m/z 213 are observed
only in ventricles. The bottom image Fig. 3m, in contrast,
shows m/z 334 localized only in the pericardium. The image
in Fig. 3q shows the high spatial distribution at 969 m/z in
both atria, aorta wall, left atrioventricular valve, and right
coronary artery. Significant peaks in our data are seen at m/z,
representing cholesterol ion [MH
whose distribu-
tion was imaged. Important tissue differences are distin-
guished upon examination of the cholesterol distribution.
Cholesterol shows higher intensity in both atria, aorta wall,
atrioventricular valves, and the coronary artery but is ob-
served with lower intensity in ventricles (Fig. 3i). VARI-
MAX rotation was used to enhance the spectral contrast of
the PCs. This axis rotation results in a higher molecular
contrast not only in the spectra, but also in a higher molec-
ular image contrast. Several signals of fatty acids show a
variation in their spatial distribution that corresponds direct-
ly to the degree of lipid unsaturation, and hence energy
catabolism. Diacylglycerol (DAG) species that were identi-
fied as [M+HOH]
)atm/z 549, [M+HOH]
)atm/z 577, [M+HOH]
)atm/z 603
and ceramide at m/z 604 could be detected. Substantial
differences are also seen in the amount of free choline
present in the various tissue parts. The distributions based
on the PCA results revealed a clear image of the different
areas where m/z 104 (choline) displayed a high intensity
signal. The m/z 104 signal strongly localizes in atria,
aorta, pulmonary artery, and the atrioventricular and
semilunar valves but has lower intensity in ventricles,
asshowninFig.4B. The phosphocholine headgroup
at m/z 184 (C
) was used to localize the
phosphocholine-containing phospholipids, i.e., sphin-
gomyelins and phosphatidylcholines [23]. The phos-
phocholine headgroup was additionally localized by
imaging a specific fragment (m/z 86). This signal
was localized in the left and right ventricles with
low intensity. In right and left atria, aortic wall, and
aorta valve, the intensity of m/z 86 signal was higher
(Fig. 3a). The localization of chemical components in
the tissue reveals structural information that can be
used for creating an atlas (Fig. 3). Separated ion images
of relevant molecules provide this image information. PCA is
used to find spatially correlated molecules; the resulting PCA
L. Fornai et al.
score images greatly enhance image contrast in comparison
with the separate ion images (Fig. 4).
3D molecular reconstruction of the rat heart
The generation of a three-dimensional dataset requires
an additional z-dimension. In our experiments, this was
achieved by successive tissue sectioning with well-
defined and measured spatial intervals [24]. Using the
micrometer scale of a cryomicrotome, an entire cryo-
preserved rat heart was sectioned. Forty sections at
irregularly spaced but well-documented intervals were
taken through the heart as shown in Figure 1A(Elec-
tronic supplementary material). The workflow of the 3D
reconstruction is show in Fig. 5.Thez-position was
recorded for each of the sections. High-resolution
metA-SIMS datasets were acquired from each section
with a 22 keV Au
primary ion beam. Each dataset
was acquired in 12 h. Forty datasets were acquired
resulting in a total of 42,949,672,960 spectra in the
raw data files. These data were subsequently combined
and processed to reveal the three-dimensional molecular
features. The processing protocol included spectral and
spatial binning to reduce the total dataset size prior to
molecular feature visualization using our datacube ex-
plorer. The 3D data volumes can be explored using the
Volume explorer software that is used to reconstruct a
3D-data grid out of the 40 individual 2D-datasets. The
molecular images of specific m/z range of interest areas
are put together into a three-dimensional volume in
which the pixels are turned into voxels. The Volume
explorer then uses volume-extraction by combining vox-
els above a certain threshold into a volume. The co-
registration of individual molecular images was obtained
by manual alignment in the Volume explorer of MSI-
section on the base of anatomical structures. This is, to
Fig. 1 A,B1,CMetal-assisted SIMS images of a sagittal section of
three different rat hearts; A1 anatomic differences from rat heart visu-
alized with H&E staining after SIMS analysis. B,B1 SIMS-MSI
images of ions detected from representative rat heart valves and
H&E-stained tissue sections. DMetal-assisted SIMS images of a
sagittal section of mouse hearts (datacube images), D1 H&E-stained
tissue sections and D2 spectrum of SIMS
Three-dimensional molecular reconstruction of rat heart with MSI
the knowledge of the authors, the first time three-
dimensional SIMS-based molecular volumes were con-
structed with a 1 μm lateral resolution and a 100 μm
depth resolution. The consistent observation of identical
correlated anatomical and molecular structures within
each of the technical replicates is shown in Figure 1B
of the Electronic supplementary material.
The images of the 3D reconstruction additionally
show a highly complementary localization between
,andanionatm/z 145. Na
is localized in
the atria, while K
is strongly localized in the ventricles
as previously reported [25](Fig.5D). Three-dimensional
reconstructions of three selected individual mass spectral
peaks can be simultaneously displayed using a Red
GreenBlue color scheme. Each color represents a spe-
cific molecular feature (Fig. 5E). The overlaid 3D mo-
lecular model obtained is a representation of the whole
heart. The result of the combination and co-registration
of 40 individual MS imaging data cubes is demonstrated
in Fig. 5F. The data volumes can be explored using
software which allows a close molecular look inside
the heart. The molecular visualization of the different
valvular structures, the pericardium, the atria, the coro-
nary arteries, and various other anatomical features pro-
vides a new tool in molecular pathology. While several
structures of the heart can be observed in individual 2D
sections analyzed by Meta-SIMS imaging, a full view of
these structures in the total heart volume can be
achieved only through the construction of the 3D heart
model. The data show significant differences in the ions
distribution in the various heart structures and reveal
distinctive molecular localization in this organ.
Mouse heart
A cross-species validation was performed through the
analysis of mouse heart sections (Fig. 1D,D1andD2).
We observed similar molecular patterns when the
mouse results were compared with the rat results. This
demonstrates, among others, the diversity in biological
systems in which this technology can be applied.
Cholesterol-related ions again demonstrated to have
specific higher intensity in atria, aorta, and pulmonary
artery and lower intensity in ventricles (Fig. 1D). The
ion at m/z 369 localized very precisely within the
pulmonary artery, tricuspid with high intensity in both
atria, aorta, pulmonary artery, and in the atrioventricu-
lar and semilunar valves but is detected in ventricles
with lower intensity. The similarity of these molecular
findings with the results obtained in the rat heart fur-
ther corroborates the across-species molecular consis-
tency at the lipid level.
