Multi-Dimensional Organic Mass Cytometry: Simultaneous Analysis of Proteins and Metabolites on Single Cells

Article abstract of DOI:10.1002/anie.202009682

Mass cytometry is attracting significant attention for enabling spatiotemporal high-throughput single-cell analysis. As the first demonstration of the simultaneous detection of single-cell proteins and untargeted metabolites, a multi-dimensional organic m

Full text of DOI:10.1002/anie.202009682

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: Multi-Dimensional Organic Mass Cytometry: Simultaneous
Analysis of Proteins and Metabolites on Single Cells
Authors: Shuting Xu, Mingxia Liu, Yu Bai, and Huwei Liu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202009682
Link to VoR: https://doi.org/10.1002/anie.202009682

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
Multi-Dimensional Organic Mass Cytometry: Simultaneous
Analysis of Proteins and Metabolites on Single Cells
Shuting Xu, Mingxia Liu, Yu Bai,* and Huwei Liu
Abstract: Mass cytometry is attracting significant attention for its
spatiotemporal high-throughput single-cell analysis. As the first
demonstration of simultaneous detection of single-cell proteins and
untargeted metabolites, a multi-dimensional organic mass cytometry
system was established by a simple microfluidic chip connected to a
nanoelectrospray mass spectrometer, providing useful cell
heterogeneous information. A series of mass probes with online-
dissociated mass tags were developed, ensuring the semi-
quantification of cell surface proteins and compatibility of endogenous
metabolite detection at single-cell level. Six cell surface antigens and
plasma mass spectrometry (ICP-MS).[7] It achieves sufficient
single-cell protein sensitivity through abundant heavy-metal
isotopes chelated on antibodies,[8] and enables the rapid and
simultaneous measurements of >40 parameters in many ground-
breaking cell biology research,[9] which is unbeatable when
compared to fluorescence-based flow cytometry. However, the
high cost of limited isotopes and the specialized nature of
elemental MS limit its versatility and universality, and the ICP-MS
displays inherent disadvantages for the detection of endogenous
metabolites.
~
100 metabolites from three ovarian cancer cell types and two breast
In recent years, efforts have been made towards single-cell
metabolite profiling using commercial organic MS.[ Electrospray
ionization (ESI) MS, with increased simplicity and versatility over
10]
cancer cell types were successfully monitored, and contributed to
highly sensitive and specific cell typing. Doxorubicin-resistant cancer
cell analysis confirmed its applications in distinguishing rare cell
phenotypes. As a new generation of mass cytometry, the proposed
system is simple, extensible, and promising for cell typing, drug-
resistance analysis of tumor cells, and clinical diagnosis and therapy
at single-cell level.
[
10b,
CyTOF, has been developed as the detector of flow cytometry.
1
0d]
Hundreds of metabolites have been successfully identified in
single cells, providing direct information to indicate cellular
behavior for potential cell subtyping and disease diagnosis.
Unfortunately, although ESI-MS has been applied in single-cell
proteomics,[4b] the direct monitoring of intact proteins along with
metabolites in single cells remains a daunting challenge because
of insufficient sensitivity and complicated matrix interference.[
Organic mass probes for protein labeling have highly increased
the sensitivity,[12] but existing probes can hardly be compatible
with mass cytometry system. Mass cytometry conducting the
simultaneous detection of proteins and metabolites would
undoubtedly yield multi-dimensional single-cell data for more
comprehensively cellular and biological understanding.
11]
Introduction
The detection and understanding of individual cells within
populations is fundamental to many basic biological processes
such as differentiation, aging, and pathopoiesis.[1] Considering the
limited size and extensive information of single cells from the
cellular genome, transcriptome, and proteome to metabolome,
technologies has surged towards single-cell-level sensitivity, high
information coverage, and spatiotemporal high-throughput.
Compared to single-cell sequencing , single-cell proteome and
metabolome[5] are of great significance since they are the direct
performers of vital movement and the final products and essential
mediators of cellular behavior, respectively. Moreover, their
detection is challenging because of their large diversity, low
abundancy, and lack of amplification ability. High sensitivity and
rich information have made mass spectrometry (MS) a powerful
In this work, we proposed a nanoelectrospray ionization mass
spectrometry (NanoESI-MS) based multi-dimensional organic
mass cytometry system (Scheme 1a and Figure S1), which
enabled the simultaneous analysis of multi-parameters of proteins
and metabolites from single cells. A series of mass probes (MPs)
were prepared by assembling rhodamine-based mass tags
[2]
[3]
[4]
(
RMTs) and specific antibodies on gold nanoparticles (GNPs).
