Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Received: 26 July 2024
Revised: 03 October 2024
Accepted: 04 October 2024
Version of Record published:
18 November 2024
Review Article
Understanding metabolic plasticity at single cell
resolution
Christina C. Abbate, Jason Hu and John G. Albeck
Department of Molecular and Cellular Biology, University of California, Davis, U.S.A.
Correspondence: John G. Albeck (jgalbeck@ucdavis.edu)
It is increasingly clear that cellular metabolic function varies not just between cells of dif-
ferent tissues, but also within tissues and cell types. In this essay, we envision how differ-
ences in central carbon metabolism arise from multiple sources, including the cell cycle,
circadian rhythms, intrinsic metabolic cycles, and others. We also discuss and compare
methods that enable such variation to be detected, including single-cell metabolomics and
RNA-sequencing. We pay particular attention to biosensors for AMPK and central carbon
metabolites, which when used in combination with metabolic perturbations, provide clear
evidence of cellular variance in metabolic function.
Introduction
It has long been known that cells of the human body can vary dramatically in their metabolic function.
Skeletal muscle cells and adipocytes, for example, face very different demands in their function and in their
requirements for nutrients. While some of these differences arise developmentally, cells also navigate ex-
treme metabolic changes as part of their physiological function, such as the acceleration of glycolysis in
T cells upon their activation [1]. Modern data have not only underscored this concept but have also re-
vealed that cellular variation in metabolism is even more pervasive. Heterogeneity in metabolic functions
can readily be found among genetically identical cells of the same type, or even within the same cell at
different points in its day-to-day life [2]. Such variation in metabolism impacts the functions of individ-
ualcellsandtheirtissueasawhole.Consequently,anadequatedescriptionofcellularmetabolismforany
organ or physiological system needs to account for cellular metabolic plasticity.
Multiple approaches to quantify cell-to-cell variation in metabolism are now emerging, with some of
them requiring fixation of cells or tissue [3] and others operating in live cells [4]. These approaches have
ledtoanewappreciationofthedynamicvariabilityofmetabolicprocesses,bothincellcultureandin
vivo. Yet, a major challenge shared across these methods is the difficulty of inferring information about
metabolic function from the measurement of a single metabolite or process. Given the many branches in
metabolism, a change in a single indicator is often ambiguous in its interpretation. One solution to this
challenge is to pay particular attention to the cellular energy sensors that evolution has produced, with the
rationale that over millions of years they have evolved to provide metabolic information that is functionally
relevant to the cell [5]. One of the best-known examples is AMPK, a key regulatory kinase that monitors
multiple internal factors including energy charge (i.e., the ratio of ATP relative to ADP and AMP), glucose
concentration, and calcium [6]. By integrating these inputs, it provides a useful single-point status indica-
tor for metabolic information. Several fluorescent protein biosensors for intracellular AMPK activity are
nowavailable,makingitpossibletomonitorthisindicatorinrealtime,forhundredsorthousandsofcells,
both in vitro and in vivo [2,7,8]. This technology has revealed a striking amount of variation across cells,
subcellular locations, and time. The key challenge now is to place this observed variability into context
and to understand its cellular and physiological significance.
As a sensor for adenylate energy charge, AMPK is closely connected to central carbon metabolism
(CCM), the process by which cells use nutrients to phosphorylate ADP to generate ATP. The two major
©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution
License 4.0 (CC BY-NC-ND).
273
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Glucose (Glc)
GLUT1 GLUT4
o
o
o
-
Glc
G6P
O
OH
HO
CH2OH
OH
OH
HK/GK
PGI
F6P
FBP
DHAP
G3P x2
BPG x2
3PG x2
Pyruvate x2
TPI
GAPDH
2PG x2
PEP x2
PFK
Ald
PGK
PGM
ENO
PK
TCA Cycle
;ul;m|-ঞom
oH
o
oH
oH
o
oH
MPC
Lactate x2
PI3K
Protein
Synthesis
AMPK
mTORC2
oH
o
oH
mTORC1
MCT
LDH
ATP
ATP
ATP
NADH
NAD+
ATP
ATP
Ribosome
ATP
to PPP
o
o
o
-
AKT
S
o
CoA
S
o
CoA
Acetyl-CoA
Citrate IC ÑŊ
S-CoA
Succinate
Fumarate
Malate
OAA
Insulin
NADH
ATP
NAD+
ATP
ATP
ATP
OxPhos
Perceval
ATP/ADP
Peredox
NADH/NAD+
ATP
G
L
U
;ul;m|-
P
P
P
P
Per
P
P
A
A
ATP/
P
P
Per
TCA
ETC
Laconic
ACO2
D
NAD
H+ H+
iATPSnFR2
$1om1;|u-ঞom
CS
IDH OGDH
SCS
SDHFH
MDH2
III III IV
ATP Synthase
1
1
2
2
3
3
3
4
4
4
5
6
7
8
9
11
8
8
11
iGlucoSnFR
sub: glucose
Keller et al. 2019
HYlight
sub: FBP
Koberstein et al. 2022
Pyronic
sub: pyruvate
Arce-Molina et al. 2020
ExRai-AKTAR
sub: AKT sub.
Mehta et al. 2018
AMPKAR
sub: AMPK sub.
"1_lb;|-ѴĺƑƏƑƑ
Laconic
sub: lactate
"-m-uঠm;|-ѴĺƑƏƐƒ
Perceval
sub: ATP/ADP
Berg et al. 2009
Peredox
sub: NADH/NAD+
Hung et al. 2011
iATPSnFR2
sub: ATP
-ubm;|-ѴĺƑƏƑƒ
6
ATP
x30-36
7
NAD+
NADH
Glycolysis
mtATeam1.03
sub: ATP
Yoshida et al. 2017
5
ExRai-AKTAR
AMPKAR
ATP
ATP
ADP
NADH
NAD+
Pyronic
mtATeam1.03
HYlight
iGlucoSnFR
TMRM
mito-membrane
ro|;mঞ-ѴŐl(ő
NADHNAD+
NADH
NAD+
NADH
NAD+
ATP
TMRM
11
12
Ehrenberg et al. 1988
SoNAR
NADH/NAD+
/
/
S
S
SoN
NAD
NAD
9
NADH
NAD+
10
10
12
9
SoNAR
sub: NADH/NAD+
Zhao et al. 2015
V
Figure 1. Schematic diagram of CCM, associated regulators and live-cell reporters.
