Download When Cells Become Organelle Donors

REVIEWS
PHYSIOLOGY 28: 414 – 422, 2013; doi:10.1152/physiol.00032.2013
When Cells Become Organelle Donors
Robert S. Rogers and Jahar Bhattacharya
Lung Biology Laboratory, Department of Medicine, and
Department of Physiology & Cellular Biophysics, Columbia
University, College of Physicians & Surgeons, New York
[email protected]
More than 40 variations of intercellular organelle transfer, such as a mitochondria or lysosome originating in one cell being transported to another
cell, have been described. We review the mechanisms and subcellular
structures by which, and conditions under which, transfer occurs, with
particular attention paid to the role of external cell stress in activating
transfer. We propose a research agenda for answering key questions in this
burgeoning field.
Intercellular Organelle Transfer:
The Basics
More than 40 variations of intercellular organelle
transfer have been described, including endoplasmic reticulum/Golgi bodies, endosomes, lysosomes or the lysosome-related melanosome, and
mitochondria originating in one cell (organelle donor) and being transported to another (organelle recipient). Table 1 summarizes the known instances
414
of intercellular organelle transfer.1 Five general
observations regarding this phenomenon can be
made.
First, in vitro intercellular organelle transfer has
been described in a highly diverse assortment of
cell types, including human and rodent, differentiated and multipotent, malignant and benign, epithelial and mesenchymal, and neuronal. Thus it is
safe to conclude that the capability of one cell to
donate or receive an organelle from another cell is
clearly widespread throughout mammalian cell
types in culture. Yet, importantly, intercellular organelle transfer is not universal. For example, although mesenchymal stem cells and fibroblasts
donate mitochondria to mitochondria-deficient
lung adenocarcinoma cells in culture, platelets
may not (27a). There are numerous examples of
both homotypic (mesenchymal stem cell to mesenchymal stem cell) and heterotypic (mesenchymal stem cell to cardiomyocyte) organelle transfer
between cells derived from both same and different species. Moreover, there are several reports
where only one type of organelle is transferred
between two given cell types, whereas there are
others in which multiple types of organelles are
transferred simultaneously. Undoubtedly, the list
of cell types and cellular organelles known to participate in organelle transfer will continue to grow.
Second, the term “transfer” defines a process in
which organelles are donated from one cell to another by ATP-dependent mechanisms that engage
dedicated cytoskeletal and associated cell machinery. This definition of transfer excludes passive
1
We compiled this list by extracting results from the
widely cited publications in the field, plus all publications
which they cited, which cited them, or were identified as
“related citations” in PubMed. In addition, we queried
PubMed and Google Scholar with the search terms “organelle,” “endosome,” “endoplasmic reticulum,” “Golgi,”
“lysosome,” and “mitochondria,” followed by “intercellular transfer,” “intercellular exchange,” “donor,” and
“recipient,” (e.g., “lysosome intercellular transfer”) and
reviewed potentially relevant articles for reports of true
intercellular organelle transfer. Omissions of other published reports of intercellular organelle transfer are
unintentional.
1548-9213/13 ©2013 Int. Union Physiol. Sci./Am. Physiol. Soc.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
Since the report by Rustom et al. that cultured
cells are capable of forming “nanotubes” that
span up to several cell diameters in length and
that they use these conduits to transport whole
organelles between cells (26), the last decade has
seen a burst of interest in the phenomenon of
intercellular organelle exchange and transfer. Traditionally considered processes of intercellular communication include cell contact-dependent and
paracrine receptor-ligand interactions, synaptic
vesicle release and uptake, and ion flow through
gap junctions (2). However, organelle transfer is
in a sense a special form of intercellular communication, because it represents the transfer not
only of signals but also of defined intracellular
structures. The study of intercellular organelle
transfer is new and has thus far raised more
questions than provided answers. In particular,
the issues relate to whether intercellular organelle transfer contributes to the cell’s adaptive
response to external stress and whether the
transfer enhances cell survival and function. In
this review, we summarize current understanding of intercellular organelle transfer, including
the types of cells and organelles in which the
transfer is described, the effects the transfer has
on donor and recipient cells, the importance of
external stress in activating the transfer, the signaling pathways and cellular machinery involved, and, finally, future research directions
that might increase understanding of this
phenomenon.