Human heart
The true relevance of a new molecular imaging technol-
ogy as described here is demonstrated through the ap-
plication of high-resolution imaging MS on explanted
human heart samples taken from patients suffering from
ventricular failure. The application of MSI in the field
Fig. 2 High-resolution
(8,192× 8,192 pixels) SIMS
total ion images in black and
white and complementary
H&E-stained images showing
the different morphological
structures of the heart observed
in both molecular imaging
L. Fornai et al.
of human cardiovascular pathology has a direct potential
for clinical diagnoses and treatments. The larger size of
the human heart combined with the time-consuming
experimental procedure is prohibitively limitative for
the generation of a full three-dimensional molecular
model of the human heart. Instead, we have opted for
the analysis of smaller subsets of the human heart, in
this case, segments of the left ventricle.
In all the experiments, the several distinctive structures in
the ventricle can be clearly seen, originating from different
lipid composition and signal intensities. An abundant signal
of cholesterol (m/z 369) [MH
can be observed
(Fig. 6). Cholesterol ion exhibit a high intensity in the
myocardium. The observed intensities in the pericardium
and endocardium are substantially lower. The m/z 104 cho-
line signal co-localizes with high intensity in myocardium
and endocardium (Fig. 6). The oleic (47 %) and palmitic
(19 %) acids are known to be the dominant fatty acids in the
human heart. Results from human and animal models of
heart failure generally support the concept of decreased fatty
acid β-oxidation in heart failure [26]. These findings are
consistent with our molecular images presented below that
show DAG species tentatively identified as [M+HOH]
)atm/z 549, [M+HOH]
577, [M+HOH]
, and ceramide at m/z 604. As expected,
they could be detected with high intensity in the pericardium
and low intensity in the myocardium. The SIMS imaging
data demonstrated that several important catabolic mole-
cules and their spatial features can be readily identified
and localized in a single MS imaging experiment (Fig. 6).
This in turn leads to a better fundamental understanding of
the pathological pathways whose molecular constituents are
Adjacent human left ventricle sections were stained
with H&E, Sirius red, and trichrome and compared
with the SIMS imaging datasets obtained (Fig. 6A
D). Figure 6BDillustrates a transverse-section of left
ventricular muscle that contains a vein structure. The
different tissue types in this section can only be dis-
tinguished by a combination of three different staining
Fig. 3 Metal-assisted SIMS images of a sagittal section of a rat heart.
Top image shows the spatial distribution of ions (a,b,c,d) in the aorta
wall. For instance, the images in eand ishow distribution for the main
cholesterol ions (m/z 369 and m/z 385) that are localized to the aorta
wall, aorta valve, right coronary artery, and right and left atria but are
not observed in both ventricles. The image in fshows the ion at 145 m/
zvery localized in pulmonary artery, right atrium, and atrioventricular
valve. Interestingly, m/z 175 and m/z 213 are observed only in ven-
tricles. The bottom image m, in contrast, shows m/z 334 localized only
in the pericardium The image in qshows the high spatial distribution of
at 969 m/z in both atria, aorta wall, left atrioventricular valve, and right
coronary artery. All ion image scale bars0100 μm
Three-dimensional molecular reconstruction of rat heart with MSI
procedures, revealing the morphology and the presence
of connective tissue. The molecular images obtained
with SIMS provide all of this information (and more)
in a single experiment. Statistical analysis of human
heart (three technical replicates) yields correlation coef-
ficients of overall spectra of A versus B 0.9934, A
versus C 0.9858, B versus C 0.9925, respectively. This
again confirms the reproducibility and selectivity of
this innovative method for molecular histology (see
Figure 2, Electronic supplementary material). The tis-
sue sections were stained with H&E after SIMS anal-
ysis to correlate the hearts molecular profile with the
observed morphological features.
SIMS imaging was able to consistently detect sodium
(23 m/z), potassium (39 m/z), choline (104 m/z),
phosphocholine (184 m/z), cholesterol (369 and 385 m/z),
DAG species (549 m/z,577m/z,603m/z), ceramide (604 m/
z), the ions at 145 m/z,175m/z,334m/z, 213 m/z,969m/z,
and several other molecules in rat, mouse, and human heart
structures. The assignments of molecules were made based
on the unique masses of single elements (e.g., Na
22.989, K
m/z 39.098), the calculated molecular weight
of more complex substances, and by comparison with
chemical standards. The lipid database was employed to
correlate specific molecules found by SIMS imaging within
the heart structures. The ability to identify specific biomole-
cules is crucial for biological, and especially pathological,
investigations. Molecular biology thrives on molecular imag-
ing techniques that aim at the investigation of the relationship
between spatial organization, structure, and function of mole-
cules in biological systems. Although impairment in calcium
homeostasis, abnormal myocyte energetics, and myocardial
remodeling have been described to be associated with cardiac
dysfunction, the underlying molecular mechanisms involved
Fig. 4 A PCA spectral results after VARIMAX optimization show a
strong contribution for among others cholesterol ([MH
at m/z
369.1 and [MH]
at m/z 385) and choline (m/z 104) showing corre-
lation with aorta wall, left and right atria, semilunar valve, atrioven-
tricular valve, left ventricle, right ventricle, and coronary artery. The
distributions based on the PCA results (B) resulted in a molecular
underpinning of the different areas that morphologically identified with
H&E staining (C). The intensity of each ion is indicated in the color
chart on the left from white (high) to dark (absent)
L. Fornai et al.
in the transition from normal cardiac function to heart failure
remain poorly understood [27]. Insight into these processes
and pathways will be important in the development of new
therapeutic strategies for treatment and prevention of heart
failure. Under normal physiologic conditions, the heart uti-
lizes fatty acids as its chief energy substrate. Because there is
limited capacity for triglyceride storage in the cardiomyocyte,
the uptake and oxidation of fatty acids is tightly coupled. The
accumulation of triglycerides in the heart, caused by a mis-
match between the uptake and the oxidation of fatty acids, is
associated with a number of pathophysiological conditions. In
animal models (rats) of obesity and diabetes, triglyceride
accumulation within cardiomyocytes is associated with im-
paired contractile function. Although it is unclear how lipids
induce cardiac dysfunction, accumulation of intramyocardial
triglycerides is associated with altered gene expression [28].