Several of the major catabolic routes involved in CCM are depicted using a simplied formalism. Protein synthesis and its network
of regulatory kinases are shown as an example of an ATP-consuming anabolic process that is closely interconnected with CCM.
Shown throughout the diagram are examples of biosensors or other reagents that have been used to track CCM function within
living cells; green cylinders represent biosensors based on cpGFP, and blue/yellow cylinders represent biosensors based on FRET
interactions between cyan and yellow uorescent proteins. Relevant references for each sensor are listed in the legend at right. The
set of biosensors shown is not exhaustive, and there are many additional tools we were unable to include due to space constraints.
branches of CCM, glycolysis and the tricarboxylic acid cycle/oxidative phosphorylation (TCA/OxPhos) pathway,
generate ATP as well as other essential intermediates (Figure 1). While OxPhos is more efficient at producing ATP,
glycolysis can predominate in situations where cells are growing rapidly or have limited oxygen supply. Prominently,
cancer cells often have a higher glycolytic rate and increased lactate secretion (also known as fermentation, Figure
1) than non-tumor cells, a shift known as the Warburg effect [9,10]. Importantly however, the balance of glycolysis
and OxPhos in a cell is not a simple “either-or”; OxPhos remains essential in tumor cells and activated T cells alike
[11]. A complete picture of CCM also requires accounting for ATP turnover - that is, how rapidly ATP is hydrolyzed
to ADP by cellular processes. The most energy-intensive cellular processes include ion transport to maintain cellular
potentials, protein synthesis by ribosomes, and nucleotide synthesis to support transcription and replication. Of these,
protein and nucleotide synthesis are the most sensitive to changes in bioenergetic status [12]. As all of these processes
impact the energy charge of the cell, AMPK activity is best viewed as a balance of all the incoming and outgoing
rates affecting ATP. Therefore, the integrative nature of AMPK comes with a challenge: identifying which of these
interlinked processes are responsible for a given change observed in AMPK activity.
Here,wediscussthepotentialsourcesofcell-to-cellvariationinCCM.Wethenconsiderthetechniquesavail-
able for tracking such variation at the single-cell level, including AMPK biosensors, other metabolite biosensors,
RNA-sequencing based methods, and mass spectrometry at single-cell resolution. We end by discussing the prospects
for developing a single-cell CCM atlas across tissues.
274 ©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons
Attribution License 4.0 (CC BY-NC-ND).
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Sources of metabolic variation
To understand cell-to-cell metabolic variation, it is first necessary to consider the potential sources and mechanisms
that underlie metabolic plasticity. We describe here factors that are in principle capable of causing cell-to-cell differ-
encesinCCMwithinthesamecelltypeofagivenindividual.
Cellular differentiation status
The body is comprised of hundreds of cell types that carry out a wide array of functions. Typically, differentiation
entails the reorganization of gene expression programs through both transcription factor binding and epigenetic
changes to the chromatin. These changes enhance the production of mRNA from genes needed for a cell’s particular
function and suppress the expression of genes that could interfere with that function. Metabolic enzymes are a key
part of these programs. For example, the liver-specific transcription factor HNF4 is responsible in part for liver dif-
ferentiation and binds to the promoter for PKL (liver-specific pyruvate kinase) [13]. The expression of PKL, which
has different kinetic properties from other forms of pyruvate kinase, helps support gluconeogenesis, an important
function of the liver.
However, it should be stressed that changes in differentiation status can occur continually even within a mature
tissue and may be incomplete within certain cells - for example, tissue-resident stem cells undergoing asymmetric
division - meaning that cells within a tissue are not permanently locked into a particular gene expression profile or
metabolic behavior [14]. Moreover, metabolism can participate actively in the differentiation process. For example,
during the formation of the neural crest, cells undergo a switch to glycolysis that enhances the rate of histone lactyla-
tion, which in turn triggers genes involved in the differentiation process [15].
Cell cycle
CCM function has been shown to vary over the course of the cell cycle, and there are various points at which
metabolic enzymes and cell cycle regulators communicate. For example, the glycolytic regulator enzyme PFKFB3
can be degraded by the APC-C/Cdh1 ubiquitin ligase complex that is active from the end of mitosis through late G1
phase [16]. Other studies have highlighted additional points of co-regulation [17], which are reviewed in detail here
[18,19]. In most adult tissues, cells progress through the cycle asynchronously, and therefore at any point in time, cell
cycle-dependent variation can be expected to contribute to a metabolically heterogeneous population.
Circadian rhythm
Over a 24-hour period, many metabolic functions vary in their activity, regulated by hormonal processes or the oscil-
lations in expression of the circadian clock genes. The regulatory proteins of the circadian clock are interlinked with
transcription regulators of metabolic enzymes [20]. For example, expression of PFKFB3 is regulated by the circadian
protein CLOCK, resulting in cyclic changes in glycolytic activity [21]. These changes may allow cells to anticipate
the metabolic needs they will encounter over the course of the day. However, while these changes are in theory syn-
chronous among all cells, in reality the variation inherent in gene expression kinetics implies that some cells will lag
behind others, creating cell-to-cell variation at any given time [22].
Autonomous metabolic cycles
There is accumulating evidence that metabolic functions undergo cyclic changes independent of both the circadian
rhythm and the cell cycle [23]. Evidence for such cycles is strongest in budding yeast, where cycles on the scale of
minutes [24,25] to several hours [26–29] have been well studied. In mammalian cells, glycolytic oscillations with
periods ranging from 1 to 20 minutes have been observed in pancreatic beta cells, where they control cyclic secretion
of insulin, and in skeletal muscle cells [25]. Cycles lasting several hours have rarely been observed in mammalian
cells; however when OxPhos is inhibited in epithelial cell lines, AMPK activity undergoes a highly regular oscillation
in activity with a period of about 3 hours, which is matched by rises and falls in glycolytic intermediates [30]. The
exact mechanism of this cycle is not clear, but the 3-hour period clearly distinguishes this cycle from either the cell
cycle or the circadian rhythm, both of which have periods of 20 hours or longer. Notably, all of these cycles are most
strongly observed under specialized conditions, such as high glucose concentrations [25] or metabolic inhibition [31];
however, biosensor measurements are now revealing more subtle fluctuations with quasi-periodic behavior occurring
on similar time scales, under more physiologically realistic concentrations of glucose [31].