The RMTs, self-assembled on the GNPs through the Au–S bond
Scheme 1b) and on-line dissociated during the nanoESI process,
(
[
4b, 5a,
provided the transformation of protein signals into the amplified
and simplified signals of RMTs and ensured the single-cell-protein
sensitivity and high throughput. Importantly, cellular metabolites
could be directly detected at the same time, and the mass-to-
charge ratios (m/z) and intensities of RMTs were compatible with
those of the metabolites, thereby facilitating the complementary
data acquisition and downstream quantitation of both proteins and
metabolites. For efficient cell dispersing and ordering, a simple
microfluidic chip (Figure S2) was connected to the
nanoelectrospray emitter (Figure S3), which established the chip-
nanoESI mass cytometry system (Scheme 1a) and further
ensured the single-cell detection with high temporal throughput.
The proposed organic mass cytometry was demonstrated with
three ovarian cancer cell types (A2780, OVCAR-3, and SK-OV-3)
and two breast cancer cell types (MCF-7 and MDA-MB-231). Cell
tool for protein and metabolite measurement in single cells.
6
]
The combination of MS and flow cytometry further promotes the
temporal and spatial throughput, which was demonstrated by the
cytometry by time-of-flight (CyTOF) based on inductively coupled
[
*]
S. Xu, Dr. Mingxia Liu, Prof. Y. Bai, Prof. H. Liu
Beijing National Laboratory for Molecular Sciences,
Key Laboratory of Bioorganic Chemistry and Molecular Engineering
of Ministry of Education,
College of Chemistry and Molecular Engineering, Peking University,
Beijing 100871, P. R. China.
Tel: +86 10 6275 8198
E-mail: yu.bai@pku.edu.cn
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
identification, cell typing, and cell heterogeneity analysis were
achieved on the level of six surface proteins and ~100 metabolites.
Heterogeneous cancer cell resistance to anticancer drugs was
presented as another successful application of the organic mass
cytometry.
assembly quantities of antibodies and RMTs on GNPs were
systematically optimized and finalized as a ratio of 1/24/9000 for
GNP/antibody/RMT (Figure S7), balancing the high specific
recognition towards the cell surface proteins and the large
amplification by the tags. The modified GNPs were verified with
obvious fluorescent signals of antibody (Figure S8-9) and
fluorescence imaging of labeled cells (Figure S10), and observed
to be ~20 nm in diameter with a modified corona of 2 nm thickness
(
Figure S11) using transmission electron microscopy (TEM)
images and fluorescence imaging. A 20 min MP incubation time
was employed, which further reduced the nonspecific adsorption
and avoided probe endocytosis (Figure S12). Take EpCAM as an
example, results from above experiments confirmed that it can be
successfully detected on the single A2780 cells but not on MDA-
MB-231 cells (negative control).
More importantly, the self-assembling RMTs on the GNPs were
proven to be on-line dissociated during nanoESI, which was
extremely significant for high-throughput mass cytometric
detection. The dominant peak of the RMT443 dimers (m/z 628.4,
z = 2) was confirmed as the distinguishable doubly charged peaks
in the mass spectra (Figure S13a). As the most major products of
Au–S bond cleavage in laser-[ and plasma- based ionization,
dimers were achieved for the first time during the nanoESI flow.
The spray voltage was the most influential parameter of RMT
dissociation during nanoESI process, and interesting dissociated
products were obtained in the range of 1–6 kV (Figure S14). The
RMT dimers (m/z 628.4, z = 2) were the main dissociated
products at the voltage 3 kV. Higher voltage caused a higher yield
of oxidative products of dissociated RMTs (e.g., m/z 620.4, z = 2;
Figure S13b). The dissociation behavior was different from the
12b]
[13]
electrospray accelerated substrate ESI-MS in which multistep
Scheme 1. (a) Schematic of the multi-dimensional chip-nanoelectrospray
ionization (Chip-nanoESI) organic mass cytometry and its workflow for single-
cell analysis, including cell labeling by mass probes, cell injection and ordering
by the chip, dissociation and ionization by nanoESI and high-resolution MS
detection. (b) Chemical structures of six rhodamine-based mass tags (RMTs:
RMT331, RMT387, RMT415, RMT443, RMT467 and RMT491) self-assembled
on gold nanoparticles (GNPs) by Au–S bond.
oxidized derivatives presented because of the air oxygen.[12d]
Simple and intensive mass reporters were achieved in this
solvent-phase on-line dissociation environment. The specific
doubly charged dimers were utilized as the final MS reporters
which offered simple and clean spectrum with the spray voltage
of 3 kV. Six homologous RMTs presented identical dissociation
abilities, providing similar MS responses without cross
interference (RMT331, m/z 516.2443; RMT387, m/z 572.3067;
RMT415, m/z 600.3380; RMT443, m/z 628.3692; RMT467, m/z
Results and Discussion
6
52.3690; RMT491, m/z 676.3689; Figure S15 and Table S1).