The lipid droplet is endured by a core of lipids, which mostly
consists of TAG (9099 %) and to a lesser amount of DAG,
FFA, phospholipids, and monoacylglycerols. Several distinc-
tive MS peaks and correlated image patterns consistent with
the expected oleic/palmitic acid ratio were observed in
consecutive SIMS experiments used for the 3D reconstruction
of rat heart.
Each molecular pattern in these sections was ana-
lyzed in a search for molecular classifiers. For exam-
ple, cholesterol was found to distinguish directly atria
from ventricles. The difference in the relative choles-
terol content in atria and ventricles was observed in a
direct comparison between rat and mouse results
(Fig. 1A, C, D). These results corroborate and expand
earlier studies performed on heart substructure homo-
genates that were analyzed with liquid chromatography
[29]. Here, for the first time, we visualize the heart
cholesterol distribution directly on histological tissue
sections without the use of any labeling approach to-
gether with a plethora of other molecules. We are not
able to assign a molecular structure to the ions at
145 m/z, 175 m/z, 334 m/z,213m/z, and 969 m/z
(Fig. 3f, h, m, n, q) showing a different distribution
in the tissue sections. In summary, SIMS is shown to
provide extensive local molecular information that
complements the conventional histological approaches
Fig. 5 Workflow of the 3D reconstruction. ATissue section, Bgold
deposition is done by sputter coating, C,C1 SIMS-ToF mass analyzer,
C,C2 mass spectral peaks, C,C3 datacube images; D,E
Reconstruction of the heart showing 3D-spatial distributions of three
different ions 145 m/z (red), 23 m/z (green), and 39 m/z (blue); Fco-
registration of 40 individual MS imaging
Three-dimensional molecular reconstruction of rat heart with MSI
for the determination of the chemical properties of
specific, known anatomical structures inside the heart.
As any analytical technique, SIMS has limitations and
advantages for molecular histology. In the next section,
we briefly discuss the most important aspects of SIMS
in cardiovascular research.
Limitations of SIMS The extreme surface sensitivity
requires careful sample treatment. Sample contamina-
tion and molecular diffusion, which can affect the
reproducibility of the data, complicate their analysis,
or affect the quality of the image, are major consider-
ations in the sample preparation protocols. The multi-
ple molecules present in a tissue section can negatively
influence each othersdesorption and ionization effi-
ciency and prevent optimal detection. This phenomenon
is called ion suppression and can limit the number of
detected molecules. The SIMS imaging technique is
only capable of a semi-quantitative measure of the
distribution of elements and molecules in tissue.
Absolute quantitation requires the use of isotopically
labeled standards. Identification of species is based on
single m/z values exclusively; as in most SIMS instru-
ments, no tandem mass spectrometry can be performed.
New instruments will overcome this limitation but are
not yet commercially available.
Advantages of SIMS
SIMS allows for label free molecular imaging of mul-
tiple species in parallel directly on tissue. SIMS instru-
ments are used for imaging unknown compounds
present in the biological sample without any a priori
knowledge or labels in a single experiment. It is a true-
discovery methodology. It has been used for imaging
of elemental species in cells at high spatial resolution
(50 nm) and can typically analyze ions up to
1,500 m/z. In the last decade, there have been a grow-
ing number of studies that show the capacity of SIMS
for analyzing biological materials, including different
Fig. 6 A Metal-assisted SIMS images of human heart. All the images
obtained are from the same section. All ion images scale bar0100 μm. B
H&E staining. CSection stained with Sirius red. Connective tissue stains
red. In this trichrome-stained specimen (image D), collagen is colored
blue and smooth muscle is red. The images in B,C,andDillustrating a
transverse-section of left ventricle muscle, clearly include an intra-
myocardial blood vessel to illustrate the staining pattern in gross sheets
of cardiomyocytes that can be seen towards the periphery of the figures
L. Fornai et al.
tissues such as mouse brain, human adipose tissue,
muscle, kidney, and single cultured cells [3035]. More
information on different MS imaging technologies for
biomedical purposes can be found in current review of
the various imaging MS technologies for molecular
pathology [11].
A major advantage is the reduced amount of sample used;
often, after SIMS analysis, the tissue surface can be ana-
lyzed or stained as if the tissue was pristine. This, in turn,
allows the direct comparison with other ex vivo techniques
(e.g., fluorescence and immunostaining) for orthogonal val-
idation of the findings. These techniques are highly specific
within the class but allow the visualization of labeled mol-
ecules only with a limited number of detectable analytes per
section [36].
Future and perspective
The application of this method in future studies can be used to
identify changing tissue regions that are indicative of human
cardiac disease. SIMS improves tissue classification necessary
to perform retrospective studies, will assist clinical studies
from the bench to bedside, and can guide therapeutic choices.
The direct application of SIMS on the same heart tissues as
used by pathologists improves and accelerates molecular di-
agnoses. Molecular tissue classification after SIMS based on
known biomarkers or using unsupervised multivariate analy-
ses can positively affect patient treatment. The evolution of a
heart disease during treatment can be monitored based on the
tissue biomarkers identified by MSI. In cases where traditional
biomarkers cannot be clearly detected in biopsies, SIMS could
become critical to the outcome. At this point, it is obvious that
further clinical studies using SIMS technology are required to
fully validate this method. Nonetheless, SIMS as well as
MALDI-MSI have opened the door to molecular tissue clas-
sification, not only for diagnostic and prognostic purposes but
also for treatment development. MS-based molecular imaging
is becoming one of the basic information providers for per-
sonalized medicine, especially when used in complement with
magnetic resonance imaging (MRI). A major advantage for
SIMS will be its coupling with positron emission tomography,
X-ray, computed tomography instrumentation, and MRI for
both preclinical and clinical research. The complementarities
between non-invasive techniques and molecular data obtained
from SIMS-MSI will result in a more precise diagnosis of the
molecular stateof a living system.
In clinical studies, the need for information on the spatial
localization of pathologically gene-encoded products has
become more pressing. The three-dimensional volume
reconstructions generated by SIMS-MSI data now offer
the possibility of comparing the molecular data with
data obtained using positron emission tomography or
MRI [37]. These multi-modal molecular imaging
approaches will strengthen the fundaments of molecu-
lar imaging research.
