Metabolite Channeling in Energy Metabolism

Abstract

Metabolite channeling is a general mechanism to increase efficiency of sequential reactions in a metabolic pathway. It describes the transfer of intermediates between sequential enzymes without equilibration of these metabolites with the surrounding bulk solution, forming so-called metabolons or microcompartments. It is structurally based on colocalization or complex formation of the participating enzymes and on specific diffusion limitations. In energy metabolism, metabolite channeling can maintain locally high and constant levels of “high energy phosphates” such as ATP or GTP, thus increasing the efficiency of the corresponding reactions. This involves enzymes like creatine kinase, nucleoside diphosphate kinase, or glycolytic enzymes.

Full text

1
Metabolite channeling in energy metabolism
Uwe Schlattner1,2, Malgorzata Tokarska-Schlattner1, Frédéric Saudou3, and Theo
Wallimann4
1) University Grenoble Alpes, Inserm U1055, Laboratory of Fundamental and Applied
Bioenergetics, and SFR Environmental and Systems Biology, Grenoble, France
2) Institut Universitaire de France, Paris, France
3) Univ. Grenoble Alpes, Inserm U1216, CHU Grenoble Alpes, Grenoble Institut
Neurosciences, Grenoble, France
4) emeritus, formerly at the Institute of Cell Biology, Biol. Dept., ETH Zürich, Switzerland
Article Outline
1. Subcellular microcompartments and mechanisms of metabolite channeling 3
2. Advantages of metabolite channeling 4
3. Creatine kinase: channeling of high-energy phosphates for efficient ATP supply 5
3.1. The creatine kinase/phosphocreatine circuit or shuttle 5
3.2. Channeling with cytosolic CK 6
3.3. Channeling in energy transducing mitochondrial microcompartments 7
4. Nucleoside diphosphate kinase: channeling GTP 8
5. Glycolytic multienzyme complexes: roles in muscle, sperm flagella and neurons 10
Further Reading 13
Glossary 18

2
Keywords: fast axonal transport, compartmentation, creatine kinase, energy channeling,
endocytosis, glycolysis, metabolite channeling, metabolon, microcompartment, mitochondria,
mitochondrial dynamics, multienzyme complexes, nucleoside diphosphate kinase, oxidative
phosphorylation, phosphocreatine.
Abstract: Metabolite channeling is a general mechanism to increase efficiency of sequential
reactions in a metabolic pathway. It describes the transfer of intermediates between
sequential enzymes without equilibration of these metabolites with the surrounding bulk
solution, forming so-called metabolons or microcompartments. It is structurally based on
colocalization or complex formation of the participating enzymes and on specific diffusion
limitations. In energy metabolism, metabolite channeling can maintain locally high and
constant levels of “high energy phosphates” like ATP or GTP, thus increasing the efficiency
of the corresponding reactions. This involves enzymes like creatine kinase, nucleoside
diphosphate kinase, or glycolytic enzymes.

3
Subcellular microcompartments, consisting of multienzyme complexes that are embedded
within the cellular, highly viscous matrix, associated with the cytoskeleton, or situated along
membranes, are operating according to exclusion principles and favor preferred pathways of
intermediates. This process, called metabolite or substrate channeling, is defined as transfer of
intermediates between sequential enzymes without equilibration of these metabolites with the
surrounding bulk solution. Such associations between two or more sequential enzyme- or
transport reactions in a microcompartment, forming a distinct functional pool of intermediates,
are also known as functionally coupled reactions or metabolons. They can be considered as a
general mechanism to increase the efficiency of sequential reactions in a metabolic pathway
(Holthuis & Ungermann, 2013, Ovadi, 1995, Ovadi & Srere, 2000). Since metabolite
channeling leads to segregation of a metabolic pathway from other cellular reactions, it
represents a specific kind of metabolic compartmentation similar to that operating within
membrane-separated organelles or by being restricted two-dimensional diffusion at surface
boundary layers. Here, metabolite channeling is described with examples from energy
metabolism, where local fueling of ATPases and GTPases is maintained by isoenzymes of
creatine kinase (CK), nucleoside diphosphate kinase (NDPK) or glycolytic enzymes.
1. Subcellular microcompartments and mechanisms of
metabolite channeling
Life most likely originated autotrophically de novo in metabolic complexes organized on
FeS2 (pyrite) mineral surfaces, the earliest form of microcompartments. Therefore, a cell
cannot be represented by a well-mixed bag of enzymes, behaving in complete equilibrium
according to solution kinetics. Because of the intricate structural and functional organization
of living cells, enzymes and metabolites do not behave as if they were freely diffusible in
solution. Instead, they may form structurally, functionally, and temporally defined subcellular
microcompartments, either via strong static, or via fickle, dynamic interactions with other
enzymes, proteins, or subcellular structures. Such a structural organization of pathway
components is a general prerequisite for metabolite channeling . It may involve (1) huge
covalently linked enzyme-complexes (or multifunctional enzymes) such as fatty acid
synthase (FAS), (2) kinetically stable multienzyme complexes like pyruvate dehydrogenase
(PDH) or bacterial and plant tryptophan synthase (TS), (3) more dynamic, reversibly