REVIEWS
Table 1. Published reports of intercellular organelle transfer
From (Donor Cell Type)
To (Recipient Cell Type)
Endosomes
Rat pheocromocytoma cell
Rat kidney cell
Rat pheocromocytoma cell
Human macrophage
Rat hippocampal astrocyte
Human retinal pigment epithelial cell
Human macrophage
Rat hippocampal astrocyte
Rat pheocromocytoma cell
Human macrophage
Mouse catecholaminergic neuron
Rat ventricular cardiomyocyte
Rat ventricular cardiomyocyte
Rat fibroblast
Rat fibroblast
Mouse endothelial progenitor cell
Human umbilical vein endothelial cell
Rat pheocromocytoma cell
Rat kidney cell
Rat pheocromocytoma cell
Human macrophage
Rat hippocampal astrocyte
Human retinal pigment epithelial cell
Human macrophage
Rat hippocampal astrocyte
Rat pheocromocytoma cell
Human macrophage
Mouse catecholaminergic neuron
Rat fibroblast
Rat ventricular cardiomyocyte
Rat ventricular cardiomyocyte
Rat fibroblast
Human umbilical vein endothelial cell
Mouse endothelial progenitor cell
26
10
5
14
31
32
15
31
26
21
9
11
11
11
11
34
34
Human melanocyte
Human keratinocyte
Mouse melanocyte
Rat neonatal cardiomyocyte
Human mesenchymal stem cell
Human skin fibroblast
Human macrophage
Rat mesechymal stem cell
Human mesenchymal stem cell
Rat kidney renal tubular cell
Human mesenchymal stem cell
Mouse endothelial progenitor cell
Rat ventricular cardiomyocyte
Rat ventricular cardiomyocyte
Rat fibroblast
Rat fibroblast
Human proximal tubular epithelial cell
Human adipose-derived stem cell
Rat hippocampal astrocyte
Human mesenchymal stem cell
Human vascular smooth muscle cell
Human retinal pigment epithelial cell
Human pleural mesothelioma cell
Human mesenchymal stem cell
Mouse mesenchymal stem cell
Human mesenchymal stem cell
Mouse mesenchymal stem cell*
Human mesenchymal stem cell*
Human keratinocyte
Human keratinocyte
Mouse keratinocyte
Human endothelial progenitor cell
Human alveolar adenocarcinoma cell
Human alveolar adenocarcinoma cell
Human macrophage
Rat lung endothelial cell
Rat cardiomyocyte
Human mesenchymal stem cell
Rat kidney renal tubular cell
Human umbilical vein endothelial cell
Rat fibroblast
Rat ventricular cardiomyocyte
Rat ventricular cardiomyocyte
Rat fibroblast
Human proximal tubular epithelial cell
Mouse cardiomyocyte
Rat hippocampal astrocyte
Human vascular smooth muscle cell
Human mesenchymal stem cell
Human retinal pigment epithelial cell
Human pleural mesothelioma cell
Human osteosarcoma cell
Mouse mesenchymal stem cell
Human mesenchymal stem cell
Mouse alveolar epithelial cell*
Mouse alveolar epithelial cell*
27
27
33
16
27a
27a
21
22
24
23
23
35
11
11
11
11
8
1
31
29
29
32
19
7
12
12
12
12
Golgi/ER
Lysosomes
Melanosomes (lysosome-related
organelles)
Mitochondria
Reference
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
Organelle
*In vivo; all other reports in vitro. Note, melanosome transfer has been described in numerous publications in vivo and in vitro; reports reviewed
here were chosen for providing insight into mechanisms involved in that transfer.
transport of cargo between connected cells (10) or
passive uptake from the extracellular environment
(27a). Moreover, there are instances of bidirectional transfer of organelles in which two cells both
donate to and receive organelles from each other
(11, 34), and of strictly unidirectional transfer in
which one cell type always donates organelles and
the other always receives them (16).
Third, from the initial report of intercellular organelle transfer via nanotubes, and also from subsequent
reports, it is clear that intercellular transfer of whole
organelles occurs concomitantly with transfer of cytosol and plasma membrane components (8), as well
as of small molecules and ions (32).
Fourth, the frequency and efficacy with which
cells transfer organelles varies across cell types
and culture conditions. In the initial report of
nanotube-mediated intercellular organelle transfer
in rat pheocromocyotma cells, 10% of cultured cells
received a transferred organelle (26). Subsequent
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
415
REVIEWS
The Consequences of Intercellular
Organelle Transfer
Intercellular organelle transfer has been demonstrated to be crucial to several experimental models of cell survival under external stress. However,
evidence for its role in cellular reprogramming is
less robust.
Enhancing Probability of Cell Survival
Spees et al. developed A549 cells devoid of intrinsic
mitochondrial function by depleting mitochondrial DNA through weeks-long treatment with
ethidium bromide. These cells were only capable
of expansion in a permissive medium supplemented with pyruvate and uridine to enhance glycolysis. Most did not survive in standard media.
When co-cultured with mesenchymal stem cells or
fibroblasts that donated functional mitochondria,
mitochondria-deficient A549 cells acquired mitochondria, established aerobic respiration, and proliferated at normal rates in standard media.
Mitochondrial acquisition by A549 cells increased
ATP, oxygen consumption, and membrane potential and decreased lactate production consistent
with return of normal mitochondrial function in
the rescued clones (27a).
Work in our laboratory has demonstrated that
mitochondrial transfer from mesenchymal stem
cells to alveolar epithelial cells protects against
endotoxin-induced acute lung injury in mice, preserving cellular ATP production, attenuating the
decline in surfactant production and enhancing
mouse survival. Transfer of viable mitochondria is
necessary for these protective effects. No protection is afforded by treatment with mesenchymal
stem cells that are fully functional yet incapable of
organelle transfer (connexin-43 mutants) or by
those that are capable of organelle transfer but
contain dysfunctional mitochondria. In sepsisinjured lungs, exogenous MSCs appear to transfer
all of their mitochondria to the alveolar epithelium.