©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution
License 4.0 (CC BY-NC-ND).
275
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Microenvironment
The local surroundings of a cell provide a milieu of fuel sources and regulatory signals. For example, cells proximal
to the bloodstream are exposed to the nutrients being carried by the blood and circulating hormones such as in-
sulin, while also receiving an abundant supply of oxygen. Cells further distal are more buffered from feeding-induced
changes in blood chemistry and have a lower availability of oxygen. Thus, even two initially identical cells placed into
different regions of a tissue could have differences in their metabolic fluxes depending on their microenvironments.
Presumably, over time these differences in fluxes could result in gene expression changes that enable the two cells
to adapt to their different environments. Metabolic outputs, such as secretion of lactate due to increased glycolysis,
can also contribute to a continual interplay between microenvironment and cell state. Tumors represent an extreme
version of this model; within the disorganized setting of a tumor there are many microenvironments, which can be
quite different from the native habitat of the original tumor cell. These microenvironments force tumor cells to adapt
with differential metabolic behavior. A key question is which mechanisms allow tumor cells to increase their capacity
for metabolic plasticity. For example, expression of the transcription factor MYC is frequently increased and regulates
the expression of many metabolic enzymes [32], suggesting a potential mechanism for this adaptability.
Mutations
Genetic changes that alter enzyme function can drastically change how metabolism works by allowing new
metabolites to be produced. For example, the TCA enzyme isocitrate dehydrogenase (IDH), which normally
converts isocitrate to alpha-ketoglutarate, is often mutated in cancers to create a neomorphic activity that pro-
duces D-2-hydroxygluatrate (D2HG) [33]. D2HG can interfere with chromatin-modifying enzymes, for which
a-ketoglutarate normally serves as a co-factor [34]. Therefore, especially in tumor contexts, CCM function can vary
from cell to cell based on the specific mutations carried by individual cells.
Noise in gene expression
While not often considered in traditional studies of metabolic regulation, the expression of any given gene is variable
among individual cells. Studies of such variation have found that the protein copy numbers of genes vary with a
lognormal (heavy tailed) distribution with coefficients of variation between 15% and 30% [35]. Practically, this means
that for any given gene there is more than a 2-fold difference in expression level between the 10th and 90th percentiles
of a cell population [36]. Such variation, considered across all the genes of a given metabolic pathway, could lead
to substantial differences in flux. Predicting how much actual impact this variation would have is challenging, as
enzymatic reactions vary in the degree that they are controlled by the concentration of enzyme vs. substrate [37].
Nonetheless, this potent source of variation, which is known to affect processes such as cell death [38], needs to be
considered.
Altogether, there is ample opportunity for cells within the same tissue to vary in metabolic function. The sources of
variation considered here are not mutually exclusive, so they can act additively to create dispersion around the central
tendency for any one cell type. However, in most cases, the data collected have been restricted to a small number of
measurements, and there is no “atlas” of single-cell metabolic function yet. Achieving such a broad view will require
accurate and accessible single-cell measurements of metabolism, which we catalog in the next section.
Detecting and evaluating single-cell variation in metabolism
Single-cell RNA sequencing
TheabilitytoquantifythecopynumberofthevastmajorityofmRNAswithinasinglecellhasopenedupmanynew
possibilities in understanding cellular states. The major utility of this method is the identification of differentiation
states of individual cells within a tissue, which has revealed a number of previously unknown cell types [39]. Aspects
of a cell’s microenvironment and signaling activity can also be inferred, by examining gene expression patterns for
enrichment of regulons downstream of signaling pathways [40]. It also stands to reason that information about the
metabolic state can be inferred as well, and several studies have explored this possibility [3,41,42]. However, there
are inherent limitations to this approach. Most importantly, metabolic flux is regulated heavily both by substrate
concentration and by posttranslational modification of enzymes, both of which are not detectable in RNA-seq data.
Furthermore, not all changes in enzyme expression affect rate limiting steps of metabolic pathways. Finally, the often
limited correlation of mRNA and protein [43] creates difficulties for the assumption that the activity of an enzyme
can be inferred from its mRNA copy number. Thus, while single-cell gene expression profiles will continue to provide
important hints of differential metabolic regulation, such observations must be interpreted with caution and should
ideallybefollowedupwithfurtheranalysisbyothermethods.
276 ©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons
Attribution License 4.0 (CC BY-NC-ND).
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Single-cell and imaging mass spectrometry
Mass spectrometry has long been the gold standard for analysis of metabolites, as it can detect thousands of metabolite
species. Advances in sensitivity have made it possible to perform analysis on single cells. However, given the very
smallamountofmaterialavailablefromasinglecell,therearestillsignificantlimitationsonwhichmetabolitescan
be detected accurately. While there are more than 200,000 metabolites in the current human metabolome database
[44], single-cell mass spectrometry studies have detected up to several hundred metabolites [45], which are typically
predominated by lipid species [46]. Mass spectrometry imaging has also advanced, making it possible to analyze
metabolite composition of cells within tissue sections [47]. However, resolution for such imaging is still limited, such
that individual cells are not fully resolvable. Another challenge for single-cell mass spectrometry is the difficulty of
implementing flux tracing of metabolites, a key method for assessing CCM function [48].