4
associating enzymes such as glycolytic complexes containing glyceraldehyde phosphate
dehydrogenase (GAPDH) or glycerol phosphate dehydrogenase (GPDH), or (4)
colocalization on subcellular particles or biological membranes. These associations allow the
transfer of intermediates between the channeling components by different mechanisms: (1)
physical hindrance or electrostatic effects prevent mixing with bulk solution and drive a
directed diffusion (e.g., TS, FAS), (2) sequential covalent binding to very close active sites in
the reaction sequence (e.g., PDH), (3) transfer of noncovalently bound intermediates between
active sites (e.g., NADH dehydrogenase), (4) transfer in dynamic multi-enzyme complexes
(GAPDH, GPDH) or in an unstirred membrane surface layer (mtCK). These mechanisms can
be fulfilled in both, static and dynamic enzyme associations. However, while static
associations often allow for an almost perfect or “tight” channeling of metabolites, dynamic
channeling is often only partial or “leaky” (Figure 1).
It remains technically challenging to access metabolite channeling experimentally, and there
are continuing efforts to develop appropriate methodologies (Abernathy et al., 2019, Zhang et
al., 2017). There is also rising interest in metabolite channeling for applications in metabolic
engineering and synthetic biology (Abernathy et al., 2017, Castellana et al., 2014, Obata,
2020).
2. Advantages of metabolite channeling
Sequestering of intermediates in a microcompartment through metabolite channeling
provides kinetic and regulatory advantages for not only the reaction sequence itself (Figure
1), but also for cellular metabolism (Ovadi, 1995, Zala et al., 2017). In general: (1) enzyme
reaction rates and equilibria are controlled by local and enzyme bound substrates, rather than
bulk phase substrate concentrations, (2) for a readily reversible reaction, local supply of
substrate and removal of product may drive the reaction in the desired direction, (3)
sequestered intermediates are present at high local concentrations and an apparently low Km
for these intermediates can be observed with the channeling complex compared to the non-
channeling situation measured with isolated components, (4) metabolites are isolated from
competing reactions, e.g., between anabolic and catabolic pathways, (5) the life-time of the
intermediate in the solvent phase is shortened relative to free diffusion, which may be
essential in case of unstable intermediates, (6) in certain cases, the unfavorable energetics of

5
desolvating the substrate that precedes binding to the enzyme is avoided, (7) channeling
components can be regulated by modulators that affect enzyme associations, and (8) a larger
degree of metabolic control of the over-all-flux of the reactions can be achieved, e.g., via
feed-back regulatory mechanisms such as substrate activation, product inhibition, and
cooperativity.
3. Creatine kinase: channeling of high-energy phosphates for
efficient ATP supply
3.1. The creatine kinase/phosphocreatine circuit or shuttle
One fundamental requirement of life is energy supply. Cellular energy demand and supply
are balanced, and tightly regulated for economy and efficiency of energy use. In eukaryotic
cells, CK is a major enzyme that helps to cope with high and fluctuating energy demands to
maintain cellular energy homeostasis in general and to guarantee stable, locally buffered
ATP/ADP ratios in particular (Bessman & Carpenter, 1985).
The enzyme catalyzes the reversible phosphoryl transfer from ATP to creatine (Cr) to
generate ADP and phosphocreatine (PCr). Thus, CK is able to conserve energy in the form of
metabolically inert PCr and vice versa, to use PCr to replenish global as well as local cellular
ATP pools. Since PCr can accumulate to much higher cellular concentrations than ATP, the
CK/PCr-system constitutes an efficient and immediately available cellular “energy buffer.” In
addition, tissue-specific CK isoenzymes are located in the cytosol (dimeric muscle-type MM-
CK and brain-type BB-CK) and within the mitochondrial intermembrane space (sarcomeric
smtCK and ubiquitous umtCK, both forming octamers and dimers). CK isoenzymes are often
associated with sites of ATP supply, where they generate PCr, or with sites of ATP
consumption, where they regenerate ATP by using PCr. Thus, together with the faster
diffusion rate of PCr as compared to ATP, the CK/PCr system also supports an intrinsic
energy transfer system (CK/PCr-circuit or -shuttle), coupling sites of energy generation
(oxidative phosphorylation or glycolysis) with sites of energy consumption (Wallimann et al.,
1992) (Figure 2A). This circuit is particularly important in cells with high or fluctuating
energy demands like cardiac and skeletal muscle or brain cells (Guzun et al., 2015), as well
as in large and/or polar cells, such as spermatozoa. In the latter, diffusional limitations of