The SIMS imaging approach can be used to detect and
probe the molecular content of tissues in an anatomical
context. Anatomical atlases based on optical images are
widely used for anatomical and physiological reference.
A series of secondary ion images obtained from succes-
sive tissue sections of rat heart can be used to produce
a 3D molecular reconstruction that contains both pieces
of information. SIMS provides detailed high-resolution
molecular images of tissue surfaces. The results reported
here represent the first 3D molecular reconstruction of
rat heart by SIMS imaging. The measurements were
extended on mouse and human heart samples. Human
tissue analysis is demonstrated to benefit from the po-
tential of SIMS imaging for the investigation of the
distribution of elements and biomolecules directly on
the surface of cardiovascular tissues.
Acknowledgments We gratefully acknowledge the assistance and In-
formatics (CWI) in the generation of the full-resolution SIMS images.
Sources of funding This work is supported by the Cardiovascular
Pathology, Department of Cardiac, Thoracic, and Vascular Sciences,
University of Padua Medical School, and by a grant from the Italian
Society of Cardiology. This work is also part of the research program
of the Stichting voor Fundamenteel Onderzoek der Materie (FOM) and
is financially supported by the Nederlandse organisatie voor Weten-
schappelijk Onderzoek (NWO).
1. McGregor E, Dunn MJ (2006) Proteomics of the heart unraveling
disease. Circ Res 98:309321
2. Mueller DR, Voshol H, Waldt A, Van Oostrum J (2007) LC-
MALDI MS and MS/MSan efficient tool in proteome analysis.
Subcell Biochem 43:355380
3. Li L, Garden RW, Sweedler JV (2000) Single-cell MALDI: a new
tool for direct peptide profiling. Trends Biotechnol 18:151160
4. Suzuki T, Nagai R (2007) Cardiovascular proteomic analysis. J
Chromatogr B Analyt Technol Biomed Life Sci 855:2834
5. Ping P (2009) Getting to the heart of proteomics. N Engl J Med
6. Dunn MJ, Corbett JM, Wheeler CH (1997) HSC-2DPAGE and the
two dimensional gel electrophoresis database of dog heart proteins.
Electrophoresis 18:27952802
7. Li XP, Pleissner KP, Scheler C, Regitz-Zagrosek V, Salnikow J,
Jungblut PR (1999) A two-dimensional electrophoresis database of
rat heart proteins. Electrophoresis 20:891897
8. Russell JC (2003) Of mice and men, rats and atherosclerosis.
Cardiovasc Res 59:810811
9. Chaurand P, Norris JL, Cornett DS, Mobley JA, Caprioli RM (2006)
New developments in profiling and imaging of proteins from tissues
by MALDI mass spectrometry. J Proteome Res 5:28892900
Three-dimensional molecular reconstruction of rat heart with MSI
10. Heeren RMA, McDonnell LA, Amstalden E, Luxembourg SL,
Altelaar AFM, Piersma SR (2006) Why dont biologists use
SIMS? A critical evaluation of imaging MS. Appl Surf Sci
11. Chughtai K, Heeren RM (2010) Mass spectrometric imaging for
biomedical tissue analysis. Chem Rev 110:32373277
12. McQuaw CM, Sostarecz AG, Zheng L, Ewing AG, Winograd
N (2006) Investigating lipid interactions and process of raft
formation in cellular membranes using ToF-SIMS. Appl Surf
Sci 252:67166718
13. Heeren RMA, Kukrer-Kaletas B, Taban IM, MacAleese L,
McDonnell LA (2008) Quality of surface: the influence of
sample preparation on MS-based biomolecular tissue imaging
with MALDI-MS and (ME-)SIMS. Appl Surf Sci 255:1289
14. Amstalden van Hove ER, Blackwell TR, Klinkert I, Eijkel GB,
Heeren RM, Glunde K (2010) Multimodal mass spectrometric
imaging of small molecules reveals distinct spatio-molecular sig-
natures in differentially metastatic breast tumor models. Cancer
Res 70:90129021
15. Pól J, Strohalm M, Havlíček V, Volný M (2010) Molecular mass
spectrometry imaging in biomedical and life science research.
Histochem Cell Biol 134:423443
16. MacAleese L, Duursma MC, Klerk LA, Fisher G, Heeren RMA
(2011) Protein identification with liquid chromatography and ma-
trix enhanced secondary ion mass spectrometry (LC-ME-SIMS). J
Proteomics 74:9931001
17. Klerk LA, Broersen A, Fletcher IW, van Liere R, Heeren RMA
(2007) Extended data analysis strategies for high resolution imag-
ing MS: new methods to deal with extremely large image hyper-
spectral datasets. Int J Mass Spectrom 260:222236
18. Smit FA, van Liere R, Fornai L, Heeren RMA (2010) An interac-
tive, multi-modal approach to analysing high-resolution image
mass spectrometry data. Proceedings of Vision, Modeling and
Visualization (VMV) 283289
19. Malmberg P, Jennische E, Nilsson D, Nygren H (2011) High-
resolution, imaging TOF-SIMS: novel applications in medical
research. Anal Bioanal Chem 399:27112718
20. Passarelli M, Winograd N (2011) Lipid imaging with time-of-flight
secondary ion mass spectrometry (ToF-SIMS). Biochim Biophys
Acta 1811:976990
21. Malmberg P, Börner K, Chen Y, Friberg P, Hagenhoff B, Månsson
J, Nygren H (2007) Localization of lipids in the aortic wall with
imaging TOF-SIMS. Biochim Biophys Acta 1771:185195
22. Tahallah N, Brunelle A, De La Porte S, Laprévote O (2008)
Lipid mapping in human dystrophic muscle by cluster-time-of-
flight secondary ion mass spectrometry imaging. J Lipid Res
23. Lands WE, Hart P (1965) Metabolism of glycerolipids VI specif-
icities of acyl coenzyme A: phospholipids acyltransfererases. J
Biol Chem 240:19051911
24. Crecelius AC, Cornett DS, Caprioli RM, Williams B, Dawant BM,
Bodenheimer B (2005) Three-dimensional visualisation of protein
expression in mouse brain structures using imaging mass spec-
trometry. J Am Soc Mass Spec 16:10931099
25. Sal'manovich VS (1962) The distribution of potassium and sodium
in the heart of various animals. Biull Eksp Biol Med 54:5862
26. Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, Stanley WC
(2010) Myocardial fatty acid metabolism in health and disease.