Biosensors for AMPK and other metabolic regulators
Fluorescent protein biosensors have the distinct advantage of providing continuous monitoring of biochemical events
in single cells over time. Numerous sensors have now been developed that allow detection of the various kinase ac-
tivities, based on several different design formats [49]. These sensors detect the kinase activity of AMPK and other
key metabolic regulators including AKT, mTORC1, or PKA. Of these, AMPK biosensors have thus far revealed the
most information about metabolic plasticity in individual cells. Notably, AMPK activity has been observed to fluc-
tuate under various forms of metabolic stresses, including low glucose or low growth factors, and in the presence of
inhibitors of glycolysis or OxPhos [31]. Cell-to-cell variation in cellular metabolism was also observed in vivo using
intravital imaging of an AMPK biosensor [2]. Multiplexed analysis of AMPK, mTOR, and AKT biosensors within the
same cells demonstrated concerted fluctuations in all of these pathways, revealing the dynamics by which these reg-
ulators help to maintain cellular homeostasis [50]. Further extending these analyses, a new AMPK biosensor design
has enabled high dynamic range measurements of AMPK activity in subcellular locations [7]. This biosensor showed
morerapidactivationwhentargetedtothelysosome,wherebothAMPKandmTORarefrequentlylocalized,relative
to the cytoplasm or mitochondria. Thus, signaling biosensor analyses reveal both spatial and temporal detail in CCM
regulation, along with substantial differences between cells.
Metabolite biosensors
The engineering of fluorescent proteins that bind to key metabolites has achieved a number of elegant successes
[51]. Many of these efforts have centered on linking a metabolite-binding protein domain to a circularly-permuted
green fluorescent protein (cpGFP), such that the conformational change occurring upon metabolite binding is cou-
pled to the physical environment (and thus the fluorescence properties) of the fluorophore (see Figure 1 for a par-
tial list of such sensors). This approach has yielded biosensors for ATP, ATP/ADP, NADH/NAD+, glucose, fructose
1,6-bisphosphate (FBP), lactate, and others [52–59]. These sensors have proved useful in many systems, for exam-
ple, demonstrating variation in glycolytic activity between cells within a tumor model in vivo [59]. Certain chemical
dyes also provide useful information, such as tetra-methyl rhodamine methyl-ester (TMRM), whose fluorescence
varies with mitochondrial membrane potential. TMRM has been widely used to track mitochondrial bioenergetics,
enabling heterogeneity in CCM flux to be linked to cell death execution, for example [60]. However, a drawback to
all of these tools is that many biosensors have overlapping excitation and emission spectra, complicating their multi-
plexed usage. Furthermore, imaging the abundance of a single metabolite is often of limited utility. Interpretation of
a single readout is difficult without additional context: does an observed increase in ATP reflect an increased rate of
ATP production, a decrease in ATP consumption, or both? Finally, some metabolites are highly buffered; for example,
intracellular ATP concentration is very stable and will often not reflect even substantial perturbations to metabolism
[61,62]. Thus, gaining insight from biosensors demands clever experimental design, rather than simple monitoring
of biosensor or fluorescent dye signals alone; we highlight several such strategies below.
Biosensors with perturbations
Because they can be sampled repeatedly in the same living cell over time, biosensors make possible a class of ex-
periment that pointedly reveals functional information, which cannot be done with destructive methods. In such
experiments, biosensor readouts are used to record each cell’s reaction to a directed perturbation, such as pharma-
cological inhibition of a particular pathway or addition/withdrawal of a carbon source. In one example of this, an
AMPK reporter was monitored as cells were treated with oligomycin, an OxPhos inhibitor [30]. The rationale for this
approach is that cells that rely heavily on OxPhos for ATP production will suffer an immediate decrease in energy
charge, resulting in AMPK activation. In contrast, cells capable of meeting their ATP needs through glycolysis alone
©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution
License 4.0 (CC BY-NC-ND).
277
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
will not experience a large change in energy charge or activation of AMPK. This concept was extensively validated
in multiple cell types, suggesting that it may serve as a broadly useful tool for interrogating CCM [30]. In another
example, a lactate biosensor was used to detect the rate of increase in intracellular lactate following pharmacological
inhibition of lactate transporters [63,64]. The rate of this rise varied between cells, indicating different rates of glycol-
ysis. Still another study used an ATP biosensor targeted to the mitochondrial matrix [65]. Upon glucose withdrawal,
the mitochondrial ATP signal rose briefly and then fell precipitously. The kinetics of this change varied between cells
and were linked to different capacities for OxPhos. Therefore, as a general strategy, observing acute biosensor signal
responses following a perturbation is one of the most powerful ways to reveal cellular metabolic variation.
Protein synthesis changes in response to metabolic perturbation
An orthogonal approach to interrogating metabolism is to make use of the connection of metabolism to protein syn-
thesis. Because protein synthesis is an energy-intensive process, its overall rate is thought to be closely connected to
ATP production processes. Total cellular protein synthesis can be assayed at the single cell level, using puromycin
(or related analogs), which is incorporated into actively translating peptides, and which can be fluorescently stained
and detected by imaging or flow cytometry following cell fixation. Arg¨
uello et al demonstrated an approach termed
SCENITH (Single Cell ENergetIc metabolism by profiling Translation inHibition), in which labeling with puromycin
inthepresenceofeither2-deoxyglucoseoroligomycin,orboth,canbeusedtodistinguishcells’relianceonglycoly-
sis, respiration, or other pathways to support protein synthesis [66]. Because this labeling can be performed in living
tissues prior to their fixation, these states can be quantified under physiological conditions. However, such interpre-
tations must be approached with caution as they assume a tight correlation of CCM processes with protein synthesis.
Such a correlation has been shown at the bulk level [12], but other studies have indicated that protein synthesis and
respiration are unlinked under some conditions [67]. Moreover, there is still little evidence available to evaluate the
strength of this correlation at the single-cell level.
The future of single-cell metabolic analysis
A key major need is now a reference dataset that compares the available single-cell technologies for metabolism.