6
adenine nucleotides, especially ADP, along the sperm tail become especially apparent.
(Wallimann et al. 2011).
3.2. Channeling with cytosolic CK
Cytosolic CK is only partially soluble. A significant fraction is structurally and functionally
associated or co-localized with different, structurally bound ATPases or ATP-regulated
processes (Schlattner et al., 2016). These include different ion pumps or the ATP-gated K+-
channel at the plasma membrane, the sarcomeric M-band of the myofibrils in muscle (Figure
2A, #4), or the calcium pump of the muscular sarcoplasmic reticulum. Cytosolic CK is also
structurally associated with the key regulatory enzyme of glycolysis, phosphofructokinase
(PFK), which itself is regulated by ATP. In all these cases, PCr is used for the local
regeneration of ATP, which is directly channeled from CK to the ATP-consuming ATPase or
ATP-regulated enzyme without major dilution by the surrounding bulk solution.
Only in some cases, the physical basis for the metabolite channeling is known. For example,
MM-CK uses a “charge clamp” consisting of four lysine residues to specifically bind to
partner molecules, myomesin and M-protein in the M-band. This allows for an isoenzyme-
specific association of MM-CK with the sarcomeric M-band of the myofibrillar apparatus,
where CK is ideally positioned to regenerate in situ during muscle contraction the ATP
hydrolyzed in the acto-myosin overlap zones that are situated symmetrically on both sides of
the M-band. (Hornemann et al. 2000)
A tight functional coupling of CK to ATPases, e.g., acto-myosin ATPase and ion pumps,
such as the K+/Na+-ATPase or the Ca2+-ATPase, has the advantage (1) that product inhibition
of the ATPase by ADP and H+ is avoided, since the latter are both substrates of the CK
reaction (PCr+ADP+H+↔Cr+ATP) and (2) that the high free energy of ATP hydrolysis
(ΔGATP) at sites of ATP hydrolysis is preserved by keeping locally very high ATP/ADP ratios
due to coupling of CK with those ATPases in situ and thus preventing energy dissipation that
would otherwise be caused by transport of ATP and mixing it with the bulk surrounding
(Wallimann et al., 1992; 2011). Transgenic mice no longer expressing cytosolic MM-CK and
mitochondrial mtCK in muscle have significant difficulties with intracellular calcium
handling and muscle relaxation. This emphasizes the physiological importance of the CK
system for the energetics of intracellular calcium homeostasis in general and the delivery of
ATP to the energetically demanding Ca2+-ATPase, in particular.

7
3.3. Channeling in energy transducing mitochondrial microcompartments
The mtCK forms mainly large, cuboidal octamers that are present (1) between the outer and
inner mitochondrial membrane (the so-called “intermembrane space” of mitochondria),
preferentially localizing in “mitochondrial contact sites” between these two membranes
(Figure 2A, #2), as well as (2) in the cristae space (Figure 2A, #1) (Schlattner et al., 2006,
Schlattner et al., 2018). This CK isoform catalyzes the direct transphosphorylation of
intramitochondrial produced ATP and Cr from the cytosol into ADP and PCr. ADP then
enters the matrix space to stimulate oxidative phosphorylation, giving rise to mitochondrial
recycling of a specific pool of ATP and ADP, while PCr is the primary “high energy”
phosphoryl compound that leaves mitochondria into the cytosol. The molecular basis for such
directed metabolite flux is channeling between the large, cuboidal mtCK octamer and two
transmembrane proteins, adenylate translocator (ANT) and mitochondrial porin or voltage-
dependent anion channel (VDAC). ANT is an obligatory antiporter for ATP/ADP exchange
across the inner mitochondrial membrane, while VDAC is a nonspecific, potential-dependent
pore in the outer mitochondrial membrane. The mtCK-linked metabolite channeling is based
on (1) colocalization, (2) direct interactions, and (3) diffusion barriers. mtCK tightly binds to
cardiolipin, an acidic phospholipid that is specific for the mitochondrial inner membrane.
Since ANT molecules are situated in a cardiolipin patches, this leads to colocalization and
metabolite channeling between both proteins, mtCK and ANT, in the cristae as well as in the
intermembrane space. Octameric mtCK in the intermembrane space further interacts with
outer membrane phospholipids and VDAC, thus virtually cross-linking inner and outer
membrane and contributing to the “mitochondrial contact sites.” Increasing the external
calcium concentrations strengthens the interaction of mtCK with VDAC, which may improve
channeling under cytosolic calcium overload as occurring at low cellular energy state.
Finally, the limited and potentially regulated permeability of VDAC and thus of the entire
outer membrane creates a dynamic microcompartmentation of metabolites in the
intermembrane space that contributes to mtCK-linked channeling and separate mitochondrial
ATP- and ADP pools. Similar to mtCK, hexokinase is able to use intramitochondrially
produced ATP by binding to VDAC from the cytosolic mitochondrial surface at contact sites
containing only ANT and VDAC. The direct functional coupling of mtCK to oxidative
phosphorylation can be demonstrated with oxygraph respirometry on skinned muscle fibers
from normal and transgenic mice lacking mtCK. While in normal muscle fibers Cr stimulates
mitochondrial respiration, this phenomenon is missing in fibers of mtCK knock-out mice.