Physiol Rev 90:207258
27. Braunwald E (2008) Biomarkers in heart failure. N Engl J Med
28. Sharma S, Adrogue J, Golfman L (2004) Intramyocardial lipid
accumulation in the failing human heart resembles the lipotoxic
rat heart. FASEB J 18:16921700
29. Charnock JS, Abeywardena MY, Mcmurchie EJ (1984) The com-
position of cardiac phospholipids in rats fed different lipid supple-
ments. Lipids 19:206213
30. Gurcan MN, Boucheron L, Can A, Madabhushi A, Rajpoot N,
Yener B (2009) Histopathological image analysis: a review. IEEE
Rev Biomed Eng 2:147171
31. Touboul D, Brunelle A, Halgand F, De La Porte S, Laprevote O
(2005) Lipid imaging by gold cluster time-of-flight secondary ion
mass spectrometry: application to Duchenne muscular dystrophy. J
Lipid Res 46:13881395
32. Nygren H, Johansson BR, Malmberg P (2004) Bioimaging TOF-
SIMS of tissues by gold ion bombardment of a silver-coated thin
section. Microsc Res Tech 65:282286
33. Richter K, Nygren H, Malmberg P, Hagenhoff B (2007) Local-
ization of fatty acids with selective chain length by imaging
time-of-flight secondary ion mass spectrometry. Microsc Res
Tech 70:640647
34. Borner K, Malmberg P, Mansson JE, Nygren H (2007) Molecular
imaging of lipids in cells and tissues. Int J Mass Spectrom
35. Ayciriex S, Touboul D, Brunelle A, Laprevote O (2011) Time-of-
flight secondary ion mass spectrometer: a novel tool for lipid
imaging. Clin Lipidol 6:437445
36. Kotze HL, Armitage EG, Fletcher JS, Henderson A, Williams KJ,
Lockyer NP, Vickerman JC (2012) ToF-SIMS as a tool for meta-
bolic profiling small biomolecules in cancer systems. Surf Inter-
face Anal. doi:10.1002/sia.5055
37. ACCF/ACR/AHA/NASCI/SCMR (2010) Expert consensus docu-
ment on cardiovascular magnetic resonance. A report of the Amer-
ican College of Cardiology Foundation Task Force on Expert
Consensus Documents. J Am Coll Cardiol 55:26142662
L. Fornai et al.

Supplementary resource (1)

... These methods generate layers of matrix crystals of different sizes and thickness. In the case of metal-assisted SIMS, a thin layer of gold can be deposited on the sample to improve molecular ion yield [31]. The attainable spatial resolution in matrix-based MSI strategies is determined by the matrix Fig. 1 Schematic overview of the workflow for MSI. ...
... Other optional mass analysers are a quadrupole TOF [35], a LTQ XL linear ion trap [36], and an orbitrap Fourier transform MS (FTMS) [37]. For SIMS MSI, a TOF-based mass spectrometer is most often used [19,25,31,[38][39][40]]. An ion source, for instance, gold or bismuth liquid metal ion gun, generates a pulse of primary ions that are accelerated towards the surface. ...
... Cardiovascular lipid imaging was mostly performed with a (Q)TOF or TOF-TOF instrument, coupled with MALDI [23,27,34,46,47] or SIMS [19,25,31,[38][39][40]. Structural information, high mass accuracy, and identification were done using an FR-ICR, LTQ XL linear ion trap, or orbitrap coupled with MALDI [17,34,36,37,47] or DESI [20,32]. ...
Mass spectrometry imaging (MSI) is a widely established technology; however, in the cardiovascular research field, its use is still emerging. The technique has the advantage of analyzing multiple molecules without prior knowledge while maintaining the relation with tissue morphology. Particularly, MALDI-based approaches have been applied to obtain in-depth knowledge of cardiac (dys)function. Here, we discuss the different aspects of the MSI protocols, from sample handling to instrumentation used in cardiovascular research, and critically evaluate these methods. The trend towards structural lipid analysis, identification, and Btop-down^ protein MSI shows the potential for implementation in (pre)clinical research and complementing the diagnostic tests. Moreover, new insights into disease progression are expected and thereby contribute to the understanding of underlying mechanisms related to cardiovascular diseases.
... In addition, intimate knowledge of histomorphological variation of samples for the selection of ROI in an individual section is necessary. For the known histomorphology in an animal model, the features of organs can be summarized, focusing on the brain including the hippocampus, cortex, thalamus, hypothalamus, and amygdala (Paxinos & Frankin, 2012), heart including left atrium, right atrium, left ventricle, right ventricle, and interventricular septum (Fornai et al., 2012;Savolainen et al., 2009), spleen such as white and red pulp (Mebius & Kraal, 2005), kidney such as renal cortex, medulla and pelvis (Nakayama et al., 2010;Slaughter et al., 2013), placenta (Miner et al., 2016;Yamashita et al., 2011), and whole-body tissue (Satyanarayana et al., 2008;Wong et al. 2015). For the unknown human tumor tissues, the selection of carcinomas and tumor-adjacent tissue at different stages is confirmed under the guidance of a surgeon. ...
... A greater stenosis is associated with plaque rupture and stroke (Patterson et al., 2016) (Figure 4A). For 3D-MSI investigation of heart, Fornai et al. (2012) reported the first 3D heart model by using SIMS imaging and localized the subregions of the heart, including the myocardium, valves, pericardium, and endocardium. ...
Full-text available
Mass spectrometry imaging (MSI) has been applied for label‐free three‐dimensional (3D) imaging from position array across the whole organism, which provides high‐dimensional quantitative data of inorganic or organic compounds that may play an important role in the regulation of cellular signaling, including metals, metabolites, lipids, drugs, peptides, and proteins. While MSI is suitable for investigation of the spatial distribution of molecules, it has a limitation with visualization and quantification of multiple molecules. 3D‐MSI, however, can be applied toward exploring metabolic pathway as well as the interactions of lipid–protein, protein–protein, and metal–protein in complex systems from subcellular to the whole organism through an untargeted methodology. In this review, we highlight the methods and applications of MS‐based 3D imaging to address the complexity of molecular interaction from nano‐ to micrometer lateral resolution, with particular focus on: (a) common and hybrid 3D‐MSI techniques; (b) quantitative MSI methodology, including the methods using a stable isotope labeling internal standard (SILIS) and SILIS‐free approaches with tissue extinction coefficient or virtual calibration; (c) reconstruction of the 3D organ; (d) application of 3D‐MSI for biomarker screening and environmental toxicological research. 3D‐MSI quantitative analysis provides accurate spatial information and quantitative variation of biomolecules, which may be valuable for the exploration of the molecular mechanism of the disease progresses and toxicological assessment of environmental pollutants in the whole organism. Additionally, we also discuss the challenges and perspectives on the future of 3D quantitative MSI.