Ideally, large samples of multiple cell types, representing each of the variation sources described above, would be in-
terrogated with multiple modalities of single-cell metabolic measurements. The resulting dataset would reveal not
only how states observed in one measurement modality correspond to states in the other, but also whether the assays
vary in their applicability to different contexts. In particular, because biosensors often provide the most function-
ally interpretable data, we would argue that there is a key need to compare biosensor readouts to scRNA-seq and
puromycin-based indications. Such a dataset is quite feasible, as scRNA-seq analysis following live-cell imaging has
been demonstrated [68]; puromycin labeling in tandem with live imaging is also practicable. Performing similar ex-
periments in physiological settings will be more challenging, but is likely achievable based on elegant in vivo work
thus far [2,59]. It should be possible to combine such imaging with post-fixation analysis by spatial transcriptomics
or mass spectrometry imaging. Together, we expect these methods to provide a high level of single-cell resolution in
metabolic function, revealing subsets of cells that contribute in functionally different ways to the overall function of
tissues [7,52–55,57,58,69–73].
Summary
•Multiple technologies for single-cell metabolism measurement are emerging.
•Biosensor experiments reveal cell-to-cell variation in central carbon metabolism.
•Cell cycle, circadian rhythm, and gene expression uctuations contribute to cellular variation in
metabolism.
•Autonomous metabolic cycles drive cellular metabolic variation but are not well understood.
•There is a key need for cross-platform comparisons in single-cell metabolic methods.
278 ©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons
Attribution License 4.0 (CC BY-NC-ND).
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
Competing Interests
The authors declare that there are no competing interests associated with the manuscript.
Funding
Funding for this work was provided by the National Institute of General Medical Sciences (R35GM139621 and T32GM007377) and
the National Heart, Lung, and Blood Institute (R01HL151983).
Open Access
Open access for this article was enabled through a transformative open access agreement between Portland Press and the Uni-
versity of California.
Abbreviations
AMPK, 5’ adenosine monophosphate-activated protein kinase; CCM, central carbon metabolism; cpGFP, circularly-permuted
green uorescent protein; IDH, isocitrate dehydrogenase; OxPhos, oxidative phosphorylation; SCENITH, single cell energetic
metabolism by proling translation inhibition; TCA, tricarboxylic acid cycle; TMRM, tetra-methyl rhodamine methyl-ester.
References
1 Rangel Rivera, G.O., Knochelmann, H.M., Dwyer, C.J., Smith, A.S., Wyatt, M.M., Rivera-Reyes, A.M. et al. (2021) Fundamentals of T Cell Metabolism
and Strategies to Enhance Cancer Immunotherapy. Front Immunol. 12, 645242, https://doi.org/10.3389/fimmu.2021.645242
2 Konagaya, Y., Terai, K., Hirao, Y., Takakura, K., Imajo, M., Kamioka, Y. et al. (2017) A Highly Sensitive FRET Biosensor for AMPK Exhibits Heterogeneous
AMPK Responses among Cells and Organs. Cell Rep. 21, 2628–2638, https://doi.org/10.1016/j.celrep.2017.10.113
3 Hartmann, F.J., Mrdjen, D., McCaffrey, E., Glass, D.R., Greenwald, N.F., Bharadwaj, A. et al. (2020) Single-cell Metabolic Profiling of Human Cytotoxic T
Cells. Nat. Biotechnol. 39, 186–197, https://doi.org/10.1038/s41587-020-0651-8
4 Koveal, D., D´
ıaz-Garc´
ıa, C.M. and Yellen, G. (2020) Fluorescent Biosensors for Neuronal Metabolism and the Challenges of Quantitation. Curr. Opin.
Neurobiol. 63, 111–121, https://doi.org/10.1016/j.conb.2020.02.011
5Gonz
´
alez, A., Hall, M.N., Lin, S.-C. and Hardie, D.G. (2020) AMPK and TOR: The Yin and Yang of Cellular Nutrient Sensing and Growth Control. Cell
Metab. 31, 472–492, https://doi.org/10.1016/j.cmet.2020.01.015
6 Lin, S.-C. and Hardie, D.G. (2017) AMPK: Sensing Glucose as well as Cellular Energy Status. Cell Metab. 27, 299–313,
https://doi.org/10.1016/j.cmet.2017.10.009
7 Schmitt, D.L., Curtis, S.D., Lyons, A.C., Zhang, J.-F., Chen, M., He, C.Y. et al. (2022) Spatial regulation of AMPK signaling revealed by a sensitive kinase
activity reporter. Nat. Commun. 13, 3856, https://doi.org/10.1038/s41467-022-31190-x
8 Tsou, P., Zheng, B., Hsu, C.-H., Sasaki, A.T. and Cantley, L.C. (2011) A fluorescent reporter of AMPK activity and cellular energy stress. Cell Metab. 13,
476–486, https://doi.org/10.1016/j.cmet.2011.03.006
9 Liberti, M.V. and Locasale, J.W. (2016) The Warburg Effect: How Does it Benefit Cancer Cells? Trends Biochem. Sci. 41, 211–218,
https://doi.org/10.1016/j.tibs.2015.12.001
10 Vander Heiden, M.G., Cantley, L.C. and Thompson, C.B. (2009) Understanding the Warburg effect: the metabolic requirements of cell proliferation.
Science 324, 1029–1033, https://doi.org/10.1126/science.1160809
11 DeBerardinis, R.J. and Chandel, N.S. (2020) We need to talk about the Warburg effect. Nat. Metab. 2, 127–129,
https://doi.org/10.1038/s42255-020-0172-2
12 Buttgereit, F. and Brand, M.D. (1995) A hierarchy of ATP-consuming processes in mammalian cells. Biochem. J. 312, 163–167,
https://doi.org/10.1042/bj3120163
13 Miquerol, L., Lopez, S., Cartier, N., Tulliez, M., Raymondjean, M. and Kahn, A. (1994) Expression of the L-type pyruvate kinase gene and the hepatocyte
nuclear factor 4 transcription factor in exocrine and endocrine pancreas. J. Biol. Chem. 269, 8944–8951,
https://doi.org/10.1016/S0021-9258(17)37059-X
14 Eberwine, J. and Kim, J. (2015) Cellular Deconstruction: Finding Meaning in Individual Cell Variation. Trends Cell Biol. 25, 569–578,
https://doi.org/10.1016/j.tcb.2015.07.004
15 Merkuri, F., Rothstein, M. and Simoes-Costa, M. (2024) Histone lactylation couples cellular metabolism with developmental gene regulatory networks.