8
By favoring ATP/ADP exchange through ANT in the mitochondrial inner membrane,
cytosolic Cr and the mtCK reaction not only stimulate the rate of mitochondrial respiration,
but also lead to improved coupling between respiration, ATP synthesis and mitochondrial
energy channeling, which finally reduces formation of potentially deleterious reactive oxygen
species (ROS; Figure 2A, #1 lower part). Both effects of mtCK-related metabolite
channeling, stimulation of mitochondrial respiration and lowering of ROS production, may
contribute to the remarkable cell- and specifically neuro-protective action of Cr
supplementation in vitro and in vivo. This is an instructive example of how efficient, multiple
metabolite channeling events can be beneficial for human health.
Cellular low-energy stress induces expression umtCK, be it caused by chronic endurance
training, fasting, Cr depletion, or pathologies in ATP generation like mitochondrial
dysfunction seen in patients with mitochondrial cytopathies. In the latter, highly up-regulated
mtCK can even crystallize into sheets to form characteristic “railway-track inclusions” within
enlarged mitochondria as a hallmark of the pathology. Similar mtCK inclusions were
observed with chronic Cr depletion in rodents. On the other hand, cellular stress that
generates reactive oxygen and nitrogen species like cardiac infarction or anthracycline
chemotherapy leads to molecular damage of mtCK. This then triggers inactivation and
dimerization of mtCK, as well as its dissociation from the mitochondrial inner membrane. All
of these impair an efficient channeling of high-energy phosphates by mtCK and thus
contribute to cardiac energy failure or specific anthracycline cardiotoxicity. Finally, mtCK
together with its substrate Cr participates in regulating the mitochondrial permeability
transition pore that is crucially involved in triggering apoptosis. Likely because of metabolite
channeling in the mtCK/ANT microcompartment, a high ADP concentrations is maintained
in the matrix space that is inhibitory for permeability transition. Thus, the channeling CK-
system may exert additional effects that are not necessarily directly related to improved cell
energetics. This may explain some of the cell- and neuro-protective effects seen with Cr
supplementation.
4. Nucleoside diphosphate kinase: channeling GTP
Many cellular processes require GTP instead of ATP as an energy source, including
biosynthesis and membrane translocation of proteins, cellular membrane dynamics or cell

9
signaling. The involved GTPases have either rather low turnover, like signaling
heterotrimeric G proteins and small GTPases, or high turnover like the motor proteins of the
dynamin family. Cellular GTP is largely supplied by members of the NME/NDPK/NM23
protein family that have nucleoside diphosphate kinase (NDPK) activity, namely isoforms
NME1 to NME4 (Boissan et al., 2018). These hexameric proteins use mainly cellular ATP to
regenerate other nucleoside triphosphates (NTPs), including GTP, in the reaction
ATP+NDP↔ADP+NTP. However, while ATP- and GTP-binding proteins have often similar
affinities for their respective nucleotide, cellular concentrations of guanine nucleotides are
largely below those of adenine nucleotides. To overcome this mismatch, NMEs localizing in
close vicinity of GTP-dependent processes locally increase GTP concentrations, or even
directly interact with and channel GTP to the guanylate binding site of specific GTPases
(Figure 2B).
The cytosol harbors three NME isoforms, NME1-NME3, that can form homo- and
heterohexamers. These are partially soluble, but partially also associate with membranes, in
particular NME3 that has a N-terminal hydrophobic membrane anchor. At the plasma
membrane, NME2/NME3 heterohexamers recruit heterotrimeric G proteins to locally provide
the GTP necessary for G protein activation via GTP/GDP exchange (Hippe et al., 2009). In
addition, NME2 can directly transfer “high energy phosphates” from its active site histidine
to a G protein histidine close to the guanylate binding site, from where it is used to regenerate
GTP from GDP directly on-site (Abu-Taha et al., 2018). This histidine phospho-relay
mechanism allows G protein activation independent of the classical receptor-induced
activation of heterotrimeric G proteins. Another example of GTP channeling is the
membrane-remodeling GTPase dynamin which drives membrane fission during clathrin-
mediated endocytosis. Here, NME1/2 localize to clathrin-coated pits at the plasma membrane
where they directly interact with the proline-rich domain of dynamin to channel the GTP
necessary for dynamin-mediated membrane constriction and efficient endocytosis (Boissan et
al., 2014) (Figure 2B, #6).
Two NME isoforms localize to mitochondria. NME4 is imported into the intermembrane and
matrix spaces of mitochondria, where it binds to the mitochondrial inner membrane (MIM)
via interaction with the phospholipid cardiolipin (Schlattner et al., 2013). NME3 anchors at
the outer surface of the mitochondrial outer membrane (MOM) via its N-terminal
hydrophobic stretch. In their respective membranes, both NMEs directly interact with and

10
channel GTP to specific dynamin-related GTPases that drive membrane remodeling and
mitochondrial fusion (Figure 2B, #5). While NME4 fuels OPA1 at MIM (Boissan et al.,
2014), NME3 fuels MFN1/2 at MOM (Chen et al., 2019). Although all the mentioned
GTPases of the dynamin family can also work in the absence of NMEs, their full efficiency is
only achieved in association with NMEs that provide high and constant local GTP levels.
5. Glycolytic multienzyme complexes: roles in muscle, sperm
flagella and neurons
In muscle, glycolytic enzymes are targeted to the actin-containing thin filaments at the
sarcomeric I-band region where they form highly complex metabolons. The I-band in
Drosophila flight muscle contains a multienzyme complex consisting of GDPH-1, aldolase,
and GAPDH. By elegant experiments with transgenic Drosophila expressing GDPH-3
instead of GDPH-1, it could be shown that all three glycolytic enzymes no longer colocalize
in the I-band to form a microcompartment. Even though the full complement of active
glycolytic enzymes was still present, their failure to colocalize in the sarcomer resulted in the
inability to fly (Wojtas et al., 1997). Thus, correct targeting and formation of multienzyme
complexes that lead to functionally coupled microcompartments and substrate/product
channeling seem to be a prerequisite for proper function of glycolysis and ultimately for
correct muscle function. In mammalian cells, also CK is participating in the glycolytic
metabolon.
Since glycolysis is able to produce ATP at low yield but high rate as compared to oxidative
phosphorylation that produce ATP with high yield but lower rate, glycolysis is often used for
activities requiring sudden high levels of ATP. An example illustrating such high energy
demand is the motility of many sperm flagella that is powered by glycolysis. Notably, the
glycolytic enzymes are localized within the flagellum and more precisely arrayed on the
fibrous sheath, the cytoskeletal structure separating the head from the sperm tail. The sperm
glycolytic enzymes differ from their somatic counterparts by containing an anchoring domain
to the fibrous sheath of the flagellum. This property of the sperm glycolytic enzyme was used
to tether ten glycolytic enzymes, from hexokinase (HK) to pyruvate kinase (PK) and lactate
dehydrogenase (LDH), to nanoparticles thus reconstituting in vitro the conversion of glucose
to lactate. While the enzymatic efficiency of each enzyme alone was higher when in solution,