... These approaches are only suitable for tissues with well-defined and visible structures in the optical or MS images. For the task of aligning and reconstructing 3D visualizations, different software applications have been used [37,[72][73][74], including ImageJ ( [75], R software (, ...
Full-text available
Mass spectrometry imaging (MSI) enables the visualization of molecular distributions on complex surfaces. It has been extensively used in the field of biomedical research to investigate healthy and diseased tissues. Most of the MSI studies are conducted in a 2D fashion where only a single slice of the full sample volume is investigated. However, biological processes occur within a tissue volume and would ideally be investigated as a whole to gain a more comprehensive understanding of the spatial and molecular complexity of biological samples such as tissues and cells. Mass spectrometry imaging has therefore been expanded to the 3D realm whereby molecular distributions within a 3D sample can be visualized. The benefit of investigating volumetric data has led to a quick rise in the application of single-sample 3D-MSI investigations. Several experimental and data analysis aspects need to be considered to perform successful 3D-MSI studies. In this review, we discuss these aspects as well as ongoing developments that enable 3D-MSI to be routinely applied to multi-sample studies.
... The method is applied by various mass spectrometry imaging and microscopy techniques. Fornai et al. pioneered SIMS ion images merging for 3D tissue investigation [22]. In this study, 40 individual tissue slices sectioned at 12 μm thickness were used for 3D reconstruction of a rat heart. ...
Currently two techniques exist for 3D reconstruction of biological samples by time-of-flight secondary ion mass spectrometry (ToF-SIMS). The first, based on microtomy and combining of successive section images, is successfully applied for tissues, while the second, based on sputter depth profiling, is widely used for cells. In the present work, we report the first successful adaptation of sectioning technique for ToF-SIMS 3D imaging of a single cell—fully grown mouse germinal vesicle (GV) oocyte. In addition, microtomy was combined with sputter depth profiling of individual flat sections for three-dimensional reconstruction of intracellular organelles. GV oocyte sectioning allowed us to obtain molecule-specific 3D maps free from artifacts associated with surface topography and uneven etching depth. Sputter depth profiling of individual flat slices revealed fine structure of specific organelles inside the oocyte. Different oocyte organelles (cytoplasm, germinal vesicle, membranes, cumulus cells) were presented on the ion images. Atypical nucleoli referred to as “nucleolus-like body” (NLB) was detected inside the germinal vesicle in PO3⁻ and CN⁻ ions generated by nucleic acids and proteins respectively. Significant difference in PO3⁻ intensity in the NLB central area and NLB border was found. This difference appears as a bright halo around the center area. The NLB size calculated for PO3⁻ and CN⁻ ion images is 12.9 ± 0.2 μm and 11.9 ± 0.2 μm respectively, which suggests that bright halo of PO3⁻ ions is a chromatin compaction on the NLB surface. Areas of approximately 1.0–2.5 μm size inside nucleoplasm with increased PO3⁻ and CN⁻ signal were registered in germinal vesicle. Observed compartments have different sizes and shapes, and they are likely attributed to chromocenters or chromosomes.
... Due to its negligible calculation time, the authors recommended the use of Varimax as a default postprocessing step after PCA for improved interpretability. Fornai et al. (2012) used PCA + Varimax in the investigation of a 3D SIMS dataset of rat heart, consisting of over 49 billion spectra collected from 40 tissue sections. Keenan (2009) has applied Varimax rotation in SIMS data on the spatial-domain components rather than the spectral components, which resulted in a higher contrast for the expression images, as shown in Figure 3. Furthermore, due to the fact that this was a relatively uncomplicated sample with only a few components present in each spatial location, the spectral components are relatively simple, making them similar to those obtained through MCR-ALS, a NMF method discussed in Section II.E. ...
Full-text available
Imaging mass spectrometry (IMS) is a rapidly advancing molecular imaging modality that can map the spatial distribution of molecules with high chemical specificity. IMS does not require prior tagging of molecular targets and is able to measure a large number of ions concurrently in a single experiment. While this makes it particularly suited for exploratory analysis, the large amount and high‐dimensional nature of data generated by IMS techniques make automated computational analysis indispensable. Research into computational methods for IMS data has touched upon different aspects, including spectral preprocessing, data formats, dimensionality reduction, spatial registration, sample classification, differential analysis between IMS experiments, and data‐driven fusion methods to extract patterns corroborated by both IMS and other imaging modalities. In this work, we review unsupervised machine learning methods for exploratory analysis of IMS data, with particular focus on (a) factorization, (b) clustering, and (c) manifold learning. To provide a view across the various IMS modalities, we have attempted to include examples from a range of approaches including matrix assisted laser desorption/ionization, desorption electrospray ionization, and secondary ion mass spectrometry‐based IMS. This review aims to be an entry point for both (i) analytical chemists and mass spectrometry experts who want to explore computational techniques; and (ii) computer scientists and data mining specialists who want to enter the IMS field. © 2019 The Authors. Mass Spectrometry Reviews published by Wiley Periodicals, Inc. Mass SpecRev 00:1–47, 2019.
... 114 a Serial sectioning Serial sectioning is currently the most commonly used technique to perform 3D visualization of SIMS-imaging results. 113,115,116 One of the rst 3D reconstructions from SIMS imaging data was published in 2012 by Fornai et al. 117 where a 3D molecular map of a rat heart was obtained from 40 serial cuts of the heart. In this example, principal component analysis (PCA) was used in order to put in evidence useful information out of very complex and large datasets. ...
This concise tutorial review provides a description of the current state of the art in the application of time-of flight based secondary ion mass spectrometry (TOF-SIMS) in the field of molecular and cellular imaging. The application of TOF-SIMS requires a choice of the appropriate beam combined with a signal enhancement method depending on the surface under investigation. The types of detected molecules and methods for molecular identification in SIMS are strongly determined by this combination of ionization method and sample preparation. The use of TOF-SIMS for single cell and three-dimensional imaging will be discussed in the context of selected applications. Finally we will discuss an outlook on the application of the TOF-SIMS technology in a multimodal molecular imaging context.