Nat. Commun. 15, 1–19, https://doi.org/10.1038/s41467-023-44121-1
16 Almeida, A., Bola ˜
nos, J.P. and Moncada, S. (2010) E3 ubiquitin ligase APC/C-Cdh1 accounts for the Warburg effect by linking glycolysis to cell
proliferation. Proc. Natl. Acad. Sci. U.S.A. 107, 738–741, https://doi.org/10.1073/pnas.0913668107
17 Takhaveev, V., ¨
Ozsezen, S., Smith, E.N., Zylstra, A., Chaillet, M.L., Chen, H. et al. (2023) Temporal segregation of biosynthetic processes is responsible
for metabolic oscillations during the budding yeast cell cycle. Nat Metab. 5, 294–313, https://doi.org/10.1038/s42255-023-00741-x
18 Zylstra, A. and Heinemann, M. (2022) Metabolic dynamics during the cell cycle. Current Opinion Systems Biol. 30, 100415,
https://doi.org/10.1016/j.coisb.2022.100415
19 Kalucka, J., Missiaen, R., Georgiadou, M., Schoors, S., Lange, C., De Bock, K. et al. (2015) Metabolic control of the cell cycle. Cell Cycle 14,
3379–3388, https://doi.org/10.1080/15384101.2015.1090068
©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons
Attribution License 4.0 (CC BY-NC-ND).
279
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
20 Guan, D. and Lazar, M.A. (2021) Interconnections between circadian clocks and metabolism. J. Clin. Invest. 131, e148278,
https://doi.org/10.1172/JCI148278
21 Chen, L., Zhao, J., Tang, Q., Li, H., Zhang, C., Yu, R. et al. (2016) PFKFB3 Control of Cancer Growth by Responding to Circadian Clock Outputs. Sci. Rep.
6, 24324, https://doi.org/10.1038/srep24324
22 Li, Y., Shan, Y., Desai, R.V., Cox, K.H., Weinberger, L.S. and Takahashi, J.S. (2020) Noise-driven cellular heterogeneity in circadian periodicity. Proc. Natl.
Acad.Sci.U.S.A.117, 10350–10356, https://doi.org/10.1073/pnas.1922388117
23 Mofatteh, M., Echegaray-Iturra, F., Alamban, A., Dalla Ricca, F., Bakshi, A. and Aydogan, M.G. (2021) Autonomous clocks that regulate organelle
biogenesis, cytoskeletal organization, and intracellular dynamics. Elife 10, e72104, https://doi.org/10.7554/eLife.72104
24 van Niekerk, D.D., van Wyk, M., Kouril, T. and Snoep, J.L. (2024) Kinetic modelling of glycolytic oscillations. Essays Biochem. 68, 15–25,
https://doi.org/10.1042/EBC20230037
25 Berridge, M.J. and Rapp, P.E. (1979) A comparative survey of the function, mechanism and control of cellular oscillators. J. Exp. Biol. 81, 217–279,
https://doi.org/10.1242/jeb.81.1.217
26 Papagiannakis, A., Niebel, B., Wit, E.C. and Heinemann, M. (2017) Autonomous Metabolic Oscillations Robustly Gate the Early and Late Cell Cycle. Mol.
Cell. 65, 285–295, https://doi.org/10.1016/j.molcel.2016.11.018
27 Tu, B.P., Mohler, R.E., Liu, J.C., Dombek, K.M., Young, E.T., Synovec, R.E. et al. (2007) Cyclic changes in metabolic state during the life of a yeast cell.
Proc.Natl.Acad.Sci.U.S.A.104, 16886–16891, https://doi.org/10.1073/pnas.0708365104
28 Tu, B.P., Kudlicki, A., Rowicka, M. and McKnight, S.L. (2005) Logic of the yeast metabolic cycle: temporal compartmentalization of cellular processes.
Science 310, 1152–1158, https://doi.org/10.1126/science.1120499
29 Slavov, N., Macinskas, J., Caudy, A. and Botstein, D. (2011) Metabolic cycling without cell division cycling in respiring yeast. Proc. Natl. Acad. Sci.
U.S.A. 108, 19090–19095, https://doi.org/10.1073/pnas.1116998108
30 Kosaisawe, N., Sparta, B., Pargett, M., Teragawa, C.K. and Albeck, J.G. (2021) Transient phases of OXPHOS inhibitor resistance reveal underlying
metabolic heterogeneity in single cells. Cell Metab. 33, 649.e8–665.e8, https://doi.org/10.1016/j.cmet.2021.01.014
31 Hung, Y.P., Teragawa, C., Kosaisawe, N., Gillies, T.E., Pargett, M., Minguet, M. et al. (2017) Akt regulation of glycolysis mediates bioenergetic stability in
epithelial cells. Elife 6, e27293, https://doi.org/10.7554/eLife.27293
32 Cargill, K.R., Stewart, C.A., Park, E.M., Ramkumar, K., Gay, C.M., Cardnell, R.J. et al. (2021) Targeting MYC-enhanced glycolysis for the treatment of
small cell lung cancer. Cancer Metab. 9, 33, https://doi.org/10.1186/s40170-021-00270-9
33 Cairns, R.A., Harris, I.S. and Mak, T.W. (2011) Regulation of cancer cell metabolism. Nat. Rev. Cancer 11, 85–95, https://doi.org/10.1038/nrc2981
34 Waitkus, M.S., Diplas, B.H. and Yan, H. (2018) Biological Role and Therapeutic Potential of IDH Mutations in Cancer. Cancer Cell. 34, 186–195,
https://doi.org/10.1016/j.ccell.2018.04.011
35 Sigal, A., Milo, R., Cohen, A., Geva-Zatorsky, N., Klein, Y., Liron, Y. et al. (2006) Variability and memory of protein levels in human cells. Nature 444,
643–646, https://doi.org/10.1038/nature05316
36 Niepel, M., Spencer, S.L. and Sorger, P.K. (2009) Non-genetic cell-to-cell variability and the consequences for pharmacology. Curr. Opin. Chem. Biol.