11
the conversion of glucose to lactate was significantly higher when the ten enzymes were
tethered to nanoparticles as compared to their soluble counterparts (Mukai et al., 2017).
These findings illustrate how substrate/product channeling is improved by the ordering of
sequential reaction in functionally coupled microcompartments.
In the brain, energy demands are primarily met by glucose oxidation through glycolysis and
oxidative phosphorylation. These ATP-generating pathways have to respond in a rapid and
localized manner to large changes in energy demand. This high demand in energy is
particularly evident at synapses since vesicles containing neurotransmitters and factors such
as the brain-derived neurotrophic factor (BDNF), are released to allow neurotransmission
upon neuronal activity. Mitochondria have been shown to be essential for sustaining synapse
function by maintaining energy metabolism at the presynapse as well as by clearing calcium.
However, glycolysis by producing ATP at high rate likely allows synapses to adapt to the
sudden needs in energy. This was demonstrated for Caenorhabditis elegans. Under
conditions of energy stress, glycolytic enzymes dynamically relocalize at synapses to form a
glycolytic metabolon thus maintaining high local levels of ATP at the synapse, the proper
functioning of the synaptic vesicle cycle and subsequent locomotion of the worms (Jang et
al., 2016). In addition to this dynamic redistribution of the glycolytic metabolon upon high
energy demand (e.g.: neuronal activity) or energy depletion, the glycolytic enzymes GAPDH
and PGK have been shown to associate to synaptic vesicles and the glycolytic machinery is
required to activate the vesicular proton pump required for vesicle glutamate reuptake at the
synapse further illustrating the compartmentalization of glycolytic pathway in neurons.
While such compartmentalization of the glycolytic metabolon appear to fulfill energy
demand at specific subcellular localizations, one long-standing question has been how motile
organelles such as vesicles or mitochondria move within neurons and what is the source of
energy for this transport. Indeed, neurons are highly polarized cells with axons that can
measure up to one meter. Neurons require efficient intracellular transport referred as fast
axonal transport (FAT) to ascertain exchange of material, mainly organelles, between the cell
body and the synapses. Dysregulated FAT plays crucial roles in different neurodegenerative
disorders (Hinckelmann et al., 2013). FAT relies on ATP-dependent motor proteins, namely
kinesin and dynein for anterograde and retrograde transport, respectively, moving cargo like
vesicles or mitochondria along microtubules. Numerous adaptor or regulatory proteins are
associated with the FAT machinery thereby modulating transport efficacy, selectivity for the

12
cargo and directionality within axons and dendrites. The observation of the specific
association and function of the glycolytic enzymes on vesicles led to the demonstration that
these enzymes provide energy for the long distance transport of vesicles within axons
(Hinckelmann et al., 2016, Zala et al., 2013). While mitochondria generate their own energy
for their locomotion, small vesicles contain the whole glycolytic metabolon. Vesicles are able
to produce energy from glucose and to self-propel on microtubules independently of
mitochondrial or bulk ATP in the cell (Figure 2C, #7). Thus, these vesicles have their energy
supply machinery “on board”.
Together, these different examples illustrate the importance of the glycolytic metabolon to
sustain a wide range of energy-consuming activities within the cell. The extraordinary high
rate activity of the glycolytic machinery makes this metabolon essential for the cells to adapt
to sudden changes in the environment.

13
Cross References
Free Radicals, Sources and Targets of: Mitochondria
Mitochondrial Channels
Mitochondrial Membranes, Structural Organization
Mitochondrial Metabolite Transporter Family
Mitochondrial Outer Membrane and the VDAC Channel
P-Type Pumps: Na+/K+ Pump