MP Preparation and On-Line Dissociation of Mass Tags.
Novel MPs were developed through self-assembling of antibodies
and specially designed mass tags, RMTs, on the GNPs (Figure
S6), which relieved the bottlenecks of organic mass cytometric
Furthermore, with the aid of crystal violet (CV) as internal
standard (IS), six MPs provided good semi-quantitation results
using their reporter signals above (Figure S16). The high
specificity, high sensitivity, online dissociation, and simultaneous
semi-quantification made the MPs ideal protein labeling probes
for the organic mass cytometry.
system including sensitivity, dissociation modes and compatibility
of existing protein labeling probes.[8, 12b]
Configuration and Performance of Chip-NanoESI Mass
Cytometry. Organic mass cytometry exhibits the advantages of
simplified and flexible ionization interfaces and good compatibility
with various commercial mass spectrometers. The chip-nanoESI
mass cytometry platform comprised four main parts (Figure S1):
the cell injection system, cell ordering chip (Figure S2), home-
made nanoESI device (Figure S3), and high-resolution Orbitrap
mass spectrometer.
High specificity and sensitivity as the two key points for single-cell
protein detection were ensured by the antibodies and RMTs
assembled on MPs. Antibodies were utilized for target protein
recognition according to their specific affinity and universality in
life analysis and clinical diagnosis. Six homologous RMTs
(
RMT331, RMT387, RMT415, RMT443, RMT467, and RMT491,
Scheme 1b and Figure S4) contributed to single-cell sensitivity
because of their large quantities on GNPs and high MS response,
one of which have been confirmed to achieve a limit of detection
of zeptomole for free proteins in our previous work.[12d] The self-
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
NanoESI with a tip size of 30 μm O.D. (Figure S3) was employed
with a 3 kV DC voltage and an adaptive flow rate of 1 μL/min
breast cancer cell types (MCF-7 and MDA-MB-231) in sequence
with the same workflow (Figure S5), achieving the single-cell
mass spectra of five different cells (Figure S20). Furthermore, a
mixed cell suspension composed of three types of cells (A2780,
OVCAR-3, and MCF7) was injected and presented single cell
signals as well (Figure S21a). Three unique mass spectra were
extracted from three individual peaks (Figure S21b-d), which were
assigned to three cell types after data analysis.
(
Figure S17), which offered multiple functions including protein
tag dissociation, cell lysis or rupture and ionization. To evaluate
its performance, the labeled A2780 cell suspension was first
attempted. A continuous low-speed vortex of cell suspension was
kept during the injection to ensure the homo- and mono-
dispersion of cells. To obtain results from a single cell, we found
the concentration of cell suspension was crucial. As depicted in
the total ion chromatograms (TICs) and extracted-ion
chromatograms (EICs) (Figure S18), a high cell concentration
6
(
>10 cells per mL) resulted in a continuous signal from more than
1
cell and even the clogging of the emitter, and isolated peak-like
signals probably from individual cells were observed when using
4
a moderate concentration (<5×10 cells per mL) compatible with
an MS screening speed of ~6 scans/s. The m/z 782.5670,
assigned
to
the
representative
cell
metabolite-
glycerophosphocholine PC(34:1), was selected as the marker of
the presence of cells. The position of peak-like signals in the EIC
of m/z 782.5670 matched well with the TICs, further assigning the
peak-like signals to single cells, and the negligible background
signal between the EIC peaks indicated the absence of any
interference in the solvent or between the cells (Figure S18b).
However, the signals from the cell clusters frequently disturbed
the single-cell detection process.
To effectively regulate cell dispersion and ordering, a five-loop
microchannel chip was connected to the system. The curved
square microchannel in the chip provided the regulating effect
based on the Dean flow.[14] As a demonstration, the cell clusters
6
in a high-concentration cell suspension (>10 cells per mL; Figure
S19a) achieved mono-dispersion, and the spacing between the
cells kept increasing after each cycle. For the MS-compatible
concentrations (~5×104 cells per mL), the introduction of the
microchannel chip almost completely resolved cell clusters and
aligned the cells in an orderly manner (Figure S19b). This
provided a series of more uniform pulse-like single-cell peaks in
the TICs and EICs [PC(34:1) and m/z 572.3066 assigned to
RMT387] (Figure 1a). For a single run, the cytometric test could
last ~ 6 minutes for the analysis of more than 200 cells with a
throughput of ~40 cells per min. Single cell analysis duration was
Figure 1. Performance of the chip-nanoelectrospray ionization (Chip-nanoESI)
mass cytometry system. (a) Total-ion chromatogram (TIC) and extracted-ion
chromatograms [EICs; glycerophosphocholine PC(34:1) of m/z 782.5670 and
RMT387 of m/z 572.3066] of the cell suspension (~5×104 cells/mL); enlarged
pulse-like peaks representing the analysis duration of individual cells (∼0.32 s
with ~40 cells/min). (b) A2780 single-cell mass spectrum extracted from the TIC
∼
0.32 s and the gap duration between cells was ~1 s (Figure 1a).