13, 556–561, https://doi.org/10.1016/j.cbpa.2009.09.015
37 Hackett, S.R., Zanotelli, V.R.T., Xu, W., Goya, J., Park, J.O., Perlman, D.H. et al. (2016) Systems-level analysis of mechanisms regulating yeast metabolic
flux. Science 354,https://doi.org/10.1126/science.aaf2786
38 Spencer, S.L., Gaudet, S., Albeck, J.G., Burke, J.M. and Sorger, P.K. (2009) Non-genetic origins of cell-to-cell variability in TRAIL-induced apoptosis.
Nature 459, 428–432, https://doi.org/10.1038/nature08012
39 Macosko, E.Z., Basu, A., Satija, R., Nemesh, J., Shekhar, K., Goldman, M. et al. (2015) Highly Parallel Genome-wide Expression Profiling of Individual
Cells Using Nanoliter Droplets. Cell 161, 1202–1214, https://doi.org/10.1016/j.cell.2015.05.002
40 Aibar, S., Gonz ´
alez-Blas, C.B., Moerman, T., Huynh-Thu, V.A., Imrichova, H., Hulselmans, G. et al. (2017) SCENIC: single-cell regulatory network
inference and clustering. Nat. Methods 14, 1083–1086, https://doi.org/10.1038/nmeth.4463
41 Alghamdi, N., Chang, W., Dang, P., Lu, X., Wan, C., Gampala, S. et al. (2021) A graph neural network model to estimate cell-wise metabolic flux using
single-cell RNA-seq data. Genome Res. 31, 1867–1884, https://doi.org/10.1101/gr.271205.120
42 Huang, Y., Mohanty, V., Dede, M., Tsai, K., Daher, M., Li, L. et al. (2023) Characterizing cancer metabolism from bulk and single-cell RNA-seq data
using METAFlux. Nat. Commun. 14,1–16, https://doi.org/10.1038/s41467-023-40457-w
43 Popovic, D., Koch, B., Kueblbeck, M., Ellenberg, J. and Pelkmans, L. (2018) Multivariate Control of Transcript to Protein Variability in Single Mammalian
Cells. Cell Syst. 7, 398.e6–411.e6, https://doi.org/10.1016/j.cels.2018.09.001
44 Wishart, D.S., Guo, A., Oler, E., Wang, F., Anjum, A., Peters, H. et al. (2022) HMDB 5.0: the human metabolome database for 2022. Nucleic. Acids. Res.
50, D622–D631, https://doi.org/10.1093/nar/gkab1062
45 Qin, S., Zhang, Y., Shi, M., Miao, D., Lu, J., Wen, L. et al. (2024) In-depth organic mass cytometry reveals differential contents of 3-hydroxybutanoic
acid at the single-cell level. Nat. Commun. 15, 4387, https://doi.org/10.1038/s41467-024-48865-2
46 Lanekoff, I., Sharma, V.V. and Marques, C. (2022) Single-cell metabolomics: where are we and where are we going? Curr. Opin. Biotechnol. 75,
102693, https://doi.org/10.1016/j.copbio.2022.102693
47 Taylor, M.J., Lukowski, J.K. and Anderton, C.R. (2021) Spatially Resolved Mass Spectrometry at the Single Cell: Recent Innovations in Proteomics and
Metabolomics. J. Am. Soc. Mass. Spectrom. 32, 872–894, https://doi.org/10.1021/jasms.0c00439
48 Jang, C., Chen, L. and Rabinowitz, J.D. (2018) Metabolomics and Isotope Tracing. Cell 173, 822–837, https://doi.org/10.1016/j.cell.2018.03.055
49 Schmitt, D.L., Mehta, S. and Zhang, J. (2020) Illuminating the kinome: Visualizing real-time kinase activity in biological systems using genetically
encoded fluorescent protein-based biosensors. Curr. Opin. Chem. Biol. 54, 63–69, https://doi.org/10.1016/j.cbpa.2019.11.005
50 Sparta, B., Kosaisawe, N., Pargett, M., Patankar, M., DeCuzzi, N. and Albeck, J.G. (2023) Continuous sensing of nutrients and growth factors by the
mTORC1-TFEB axis. Elife 12, e74903, https://doi.org/10.7554/eLife.74903
280 ©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons
Attribution License 4.0 (CC BY-NC-ND).
Essays in Biochemistry (2024) 68 273–281
https://doi.org/10.1042/EBC20240002
51 Liu, Y., Liu, Y. and Wang, M. (2017) Design, Optimization and Application of Small Molecule Biosensor in Metabolic Engineering. Front Microbiol. 8,
2012, https://doi.org/10.3389/fmicb.2017.02012
52 Tantama, M., Mart´
ınez-Franc¸ ois, J.R., Mongeon, R. and Yellen, G. (2013) Imaging energy status in live cells with a fluorescent biosensor of the
intracellular ATP-to-ADP ratio. Nat. Commun. 4, 2550, https://doi.org/10.1038/ncomms3550
53 Hung, Y.P., Albeck, J.G., Tantama, M. and Yellen, G. (2011) Imaging cytosolic NADH-NAD(+) redox state with a genetically encoded fluorescent
biosensor. Cell Metab. 14, 545–554, https://doi.org/10.1016/j.cmet.2011.08.012
54 Zhao, Y., Hu, Q., Cheng, F., Su, N., Wang, A., Zou, Y. et al. (2015) SoNar, a Highly Responsive NAD+/NADH Sensor, Allows High-Throughput Metabolic
Screening of Anti-tumor Agents. Cell Metab. 21, 777–789, https://doi.org/10.1016/j.cmet.2015.04.009
55 Koberstein, J.N., Stewart, M.L., Smith, C.B., Tarasov, A.I., Ashcroft, F.M., Stork, P.J.S. et al. (2022) Monitoring glycolytic dynamics in single cells using a
fluorescent biosensor for fructose 1,6-bisphosphate. Proc. Natl. Acad. Sci. U.S.A. 119, e2204407119, https://doi.org/10.1073/pnas.2204407119
56 Imamura, H., Nhat, K.P.H., Togawa, H., Saito, K., Iino, R., Kato-Yamada, Y. et al. (2009) Visualization of ATP levels inside single living cells with