14
Cited references and further reading
Abernathy, M., Wan, N., Shui, W. and Tang, Y. J. (2019). Dynamic 13C labeling of fast
turnover metabolites for analysis of metabolic fluxes and metabolite channeling. In Baidoo,
E. E. K. (ed) Microbial Metabolomics. Vol. 1859, pp. 301-316 New York: Humana Press,
Springer.
Abernathy, M. H., He, L. and Tang, Y. J. (2017). Channeling in native microbial pathways:
Implications and challenges for metabolic engineering. Biotechnol Adv 35, 805-814.
Abu-Taha, I. H., Heijman, J., Feng, Y., Vettel, C., Dobrev, D. and Wieland, T. (2018).
Regulation of heterotrimeric G-protein signaling by NDPK/NME proteins and caveolins: an
update. Lab Invest 98, 190-197.
Bessman, S. P. and Carpenter, C. L. (1985). The creatine-creatine phosphate energy shuttle.
Annu Rev Biochem 54, 831-862.
Boissan, M., Montagnac, G., Shen, Q., Griparic, L., Guitton, J., Romao, M., Sauvonnet, N.,
Lagache, T., Lascu, I., Raposo, G., Desbourdes, C., Schlattner, U., Lacombe, M. L., Polo, S.,
van der Bliek, A. M., Roux, A. and Chavrier, P. (2014). Membrane trafficking. Nucleoside
diphosphate kinases fuel dynamin superfamily proteins with GTP for membrane remodeling.
Science 344, 1510-1515.
Boissan, M., Schlattner, U. and Lacombe, M. L. (2018). The NDPK/NME superfamily: state
of the art. Lab Invest 98, 164-174.
Castellana, M., Wilson, M. Z., Xu, Y., Joshi, P., Cristea, I. M., Rabinowitz, J. D., Gitai, Z.
and Wingreen, N. S. (2014). Enzyme clustering accelerates processing of intermediates
through metabolic channeling. Nat Biotechnol 32, 1011-1018.
Chen, C. W., Wang, H. L., Huang, C. W., Huang, C. Y., Lim, W. K., Tu, I. C., Koorapati, A.,
Hsieh, S. T., Kan, H. W., Tzeng, S. R., Liao, J. C., Chong, W. M., Naroditzky, I., Kidron, D.,
Eran, A., Nijim, Y., Sela, E., Feldman, H. B., Kalfon, L., Raveh-Barak, H., Falik-Zaccai, T.
C., Elpeleg, O., Mandel, H. and Chang, Z. F. (2019). Two separate functions of NME3

15
critical for cell survival underlie a neurodegenerative disorder. Proc Natl Acad Sci U S A 116,
566-574.
Guzun, R., Kaambre, T., Bagur, R., Grichine, A., Usson, Y., Varikmaa, M., Anmann, T.,
Tepp, K., Timohhina, N., Shevchuk, I., Chekulayev, V., Boucher, F., Dos Santos, P.,
Schlattner, U., Wallimann, T., Kuznetsov, A. V., Dzeja, P., Aliev, M. and Saks, V. (2015).
Modular organization of cardiac energy metabolism: energy conversion, transfer and
feedback regulation. Acta Physiol (Oxf) 213, 84-106.
Hinckelmann, M. V., Virlogeux, A., Niehage, C., Poujol, C., Choquet, D., Hoflack, B., Zala,
D. and Saudou, F. (2016). Self-propelling vesicles define glycolysis as the minimal energy
machinery for neuronal transport. Nat Commun 7, 13233.
Hinckelmann, M. V., Zala, D. and Saudou, F. (2013). Releasing the brake: restoring fast
axonal transport in neurodegenerative disorders. Trends Cell Biol 23, 634-643.
Hippe, H. J., Wolf, N. M., Abu-Taha, I., Mehringer, R., Just, S., Lutz, S., Niroomand, F.,
Postel, E. H., Katus, H. A., Rottbauer, W. and Wieland, T. (2009). The interaction of
nucleoside diphosphate kinase B with Gbetagamma dimers controls heterotrimeric G protein
function. Proc Natl Acad Sci U S A 106, 16269-16274.
Holthuis, J. C. and Ungermann, C. (2013). Cellular microcompartments constitute general
suborganellar functional units in cells. Biol Chem 394, 151-161.
Hornemann T, Stolz M, Wallimann T.
Isoenzyme-specific interaction of muscle-type creatine kinase with the sarcomeric M-line is
mediated by NH(2)-terminal lysine charge-clamps.
J Cell Biol. 2000 Jun 12;149(6):1225-34. doi: 10.1083/jcb.149.6.1225.
Jang, S., Nelson, J. C., Bend, E. G., Rodriguez-Laureano, L., Tueros, F. G., Cartagenova, L.,
Underwood, K., Jorgensen, E. M. and Colon-Ramos, D. A. (2016). Glycolytic Enzymes
Localize to Synapses under Energy Stress to Support Synaptic Function. Neuron 90, 278-291.
Mukai, C., Gao, L., Nelson, J. L., Lata, J. P., Cohen, R., Wu, L., Hinchman, M. M.,
Bergkvist, M., Sherwood, R. W., Zhang, S. and Travis, A. J. (2017). Biomimicry promotes
the efficiency of a 10-Step sequential enzymatic reaction on nanoparticles, converting
glucose to lactate. Angew Chem Int Ed Engl 56, 235-238.