For semi-quantification of the cellular protein tags and metabolites,
the CV intensities (m/z 372.2433 in Figure 1b) were utilized for
signal normalization, which reduced the possible variability during
the ionization process (relative standard deviation (RSD) for CV =
(pulse-like peak marked with the red dotted line), providing cellular metabolite,
protein tag, and internal standard (IS) signals. (c) Signal plots of m/z 782.5670
and 572.3066 from 151 sequential detected A2780 cells using the signal
intensity ratios of target signal and IS for semi-quantification. (d) Intensity ratio
distribution of m/z 782.5670 and 572.3066 from 151, 142, and 150 A2780 cells
8
.7%, n = 151), and highlighted the intercellular differences. The
normalized signals of m/z 782.5670 and 572.3066 from the
sequentially detected cell 1 to cell 151 in one injection are
illustrated in Figure 1c. Significant signal differentiation among the
individual cells confirmed the necessity of single-cell analysis.
The signal distribution within the cell populations were confirmed
by ~500 cells from three tests (151, 142, and 150 cells). The
percentage of A2780 cell populations with certain intensity ratios
in three tests presented acceptable RSDs (11% and 6% for the
highest percentages for signals m/z 782.5670 and 572.3066, n =
(
most RSDs <20%, n = 3). RSD: relative standard deviation.
Surface Antigen Detection in Single Cells. Six antigens on cell
surfaces, including three important cancer biomarkers, cancer
antigen 125 (CA125), carcinoembryonic antigen (CEA), and
epithelial cell adhesion molecule (EpCAM), and three commonly
used cluster of differentiation (CD) antigens, CD24, CD44, and
CD133, were labeled with six MPs, detected by the organic mass
cytometry and then semi-quantitated.
3
; Figure 1d). All the above data indicated good reliability and
stability for this organic mass cytometry platform.
More validations for the platform were conducted with two more
ovarian cancer cell types (OVCAR-3, and SK-OV-3) and two
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
Benefiting from the high sensitivity RMTs and large signal
amplification ability of the MPs, the six cell surface antigens were
successfully detected in the single-cell level (Figure 1b, green bar
peaks and Figure S20, pink bar peaks). The mass reporter signals
facilitated rapid identification in the complex matrix because of
their characteristic doubly charged peaks and the absence of
interference with the signals from cellular endogenous
compounds. The semi-quantification of the surface antigens was
achieved with the intensity ratios of the mass reporters and IS.
The expression of the surface antigens among the five cell types
and subtypes was presented with significant differences (Figure
confirmed the specific labeling and detection of the single-cell
proteins using MP-based organic mass cytometry.
These differences in cell surface antigens might be closely related
to cellular function, such as proliferation and differentiation,
epithelial cell adhesion, metastasis, and signaling regulation for
different cell types and subtypes. As the typical indicators for cell
typing,[
15]
different cell types and subtypes including ovarian
cancer cells (A2780, OVCAR-3, and SK-OV-3) and breast cancer
cells (MCF-7 and MDA-MB-231) were clearly distinguished with
cellular antigen information after hierarchical clustering and
visualized by principal component analysis (PCA; Figure S24),
representing high sensitivity and specificity for most cell lines but
a little bit confused for the identification of MDA-MB-231 cells
(Most of the error scatters in Figure S24b).
Metabolite Profiling in Single Cells and Cell Typing. Apart
from cell surface proteins, cellular metabolites were
simultaneously identified in the single-cell mass spectra (Figure
S20), taking full advantages of organic mass cytometry. Among
the >500 detected peaks in the mass spectra of single cells, ~84
potential cellular metabolites were successfully identified based
on the exact m/z ratio in the m/z 80–1200 under positive mode
2
a). For example, as an important functional protein of epithelial
carcinogenesis, EpCAM was detected in most epithelial
carcinoma cell lines, including SK-OV-3, MCF-7, A2780, and
OVCAR-3 but not in the MDA-MB-231 cell line.
(
Table S2). Since direct tandem MS of the metabolites from single
cells was not performed because of the limited scan time (∼0.32
s) for one cell, metabolite extractions of the bulk cells were
analyzed to obtain the exact mass and tandem MS information to
assist metabolite assignment (Figure S25). The assigned
metabolites mostly included amino acids or peptides (proline,
valine, threonine, leucine, creatine, glutamine, arginine,
cysteinylglycine, and glutathione) and lipids [including PC,
glycerophosphoglycerols (PG), glycerophosphoethanolamines
(
PE), triglycerides (TG), fatty acid esters. and fatty amides; Table
S3]. Notably, most metabolites were extracted or released from
the cytomembrane and cytoplasm (Table S4). This was mainly
attributed to the short interaction time for cell lysis/rupture and
metabolite extraction during electrospray.