Full-text available
Acute myocardial infarction (MI) is a cardiovascular disease which remains a major cause of morbidity and mortality world-wide despite advances in its prevention and treatment. During acute myocardial ischemia, the lack of oxygen switches the cell metabolism to anaerobic respiration, with lactate accumulation, ATP depletion, Na+ and Ca2+ overload and inhibition of myocardial contractile function, which drastically modifies the lipid, protein and small metabolite profile in the myocardium. Imag-ing mass spectrometry (IMS) is a powerful technique to comprehensively elucidate the spatial distribution patterns of lipids, peptides, and proteins in biological tissue sections. In this work, we demonstrate an application of multimodal chemical imag-ing using matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) which provided comprehensive molecu-lar information in situ within the same mouse heart tissue sections with myocardial infarction. MALDI-IMS (at 30 μm per pix-el) revealed infarct-associated spatial alterations of several lipid species of sphingolipids, glycerophospholipids, lysophospho-lipids and cardiolipins along with the acyl carnitines. Further, we performed trimodal MALDI-IMS (IMS3) where dual polarity lipid imaging was combined with subsequent protein MALDI-IMS analysis (at 30 μm per pixel) within the same tissue sections which revealed accumulations of core histone proteins H4, H2A and H2B along with post-translational modification products, acetylated H4 and H2A on the borders of the infarcted region. This methodology allowed us to interpret the lipid and protein molecular pathology of the very same infarcted region in a mouse model of myocardial infarction. Therefore, the presented data highlight the potential of multimodal MALDI imaging mass spectrometry of the same tissue sections as a powerful approach for simultaneous interrogation of spatial infarct-associated lipid and protein changes of myocardial infarction.
Full-text available
Three-dimensional (3D) histology is a new advanced tool for cancerology. The whole chemical profile and physiological characteristics of a tissue is essential to understand the rationale of pathology development. However, there is no analytical technique, in vivo or histological, that is able to discover such abnormal features and provide a 3D distribution at microscopic resolution.Here, we introduce a unique high- throughput infrared (IR) microscopy method that combines automated image correction and subsequent spectral data analysis for 3D-IR image reconstruction. I performed spectral analysis of a complete organ for a small animal model, a mouse brain with animplanted glioma tumor. The 3D-IR image is reconstructed from 370 consecutive tissue sectionsand corrected using the X-ray tomogram of the organ for an accurate quantitative analysis of thechemical content. A 3D matrix of 89 x 106 IR spectra is generated, allowing us to separate the tumor mass from healthy brain tissues based on various anatomical, chemical, and metabolic parameters. I demonstrate for the first time that quantitative metabolic parameters (glucose, glycogen and lactate) can be extracted and reconstructed in 3D from the IR spectra for the characterization of the brain vs. tumor metabolism (assessing the Warburg effect in tumors). Our method can be further exploited by searching for the whole spectral profile, discriminating different anatomical landmarks in the brain. I demonstrate this by the reconstruction of the corpus callosum and basal ganglia region of the brain.
Full-text available
Mass spectrometry imaging is becoming an important complementary alternative technique to more well-established imaging techniques. It can determine the spatial distribution of a large number of intact molecular or fragment ions generated directly from tissue sections or single cells. Secondary ion mass spectrometry coupled with a time-of-flight mass spectrometer imaging technique, using a cluster ion beam as primary bombardment particles, is now an analytical tool of reference for high lateral resolution imaging (1–2 µm) of biological surfaces. It exhibits a high efficiency for lipids and inorganic species. While numerous studies have been performed with cells in culture or with animal models, few clinical studies dedicated to lipid perturbations have been performed, to date. This article will compare the main advantages of time-of-flight mass spectrometry imaging over other mass spectrometry imaging methods and several biological applications will be detailed.
Full-text available
Over the past decade, dramatic increases in computational power and improvement in image analysis algorithms have allowed the development of powerful computer-assisted analytical approaches to radiological data. With the recent advent of whole slide digital scanners, tissue histopathology slides can now be digitized and stored in digital image form. Consequently, digitized tissue histopathology has now become amenable to the application of computerized image analysis and machine learning techniques. Analogous to the role of computer-assisted diagnosis (CAD) algorithms in medical imaging to complement the opinion of a radiologist, CAD algorithms have begun to be developed for disease detection, diagnosis, and prognosis prediction to complement the opinion of the pathologist. In this paper, we review the recent state of the art CAD technology for digitized histopathology. This paper also briefly describes the development and application of novel image analysis technology for a few specific histopathology related problems being pursued in the United States and Europe.
Time-of-flight secondary ion mass spectrometry (ToF-SIMS) is emerging as a tool for studying the metabolism of disease. ToF-SIMS enables chemical specificity in addition to high spatial resolution imaging of biological samples from cells to tissue. Here, ToF-SIMS has been used to investigate the metabolic regulation of hypoxia-induced chemoresistance to doxorubicin treatment using multicellular tumour spheroids. Imaging principal component analysis (PCA) was used as a tool to identify the regions of chemistry present within the image that differ as a result of drug treatment. A series of metabolite ToF-SIMS spectra were acquired, which were used to identify quasi-molecular ions and fragments correlated to the PCA loading plots. Metabolite patterns have been identified as potential biomarkers of hypoxia-induced chemoresistance. Copyright © 2012 John Wiley & Sons, Ltd.
Imaging mass spectrometry technology is rapidly developing. New desorption and ionization methods allow access to an increasingly large number of biomolecular systems. Sensitivity, speed and spatial resolution are continuously improving with new technologies such as cluster ion sources and microscope mode imaging approaches. MALDI imaging and SIMS are providing complementary molecular insights into biomolecular distributions on the surfaces of tissue sections and cells and demonstrate the great promise of biomolecular mass spectrometric imaging.There is one element, all MS imaging researchers agree upon. Sample preparation is crucial to the success of the method. Unfortunately, each application and each MS imaging technology requires a different type of sample preparation. Matrix coating, metal coating, sample morphology, sample history, and local chemical environment all influence the desorption and ionization mechanisms that lie at the basis of all imaging techniques. As images by themselves are a semi-quantitative representation of molecular distributions this obviously raises questions on their reliability.In this paper, a discussion will be devoted to the basic sample preparation requirements for biomedical imaging mass spectrometry. The differences and similarities for the two major imaging MS methods, SIMS and MALDI, will be addressed. Examples of extreme compound suppression as well as different matrix preparation methods for ME-SIMS and MALDI will be discussed.