16
Obata, T. (2020). Toward an evaluation of metabolite channeling in vivo. Curr Opin
Biotechnol 64, 55-61.
Ovadi, J. (1995). Cell architecture and metabolic channeling. Heidelberg: Springer.
Ovadi, J. and Srere, P. A. (2000). Macromolecular compartmentation and channeling. Int Rev
Cytol 192, 255-280.
Schlattner, U., Kay, L. and Tokarska-Schlattner, M. (2018). Mitochondrial proteolipid
complexes of creatine kinase. Subcell Biochem 87, 365-408.
Schlattner, U., Klaus, A., Ramirez Rios, S., Guzun, R., Kay, L. and Tokarska-Schlattner, M.
(2016). Cellular compartmentation of energy metabolism: creatine kinase
microcompartments and recruitment of B-type creatine kinase to specific subcellular sites.
Amino Acids 48, 1751-1774.
Schlattner, U., Tokarska-Schlattner, M., Ramirez, S., Tyurina, Y. Y., Amoscato, A. A.,
Mohammadyani, D., Huang, Z., Jiang, J., Yanamala, N., Seffouh, A., Boissan, M., Epand, R.
F., Epand, R. M., Klein-Seetharaman, J., Lacombe, M. L. and Kagan, V. E. (2013). Dual
function of mitochondrial Nm23-H4 protein in phosphotransfer and intermembrane lipid
transfer: a cardiolipin-dependent switch. J Biol Chem 288, 111-121.
Schlattner, U., Tokarska-Schlattner, M. and Wallimann, T. (2006). Mitochondrial creatine
kinase in human health and disease. Biochim Biophys Acta 1762, 164-180.
Wallimann, T., Wyss, M., Brdiczka, D., Nicolay, K. and Eppenberger, H. M. (1992).
Intracellular compartmentation, structure and function of creatine kinase isoenzymes in
tissues with high and fluctuating energy demands: the 'phosphocreatine circuit' for cellular
energy homeostasis. Biochem J 281 ( Pt 1), 21-40.
Wallimann T, Tokarska-Schlattner M, Schlattner U.
The creatine kinase system and pleiotropic effects of creatine.
Amino Acids. 2011 May;40(5):1271-96. doi: 10.1007/s00726-011-0877-3
Wojtas, K., Slepecky, N., von Kalm, L. and Sullivan, D. (1997). Flight muscle function in
Drosophila requires colocalization of glycolytic enzymes. Mol Biol Cell 8, 1665-1675.

17
Zala, D., Hinckelmann, M. V., Yu, H., Lyra da Cunha, M. M., Liot, G., Cordelieres, F. P.,
Marco, S. and Saudou, F. (2013). Vesicular glycolysis provides on-board energy for fast
axonal transport. Cell 152, 479-491.
Zala, D., Schlattner, U., Desvignes, T., Bobe, J., Roux, A., Chavrier, P. and Boissan, M.
(2017). The advantage of channeling nucleotides for very processive functions. F1000Res 6,
724.
Zhang, Y., Beard, K. F. M., Swart, C., Bergmann, S., Krahnert, I., Nikoloski, Z., Graf, A.,
Ratcliffe, R. G., Sweetlove, L. J., Fernie, A. R. and Obata, T. (2017). Protein-protein
interactions and metabolite channelling in the plant tricarboxylic acid cycle. Nat Commun 8,
15212.

18
Glossary
metabolite channeling
Local transfer of metabolic intermediates between sequential enzyme or transport
reactions without equilibration with bulk solution.
metabolic compartmentation
Segregation of intermediates and enzymes of a metabolic pathway by membranes,
binding to a specific surface or direct interaction in protein complexes allowing
metabolite channeling.
microcompartment
Structural unit allowing metabolic compartmentation, also called “metabolon.”
mitochondrial contact sites
Close adhesions of inner and outer mitochondrial membrane that can be observed by
electron microscopy and can be isolated as a separate microcompartment. Contact
sites consist of multi-lipid/protein complexes with variable composition and are
involved in energy transduction (e.g., containing ANT/VDAC or
ANT/mtCK/VDAC) or protein import.
mitochondrial permeability transition pore
A multiprotein complex of the inner mitochondrial membrane that is crucially
involved in early events that trigger apoptosis like release of cytochrome c and other
apoptosis-inducing factors into the cytosol.

19
Uwe Schlattner is Professor and Director of the Laboratory of Fundamental and
Applied Bioenergetics, Inserm U1055, at the University Grenoble Alpes, France, and
senior member of the Institut Universitaire de France. He studied biology in
Germany and Switzerland, and worked from 1996 to 2005 at ETH Zürich,
Switzerland. His research interest is in molecular mechanisms of cellular energy
homeostasis, their spatial and temporal integration within the cell, and their role in
human health and disease. Web: https://lbfa.univ-grenoble-alpes.fr/research/axis-1-
energy-signaling-systems-bioenergetics
Malgorzata Tokarska-Schlattner is senior scientist at the Laboratory of Fundamental
and Applied Bioenergetics, Inserm U1055, at the Grenoble Alpes, France. She
studied biophysics and philosophy in Poland, and continued postdoctoral work at the
University of Geneva and at ETH Zurich, Switzerland. Her interests are in membrane
biophysics and cellular energy stress.
Frédéric Saudou is Professor and hospital practitioner at the University Grenoble
Alpes & University Hospital Grenoble Alpes. He is also director of the Grenoble
Institute of Neurosciences, Inserm U1216. He studied molecular neurobiology in
Strasbourg and after a postdoc at Harvard Med. School, Boston, he developed his
group at the Institut Curie in Paris from 2000 to 2014 before moving to Grenoble.
His main interests are Huntington’s disease, axonal transport and energy metabolism.
Theo Wallimann is a Professor emeritus, formerly working at the Institute of Cell
Biology at the ETH in Zürich, Switzerland, where he also received his Ph.D. After a
postdoctoral stay at Brandeis University, Boston, U.S.A. from 1975-1981, he
returned to ETH Zürich. His research interests are in cellular energetics, especially
creatine kinases and AMP-activated protein kinases, as well as creatine
supplementation in health and disease. Web: https://mhs.biol.ethz.ch/about-
us/emeriti-formermembers/wallimann.html

20
Figure 1
Figure 1. Free diffusion (no channeling) versus loose (leaky) or tight metabolite channeling.
Compartmentation of a reaction sequence without (full) equilibration with bulk solution leads
to shorter transition times and further advantages (see text).