Significant statistical differences in the cellular metabolites along
with six surface antigens were observed among the different cell
types and subtypes (Figure 3a and Figure S26, 27). Several
amino acids and lipids displayed significant differences (Figure
S28, 29), for example PC(32:1), proline, PE(P-40:6), and
TG(52:2) between the A2780 and OVCAR-3 cell lines (Figure 3b).
Notably, different lipid expression levels among the different cell
lines were observed. Only three TGs, namely TG(50:2), TG(52:2),
and TG(52:3), were present within the A2780 cells with very high
intensities, while large amounts of high-intensity PCs, such as
PC(38:2), PC(36:2), and PC(30:0), were found within the MCF-7
cells. Most lipids were present at much lower expression levels in
the SK-OV-3 cells. The same tendencies were confirmed when
unlabeled cells were analyzed (Figure S30-31), thereby
eliminating the effect on metabolite detection induced by the MP
addition. As the lipids perform significant functions to form cellular
structures and participate in cell proliferation/differentiation and
multiple signaling pathways, the lipid differences of these cancer
cells were associated with different cell morphologies, generation
cycles, and functions to aid cancer classification and staging.[
The cellular metabolites were potential biomarkers for
distinguishing cells and could also be applied to assist cell type
and subtype identification. With the assigned 84 metabolites, five
cell types and subtypes were also successfully clustered by
Figure 2. Single-cell surface antigen detection. (a) Spider chart and heatmap of
the expression of six proteins: cluster of differentiation (CD) antigens (CD24,
CD133, and CD24), epithelial cell adhesion molecule (EpCMA),
carcinoembryonic antigen (CEA), cancer antigen 125 (CA125), in five cell types
and subtypes (A2780, OVCAR-3, SK-OV-3, MCF-7, and MDA-MB-231).
Distribution curves of the EpCAM expression in the five cell lines detected by
(b) organic mass cytometry and (c) fluorescence-based flow cytometry.
To verify the semi-quantification results, gold-standard
fluorescence-based flow cytometry was employed. The inherent
fluorescent signals from RMTs made these MPs difunctional
probes additionally for fluorescence-based flow cytometry.
Therefore, the MP-labeled cells were directly subjected to
commercial flow cytometry. The semi-quantification results
attained by fluorescence intensity measurements (Figure S22)
were well consistent with those afforded by the organic mass
cytometry, such as the relative expression of EpCAM visualized
by the distribution curves based on the MS intensity ratios (Figure
2
b) and fluorescence intensities (Figure 2c). Furthermore, the
16]
results attained using commercial fluorescent-labeled antibodies
were in satisfactory agreement with those attained with the MPs
in protein semi-quantification (Figure S23), which further
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
hierarchical clustering and visualized in PCA (Figure S32);
however, these still displayed a lower degree of discrimination for
MDA-MB-231 cells (Most of the error scatters in Figure S32b).
Since the organic mass cytometry system provided rich cellular
information including cell surface antigens and metabolites, a
heightened cell typing for the five cell lines with both cellular
proteins and metabolites was demonstrated for the first time
(
Figure S33). Thus, the combined protein/metabolite information
resulted in the most sensitive and specific differentiation results
for cell typing and identification, especially 100% for both
sensitivity and specificity for the MDA-MB-231 cells (Table S5).
Furthermore, the cell suspension of mixed A2780, OVCAR-3 and
MCF-7 cells were also successfully distinguished with
protein/metabolite information, which was preliminarily gathered
to three clusters in the PCA (Orange triangle scatters in Figure
S34), and highly consistent with the three identified cell clusters.
After hierarchical clustering using identified cells as reference, the
mixed unknown cells were clearly clustered into three cell types
as 74 proposed A2780 cells, 90 proposed OVCAR-3 cells, and 87
proposed MCF-7 cells (Figure 3c), which was also consistent with
the initial mixed ratios of cell suspension sample. The detection
and identification of mixed sample expends the platform for more
complicated and unknown system. All the above demonstrated
the organic mass cytometry platform being more comprehensive,
accurate and efficient for cell typing, and high potential for further
biological analysis at single-cell level.
Cell Heterogeneity Analysis for Chemotherapy Resistance.
Development of single-cell measurement techniques reveals cell
heterogeneity among cell types and more importantly in a clonal
population, which is crucial for drug resistance analysis and stem
cell research in cancer therapy.[1d, 17] To investigate the drug
resistant at the single-cell level, wild-type MCF-7 (wt MCF-7) cells
exposed to doxorubicin were employed to induce doxorubicin-
[18]
chemoresistant phenotype (DOX-res) MCF-7 cells.