fluorescence resonance energy transfer-based genetically encoded indicators. Proc. Natl. Acad. Sci. U.S.A. 106, 15651–15656,
https://doi.org/10.1073/pnas.0904764106
57 Aburto, C., Galaz, A., Bernier, A., Sandoval, P.Y., Holtheuer-Gallardo, S., Ruminot, I. et al. (2022) Single-Fluorophore Indicator to Explore Cellular and
Sub-cellular Lactate Dynamics. ACS Sens. 7, 3278–3286, https://doi.org/10.1021/acssensors.2c00731
58 Arce-Molina, R., Cort ´
es-Molina, F., Sandoval, P.Y., Galaz, A., Alegr´
ıa, K., Schirmeier, S. et al. (2020) A highly responsive pyruvate sensor reveals
pathway-regulatory role of the mitochondrial pyruvate carrier MPC. Elife 9, e53917, https://doi.org/10.7554/eLife.53917
59 Kondo, H., Ratcliffe, C.D.H., Hooper, S., Ellis, J., MacRae, J.I., Hennequart, M. et al. (2021) Single-cell resolved imaging reveals intra-tumor
heterogeneity in glycolysis, transitions between metabolic states, and their regulatory mechanisms. Cell Rep. 34, 108750,
https://doi.org/10.1016/j.celrep.2021.108750
60 Huber, H.J., Dussmann, H., Kilbride, S.M., Rehm, M. and Prehn, J.H.M. (2011) Glucose metabolism determines resistance of cancer cells to
bioenergetic crisis after cytochrome-c release. Mol. Syst. Biol. 7, 470, https://doi.org/10.1038/msb.2011.2
61 Gowans, G.J., Hawley, S.A., Ross, F.A. and Hardie, D.G. (2013) AMP is a true physiological regulator of AMP-activated protein kinase by both allosteric
activation and enhancing net phosphorylation. Cell Metab. 18, 556–566, https://doi.org/10.1016/j.cmet.2013.08.019
62 Hao, W., Chang, C.-P.B., Tsao, C.-C. and Xu, J. (2010) Oligomycin-induced bioenergetic adaptation in cancer cells with heterogeneous bioenergetic
organization. J. Biol. Chem. 285, 12647–12654, https://doi.org/10.1074/jbc.M109.084194
63 Wu, D., Harrison, D.L., Szasz, T., Yeh, C.-F., Shentu, T.-P., Meliton, A. et al. (2021) Single-cell metabolic imaging reveals a SLC2A3-dependent glycolytic
burst in motile endothelial cells. Nat Metab. 3, 714–727, https://doi.org/10.1038/s42255-021-00390-y
64 Harrison, D., Wu, D., Huang, J. and Fang, Y. (2021) Single-cell lactate production rate as a measure of glycolysis in endothelial cells. STAR Protoc. 2,
100807, https://doi.org/10.1016/j.xpro.2021.100807
65 Depaoli, M.R., Karsten, F., Madreiter-Sokolowski, C.T., Klec, C., Gottschalk, B., Bischof, H. et al. (2018) Real-Time Imaging of Mitochondrial ATP
Dynamics Reveals the Metabolic Setting of Single Cells. Cell Rep. 25, 501.e3–512.e3, https://doi.org/10.1016/j.celrep.2018.09.027
66 Arg ¨
uello, R.J., Combes, A.J., Char, R., Gigan, J.-P., Baaziz, A.I., Bousiquot, E. et al. (2020) SCENITH: A Flow Cytometry-Based Method to Functionally
Profile Energy Metabolism with Single-Cell Resolution. Cell Metab. 32, 1063.e7–1075.e7, https://doi.org/10.1016/j.cmet.2020.11.007
67 Goron, A., Lamarche, F., Blanchet, S., Delangle, P., Schlattner, U., Fontaine, E. et al. (2019) Citrulline stimulates muscle protein synthesis, by
reallocating ATP consumption to muscle protein synthesis. J. Cachexia Sarcopenia Muscle 10, 919–928, https://doi.org/10.1002/jcsm.12435
68 Lane, K., Van Valen, D., DeFelice, M.M., Macklin, D.N., Kudo, T., Jaimovich, A. et al. (2017) Measuring Signaling and RNA-Seq in the Same Cell Links
Gene Expression to Dynamic Patterns of NF-κB Activation. Cell Syst.,https://doi.org/10.1016/j.cels.2017.03.010
69 Keller, J.P., Marvin, J.S., Lacin, H., Lemon, W.C., Shea, J., Kim, S. et al. (2021) In vivo glucose imaging in multiple model organisms with an engineered
single-wavelength sensor. Cell Rep. 35, 109284, https://doi.org/10.1016/j.celrep.2021.109284
70 Marvin, J.S., Kokotos, A.C., Kumar, M. and Pulido, C. (2023) iATPSnFR2: a high dynamic range fluorescent sensor for monitoring intracellular ATP.
bioRxiv,https://doi.org/10.1101/2023.08.24.554624
71 Yoshida, T., Alfaqaan, S., Sasaoka, N. and Imamura, H. (2017) Application of FRET-based biosensor “ATeam” for visualization of ATP levels in the
mitochondrial matrix of living mammalian cells. Methods Mol. Biol. 1567, 231–243, https://doi.org/10.1007/978-1-4939-6824-4˙14
72 Ehrenberg, B., Montana, V., Wei, M.D., Wuskell, J.P. and Loew, L.M. (1988) Membrane potential can be determined in individual cells from the nernstian
distribution of cationic dyes. Biophys. J. 53, 785–794, https://doi.org/10.1016/S0006-3495(88)83158-8
73 Chen, M., Sun, T., Zhong, Y., Zhou, X. and Zhang, J. (2021) A Highly Sensitive Fluorescent Akt Biosensor Reveals Lysosome-Selective Regulation of
Lipid Second Messengers and Kinase Activity. ACS Cent Sci. 7, 2009–2020, https://doi.org/10.1021/acscentsci.1c00919
©2024 The Author(s). This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution
License 4.0 (CC BY-NC-ND).
281