21
Figure 2
Figure 2. High energy phosphate channeling in specific cellular microcompartments. Colored
symbols depict proteins that transport energy phosphates (yellow, orange) or convert high
energy phosphates (ATP generation, blue; ATP/PCr conversion, red; ATP/GTP conversion,
magenta; ATP or GTP consumption, green). Metabolite fluxes are depicted as arrows.
(A) Creatine kinase (CK) isoforms, together with easily diffusible, highly concentrated
phosphocreatine (PCr; up to 30 mM), maintain a unique energy buffer and shuttle system
between ATP-providing and -consuming processes, in particular in cells that are polarized
and/or have very high or localized ATP consumption. Isoenzymes of CK are found in
mitochondria (octameric mtCK 12) and the cytosol (dimeric cytCK 34), either soluble (not
shown) or structurally associated with ATP-delivering or -consuming sites. Metabolite
channeling occurs where CK is associated with ATP-providing or -consuming transporters,
pumps or metabolic enzymes. 12 Mitochondrially generated ATP is exported from the matrix
through the mitochondrial inner membrane (MIM) via adenine nucleotide translocase (ANT)
and then channeled to mtCK, both proteins being in close proximity via their preferential
interaction with the mitochondrial phospholipid cardiolipin. 1 In cristae, ANT and mtCK can
be part of supercomplexes that also include F0F1-ATPase and respiratory chain complexes.

22
This maintains a local, efficient ADP/ATP circuit favoring correct respiratory function and
reducing reactive oxygen species (ROS). Generated PCr then diffuses along cristae and cristae
junctions (CJ) to cross the mitochondrial outer membrane (MOM) via large channels (VDAC)
into the cytosol. 2 In the intermembrane space, mtCK associates simultaneously with the
mitochondrial inner membrane (MIM) and VDAC, forming so-called mitochondrial contact
sites between MIM and MOM for direct, vectorial channeling of PCr into the cytosol. 3 In the
cytosol, cytCK can associate with glycolytic ATP-generating enzymes or enzyme assemblies
for PCr synthesis. While in oxidative tissues like heart, PCr is mainly produced by mtCK using
ATP from oxidative phosphorylation, PCr in fast-twitch glycolytic muscle mainly comes from
cytCK and glycolytic ATP. PCr is then used by cytCK to buffer global cellular ATP/ADP ratios
(not shown). 4 More specifically, cytCK associated with or close to cytosolic ATPases uses
PCr to keep local ATP/ADP ratios high for maximal DG0#of#ATP#hydrolysis.#In#muscle,#for#
example,#the#cytosolic#MCK#isoenzyme#is#bound#to#myomesin#in#the#myofibrillar#M-band#
to# regenerate# ATP# close# to# the# myosin# head# ATPases# for# contraction.# In# addition,# a#
significant# fraction# of# cytosolic# MCK# is# also# specifically# bound# to# the# sarcoplasmic#
reticulum# (SR)-Ca2+ATPase# to# support# energy# demanding# Ca2+# pumping# (not# shown#
here).#
(B) Isoenzymes of nucleoside diphosphate kinase (NDPK, also called NME or NM23) use ATP
to maintain the cellular levels of different NTPs. Part of these NDPKs associate with GTPases
or G-proteins for direct channeling of the required GTP. 5 In mitochondria, NDPK-D (NME4)
directly interacts with the dynamin-related GTPase OPA1 to fuel GTP for MIM dynamics. 6
At the plasma membrane, cytosolic NDPK isoenzymes, NDPK-A (NME1) and NDPK-2
(NME2) associate with dynamin to fuel GTP for membrane constriction and dynamin-mediated
endocytosis.
(C) The ten enzymes of the glycolytic pathway (including glyceraldehyde-3-phosphate
dehydrogenase, phosphoglycerate kinase, pyruvate kinase) associate with vesicles that travel
along axons. This fast-axonal transport (FAT) is realized by the vesicle-bound molecular
motors kinesin and dynein that attach to microtubule filaments and propel vesicles in either
anterograde or retrograde direction, respectively. 7 The ATP for efficient FAT by kinesin or
dynein ATPases is mainly channeled from on-board vesicular glycolysis, but is not generated
by mitochondria or in bulk cytosol.

23
Change history
- Some minor revisions all over the text
- New paragraph in chapter 1 on recent methodological developments
- Chapter 3.4 deleted and integrated in chapter 3.3
- New chapters 4 and 5
- Revised Figure 1
- New Figure 2, replacing earlier Figures 2 and 3
- Deleted Figure 4
- In-text references, added references (preferentially reviews and book chapters)