The
differentiation of DOX-res and wt MCF-7 cells at the single-cell
level was conducted using our platform. The DOX-res cells
presented marked differentiation in the surface antigen
expression profiles, particularly the low expression level of CD24
(
CD24low/-), which has been reported as a specific marker for
chemoresistant phenotypes.[
17, 19]
The CD44 CD24
+
low/-
DOX-res
MCF-7 cells (Figure 4a, triangle scatters) were clearly
+
+
distinguished from most of the CD44 CD24 MCF-7 cells (Figure
a, circle scatters).
4
Representative metabolites with statistical differences were also
singled out between the DOX-res phenotype and wt MCF-7 cells,
including amino acids (arginine, glutamine, and glutamic acid)
and lipids [sphingomelin SM(D34:1), PC(38:4), and PC(O-32:0);
Figure 4b]. The differential metabolites could be associated with
the cellular metabolic pathways and special cell morphology,
which were potential indicators for the recognition of drug
resistant cells and provided phenotypic evidence for the analysis
of the drug resistance mechanism and the survival, differentiation,
and aging of cancer cells.
Figure 3. Metabolite profiling and cell typing. (a) Volcano plot of the correlations
between the P-values and fold changes (FD) of normalized intensities for the
detected mass tag and metabolite signals within A2780 and OVCAR-3 cells;
differential compounds (T test, P <0.05; FD >2) depicted in red and blue dots.
(
7
(
b) Distribution of the representative differential metabolites: PC(32:1) (m/z
34.5680), proline (m/z 116.0708), glycerophosphoethanolamines PE(P-40:6)
m/z 798.5403) and triglycerides TG(52:2) (m/z 897.7286) in A2780 and
Interestingly, in the results of the wt MCF-7 cell populations, a
very low number of cells with relatively low CD24 expression were
OVCAR-3 cells. (c) Principal component analysis (PCA) of five known cell types
and mixed cell sample based on the detected six mass tag and 84 metabolite
signals. Cells marked based on the results of hierarchical clustering (Ward’s
method, square Euclidean distance), mixed cells clustered to three groups:
proposed A2780 cells (red triangle scatters), proposed OVCAR-3 cells (brown
triangle scatters) and proposed MCF-7 cells (blue triangle scatters).
+
low/-
observed (circle scatters of CD44 CD24
in Figure 4a), and the
results were identical with those observed with fluorescence-
based flow cytometry (Figure S35). These small clusters in the wt
MCF-7 cells might be drug resistant cells and could have other
stem-like features, like self-renewal and differentiation, which
have been of great concern in cancer therapy.[20] The analysis of
drug-resistant cancer cells further demonstrated the organic mass
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
cytometry system as
technique for cell heterogeneity analysis and cell phenotype
identification and recognition.
a
powerful single-cell measurement
(2016YFF0100303) and Beijing Natural Science Foundation
Essential Research Project (Z170002).
Keywords: mass cytometry • single-cell heterogeneity• mass
tags • surface protein • metabolites
[
1]
a) A. Colman-Lerner, A. Gordon, E. Serra, T. Chin, O. Resnekov, D.
Endy, C. G. Pesce, R. Brent, Nature 2005, 437, 699-706; b) L. Bintu, J.
Yong, Y. E. Antebi, K. McCue, Y. Kazuki, N. Uno, M. Oshimura, M. B.
Elowitz, Science 2016, 351, 720-724; c) M. Yazawa, B. Hsueh, X. Jia,
A. M. Pasca, J. A. Bernstein, J. Hallmayer, R. E. Dolmetsch, Nature
2011, 471, 230-234; d) B. M. Seo, M. Miura, S. Gronthos, P. M. Bartold,
S. Batouli, J. Brahim, M. Young, P. G. Robey, C. Y. Wang, S. T. Shi,
Lancet 2004, 364, 149-155.
[
[
2]
3]
a) D. Wang, S. Bodovitz, Trends Biotechnol. 2010, 28, 281-290; b) A.
Schmid, H. Kortmann, P. S. Dittrich, L. M. Blank, Curr. Opin.
Biotechnol. 2010, 21, 12-20.
a) E. Z. Macosko, A. Basu, R. Satija, J. Nemesh, K. Shekhar, M.
Goldman, I. Tirosh, A. R. Bialas, N. Kamitaki, E. M. Martersteck, J. J.
Trombetta, D. A. Weitz, J. R. Sanes, A. K. Shalek, A. Regev, S. A.
McCarroll, Cell 2015, 161, 1202-1214; b) A. P. Patel, I. Tirosh, J. J.
Trombetta, A. K. Shalek, S. M. Gillespie, H. Wakimoto, D. P. Cahill, B.