The total potassium and sodium content was studied in various portions and tissues of the heart of 25 frogs, 40 rats, 26 rabbits, 3 cats, 2 dogs and 18 oxen. The myocardium of different portions of the heart may be placed in the following order according to the rise of the potassium content: the right and left auricle, the right and left ventricle. The sodium content in these portions decreases in reverse order to the potassium. Level K/Na coefficient for auricular myocardium approaches 1, and for ventricular myocardium, 2. In specific muscles of the ox heart the content of potassium is lower than in the myocardium and increases from the sinus and atriventricular nodes to the bundle of His and the fasciculi of the bundle of His. The sodium content therein has a reverse distribution gradient. Their K/Na coefficients are correspondingly 0.4, 0.6 and 0.8. The sum of potassium and sodium content in the myocardium of auricles and ventricles is the same. The sinus and atrioventricular nodes are characterized by the highest sum of potassium and sodium level as compared with other portions of the conduction system and myocardium.
The distribution pattern of lipid species in biological tissues was analyzed with imaging mass spectrometry (TOF-SIMS; time-of-flight secondary ion mass spectrometry). The first application shows distribution of a glycosphingolipid, the galactosylceramide-sulfate (sulfatide) with different hydrocarbon chain lengths and the fatty acids palmitate and oleate in rat cerebellum. Sulfatides were seen localized in regions suggested as paranodal areas of rat cerebellar white matter as well as in the granular layer, with highest concentrations at the borders of the white matter. Different distribution patterns could be shown for the fatty acid C16:0 palmitate and C18:1 oleate in rat cerebellum, which seem to origin partly from the hydrocarbon chains of phosphatidylcholine. Results were shown for two different tissue preparation methods, which were plunge-freezing and cryostat sectioning as well as high-pressure freezing, freeze-fracturing and freeze-drying.The second application shows TOF-SIMS analysis on a biological trial of choleratoxin treatment in mouse intestine. The effect of cholera toxin on lipids in the intestinal epithelium was shown by comparing control and cholera toxin treated mouse intestine samples. A significant increase of the cholesterol concentration was seen after treatment. Cholesterol was mainly localized to the brush border of enterocytes of the intestinal villi, which could be explained by the presence of cholesterol-rich lipid rafts present on the microvilli or by relations to cholesterol uptake. After cholera toxin exposure, cholesterol was seen increased in the nuclei of enterocytes and apparently in the interstitium of the villi.We find that imaging TOF-SIMS is a powerful tool for studies of lipid distributions in cells and tissues, enabling the elucidation of their role in cell function and biology.
Matrix-assisted laser desorption-ionization (MALDI) mass spectrometry (MS) is a rapid and sensitive analytical approach that is well suited for obtaining molecular weights of peptides and proteins from complex samples. MALDI-MS can profile the peptides and proteins from single-cell and small tissue samples without the need for extensive sample preparation, except for the cell isolation and matrix application. Strategies for peptide identification and characterization of post-translational modifications are presented. Furthermore, several recent enhancements in MALDI-MS technology, including in situ peptide sequencing as well as the direct spatial mapping of peptides in cells and tissues are discussed.
Secondary ion mass spectrometry is commonly used to study many different types of complex surfaces. Yet, compared with MALDI and ESI–MS, SIMS has not made a significant impact in biological or biomedical research. The key features of the technique, namely high spatial resolution, high detection efficiency of ions spanning a wide m/z range, surface sensitivity and the high scan rates seem to match ideally with several questions posed at the cellular level. To this date, SIMS has had only limited success in the biological arena. Why is this and what is needed to change this? This discussion paper will critically review the advances and the usefulness of SIMS in biomedical research and compare it to other approaches that offer spatially resolved molecular information available to a researcher with a biological interest. We will demonstrate that the type of information generated by the various incarnations of SIMS is strongly dependent on sample preparation and surface condition and these phenomena are only poorly understood. Modern approaches such as the cluster gun developments, ME-SIMS, gold coating and MALDI stigmatic imaging on a SIMS instrument might change the perception of SIMS being a tool for semiconductor manufacturers and physicists, and might persuade biologists to use these innovative mass spectrometric imaging tools.
Conference Paper
The output resolution of imaging mass spectrometers is increasing rapidly due to advances in engineering and the use of tiling. Imaging-MS data is often displayed as a total-ion-count (TIC) image; however, anatomical structures are not easily identifiable from TIC images. For this purpose, additional high-resolution images that originate from different imaging modalities, such as stained histological data, are preferred. These modalities are most useful when fused; i.e., when the corresponding images are spatially aligned with respect to each other. The viewing and analysis of such data is ideally performed in real-time and at the highest possible resolution, allowing users to interactively query the combination of all fused data at the highest detail. However, proper alignment between modalities and interactively presenting large volumes of data is as of yet a challenge. We present a system for the simultaneous viewing and analysis of high-resolution data from different imaging modalities. Fusion is provided in such a way that interaction in one modality can be mapped to different modalities. For example, anatomical structures can be identified from histological data and their spatial extent mapped to a corresponding region-of-interest in the image MS data, allowing the analysis of its chemical compounds. In turn, the MS data can be analysed and filtered, for example using multi-variate analysis such as PCA, and the result mapped back to structures in other modalities. Level-of-detail, region-of-interest and asynchronous data processing algorithms ensure that the system can be operated interactively at the highest resolution.
A two-dimensional gel electrophoresis database of dog (Canis familiaris) proteins is presented. The database contains 1212 protein spots which have been characterised in terms of their pI and Mr. This database has been integrated into the HSC-2DPAGE database which is accessible on the Internet via the World Wide Web with the uniform resource location (URL): ( protein/index.html). Identifications for 80 of the protein spots have been obtained by visual cross-matching with the human heart protein database in HSC-2DPAGE (42 spots), N-terminal microsequence analysis (25 spots) and peptide mass fingerprinting (20 spots). This database is being used in studies of alterations in protein expression in models of heart failure and heart disease.