V. Nahed, W. T. Curry, R. L. Martuza, D. N. Louis, O. Rozenblatt-
Rosen, M. L. Suva, A. Regev, B. E. Bernstein, Science 2014, 344,
Figure 4. Chemotherapy resistance subtype analysis. (a) Scatter plot of MCF-
7
(circles) and DOX-res MCF-7 (triangles) cells based on CD24 and CD44
expression. (b) Volcano plot of the correlations between the P-values and fold
changes (FD) of normalized intensities for the detected mass tag and metabolite
signals within MCF-7 and DOX-res MCF-7 cells, differential compounds (T test,
P <0.05; FD >2) illustrated as red and blue dots.
1
396-1401; c) P. Zhang, X. Han, J. Yao, N. Shao, K. Zhang, Y. Zhou,
Y. Zu, B. Wang, L. Qin, Angew. Chem. Int. Ed. 2019, 58, 13700-13705.
a) A. J. Hughes, D. P. Spelke, Z. Xu, C.-C. Kang, D. V. Schaffer, A. E.
Herr, Nat. Methods 2014, 11, 749-755; b) Y. Zhu, G. Clair, W. B.
Chrisler, Y. Shen, R. Zhao, A. K. Shukla, R. J. Moore, R. S. Misra, G. S.
Pryhuber, R. D. Smith, C. Ansong, R. T. Kelly, Angew. Chem. Int. Ed.
2018, 57, 12370-12374.
[4]
Conclusion
[
5]
a) R. Zenobi, Science 2013, 342, 1243259; b) M. Fessenden, Nature
2016, 540, 153-155.
In summary, a multi-dimensional organic mass cytometry platform
was developed for the simultaneous analysis of single-cell
proteins and metabolites. Sufficient sensitivity for single-cell
protein detection was achieved with six mass tags-RMTs
assembled on GNPs for signal transformation and amplification.
The online dissociation of RMTs and specific recognition using
antibodies provided the highly sensitive and specific semi-
quantification of six cell surface antigens at the single-cell level. A
facile integrated setup comprising cell injection, cell ordering, and
ionization was established, which could be easily coupled with a
high-resolution mass spectrometer for cytometric measurement.
Cell suspension was analyzed with a throughput of ~40 cells per
minute, providing six protein parameters and ~100 metabolite
parameters at single-cell resolution. Cancer cell phenotypes and
substantial heterogeneity were better distinguished and identified
based on the comprehensive cell surface antigens and cellular
metabolites, achieving more than 95% sensitivity and specificity
for cell typing. Moreover, the drug resistant and stem-like cells
within MCF-7 were recognized based on antigen markers, and
metabolic differences were found for further drug resistance
analysis and stem cell research. Combining significant protein
targets with hundreds of downstream metabolites, the multi-
dimensional mass cytometry offers a high possibility for the deep
understanding of fundamental biological processes such as
differentiation, aging, and pathopoiesis at single-cell level.
[6]
a) T. J. Comi, T. D. Do, S. S. Rubakhin, J. V. Sweedler, J. Am. Chem.
Soc. 2017, 139, 3920-3929; b) L. Zhang, A. Vertes, Angew. Chem. Int.
Ed. 2018, 57, 4466-4477.
M. H. Spitzer, G. P. Nolan, Cell 2016, 165, 780-791.
R. Liu, S. Zhang, C. Wei, Z. Xing, S. Zhang, X. Zhang, Acc. Chem. Res.
[7]
[
[
8]
9]
2016, 49, 775-783.
a) S. C. Bendall, E. F. Simonds, P. Qiu, E.-a. D. Amir, P. O. Krutzik, R.
Finck, R. V. Bruggner, R. Melamed, A. Trejo, O. I. Ornatsky, R. S.
Balderas, S. K. Plevritis, K. Sachs, D. Pe'er, S. D. Tanner, G. P. Nolan,
Science 2011, 332, 687-696; b) C. Boettcher, S. Schlickeiser, M. A. M.
Sneeboer, D. Kunkel, A. Knop, E. Paza, P. Fidzinski, L. Kraus, G. J. L.
Snijders, R. S. Kahn, A. R. Schulz, H. E. Mei, E. M. Hol, B. Siegmund,
R. Glauben, E. J. Spruth, L. D. de Witte, J. Priller, N. B. B. Psy, Nat.
Neurosci. 2019, 22, 78-90.
[10] a) E. K. Neumann, T. J. Comi, S. S. Rubakhin, J. V. Sweedler, Angew.
Chem. Int. Ed. 2019, 58, 5910-5914; b) H. Yao, H. Zhao, X. Zhao, X.
Pan, J. Feng, F. Xu, S. Zhang, X. Zhang, Anal. Chem. 2019, 91, 9777-
9783; c) Z. Yin, X. Cheng, R. Liu, X. Li, L. Hang, W. Hang, J. Xu, X.
Yan, J. Li, Z. Tian, Angew. Chem. Int. Ed. 2019, 58, 4541-4546; d) Q.
Huang, S. Mao, M. Khan, L. Zhou, J. M. Lin, Chem. Commun. 2018, 54,
2595-2598; e) Q. Huang, S. Mao, M. Khan, W. Li, Q. Zhang, J. M. Lin,
Chem. Sci. 2020, 11, 253-256.
[
[
11] G. Li, S. Yuan, S. Zheng, Y. Liu, G. Huang, Anal. Chem. 2018, 90,
3409-3415.
12] a) J. R. Lee, A. Lee, S. K. Kim, K. P. Kim, H. S. Park, W.-S. Yeo,
Angew. Chem. Int. Ed. 2008, 47, 9518-9521; b) Y. Wang, R. Du, L.
Qiao, B. Liu, Chem. Commun. 2018, 54, 9659-9662; c) W. Ma, S. Xu,
H. Nie, B. Hu, Y. Bai, H. Liu, Chem. Sci. 2019, 10, 2320-2325; d) S. Xu,
W. Ma, Y. Bai, H. Liu, J. Am. Chem. Soc. 2019, 141, 72-75.
13] R. K. Manova, S. Joshi, A. Debrassi, N. S. Bhairamadgi, E. Roeven, J.
Gagnon, M. N. Tahir, F. W. Claassen, L. M. W. Scheres, T. Wennekes,
K. Schroen, T. A. van Beek, H. Zuilhof, M. W. F. Nielen, Anal. Chem.
2014, 86, 2403-2411.
[
[
[
[
14] E. W. M. Kemna, R. M. Schoeman, F. Wolbers, I. Vermes, D. A. Weitz,
A. van den Berg, Lab Chip 2012, 12, 2881-2887.
15] N. Samusik, Z. Good, M. H. Spitzer, K. L. Davis, G. P. Nolan, Nat.
Methods 2016, 13, 493-496.
16] a) Y. Sunami, A. Rebelo, J. Kleeff, Cancers 2017, 10, 3; b) E. Currie, A.
Schulze, R. Zechner, T. C. Walther, R. V. Farese, Jr., Cell Metab. 2013,
Acknowledgements
18, 153-161.
[
17] K. Meirelles, L. A. Benedict, D. Dombkowski, D. Pepin, F. I. Preffer, J.
Teixeira, P. S. Tanwar, R. H. Young, D. T. MacLaughlin, P. K.
This work was financially supported by the National Natural
Science Foundation of China (21874003, 21527809, and
Donahoe, X. Wei, Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 2358-2363.
2
1728501), National Key R&D Program of China
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
[
18] P. C. Marinello, C. Pannis, T. N. Xavier Silva, R. Binato, E. Abdelhay, J.
A. Rodrigues, A. L. Mencalha, N. M. Dias Lopes, R. C. Luiz, R.
Cecchini, A. L. Cecchini, Scientific Reports 2019, 9.
[
19] a) E. Amir, A. Ocana, O. Freedman, M. Clemons, B. Seruga, Nat. Rev.
Clin. Oncol. 2010, 7, 79-80; b) S. Liu, Y. Cong, D. Wang, Y. Sun, L.
Deng, Y. Liu, R. Martin-Trevino, L. Shang, S. P. McDermott, M. D.
Landis, S. Hong, A. Adams, R. D'Angelo, C. Ginestier, E. Charafe-
Jauffret, S. G. Clouthier, D. Birnbaum, S. T. Wong, M. Zhan, J. C.
Chang, M. S. Wicha, Stem Cell Rep. 2014, 2, 78-91; c) A. M. Calcagno,
C. D. Salcido, J.-P. Gillet, C.-P. Wu, J. M. Fostel, M. D. Mumau, M. M.
Gottesman, L. Varticovski, S. V. Ambudkar, J. Natl. Cancer Inst. 2010,
102, 1637-1652.
[
20] G. Farnie, F. Sotgia, M. P. Lisanti, Oncotarget 2015, 6, 30472-30486.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202009682
RESEARCH ARTICLE
Entry for the Table of Contents
RESEARCH ARTICLE
A multi-dimensional organic mass
cytometry was established, enabling
simultaneous monitoring of six cell
surface proteins and ~100
Shuting Xu, Mingxia Liu, Yu Bai,* and
Huwei Liu
Page No. – Page No.
metabolites (including lipids and
amino acids) at single-cell level with
a throughput of ~40 cells/min. More
comprehensive protein and
metabolite parameters were applied
for heterogeneous single-cell
analysis, including cell typing,
chemotherapy resistance, and
stemness recognition.
Multi-Dimensional Organic Mass
Cytometry: Simultaneous Analysis
of Proteins and Metabolites on
Single Cells
This article is protected by copyright. All rights reserved.