416
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
The subsequent fate of the mitochondria-depleted
MSC cell body presently remains unclear (12).
Transfer of lysosomes has also been shown to
enhance cell survival under certain conditions of
stress. Yasuda et al. (34) demonstrated that treatment with glycated collagen type I induces lysosomal dysfunction (an increase in lysosomal pH
and decreased autophagy), leading to reduced survival in human umbilical vein endothelial cells.
When endothelial progenitor cells are added to
create a co-culture, a protective effect is seen, mediated by the visualized transfer of lysosomes from
the endothelial progenitor cell to the human umbilical vein endothelial cell. This protective effect is
lost when lysosomal transfer is prevented by the
inhibition of nanotube formation (34).
Although most reports note that mitochondrial
transfer through nanotubes is beneficial for cell
survival, in high concentrations, donated mitochondria might be directly cytotoxic. When rat mesenchymal stem cells and rat lung microvascular endothelial
cells are co-cultured in a 1:1 ratio, the mesenchymal
stem cells induce apoptosis of the endothelial cells.
Mitochondrial transfer from the mesenchymal stem
cell to the endothelial cell has been visualized in this
experimental setting. These studies provide suggestive evidence that this mitochondrial transfer is an
essential mediator of the observed cytotoxicity,
which occurs through a cell contact-dependent increase in reactive oxygen species in the endothelial
cell (22).
Cellular Reprogramming
Intercellular organelle transfer has been studied
extensively in the context of cellular reprogramming, and several reports have detailed instances
in which a cell type’s differentiation or dedifferentiation is accompanied by intercellular organelle
transfer. Plotnikov et al. (23, 24) show that coculturing mesenchymal stem cells with renal tubular cells or cardiomyocytes leads to mitochondrial
transfer from the mesenchymal stem cell to the
differentiated cell and that, in time, the mesenchymal stem cell expresses proteins characteristic of
the differentiated cell (Tamm-Horfsall protein and
heavy chain myosin, respectively). However, it is
unknown whether selectively blocking the mitochondrial transfer would be sufficient to inhibit
reprogramming of the mesenchymal stem cell.
Acquistapace et al. (1) demonstrated that terminally differentiated mouse cardiomyocytes can
assume a more progenitor-like state when cocultured with human adipose-derived or bone marrow-derived stem cells. The process involves cell
fusion and mitochondrial transfer from adiposederived cells to cardiomyocytes. These effects require
functional mitochondria, since there is a marked reduction in the reprogramming process when the
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
reports more or less confirm this transfer rate (5,
34). Among cell subsets that form nanotube connections, up to three-quarters participate in organelle transfer (26) at transfer rates of 1–15 ␮m/min
(5, 10, 26, 31). Organelle transfer commences
within hours after cell plating (5, 26), reaching a
peak in 2–3 days (11, 24, 34).
Fifth, the vast majority of reports of intercellular
organelle transfer have been obtained through
studies in vitro (17). Exceptions are a report that a
canine venereal tumor acquires mitochondria
from its host (25) and that exogenously administered bone marrow-derived mesenchymal stem
cells donate mitochondria to injured alveolar epithelial cells of the mouse lung (12).
REVIEWS
The Connection Between Cellular
Stress and Intercellular Organelle
Transfer
Emerging understanding indicates that cell stress
might be a major determinant of intercellular organelle transfer. The transfer of the melanosome
from the melanocyte to the keratinocyte, a physiological process that regulates skin and hair pigmentation, is enhanced by UV stress (27). When
human umbilical vein endothelial cells (HUVECs)
and mouse endothelial progenitor cells (EPCs) are
co-cultured in standard media, 17% of the HUVECs
receive mitochondria from EPCs after 24 h. This
transfer doubles when the HUVECs are stressed by
exposure to the cytotoxic drug adriamycin (35). Of
note, under these conditions, mitochondrial transfer is unidirectional, from the EPC to the HUVEC.
However, after 72 h of co-culture, bi-directional exchange of lysosomes occurs under basal conditions,
such that ⬃15% of EPCs receive lysosomes from
HUVECs and vice versa (34). When HUVECs are
pretreated with glycated collagen type I, which induces lysosomal dysfunction, the fraction of cells
participating in lysosomal exchange doubles. Interestingly, the increase is driven entirely by lysosomal transfer in one direction, from the EPC to
the HUVEC (34).
Under some conditions, cell stress might be an
absolute prerequisite for organelle transfer to occur. Wang et al. (31) show that hippocampal astrocytes and neurons form nanotubes, and hence
probably induce intercellular organelle transfer,
only when the cells are stressed by serum starvation or hydrogen peroxide exposure. Cho et al. (7)
report that mesenchymal stem cells transfer mitochondria to osteosarcoma cells only if the cells are
completely depleted of mitochondria by pretreatment with ethidium bromide or rhodamine. Notably, no transfer occurs when transfection with
mutant mitochondrial DNA partially depletes mitochondrial function (7). These findings suggest
that, under certain conditions, major mitochondrial dysfunction in the recipient cell predetermines mitochondrial transfer from a donor.
Work in our laboratory suggests that cell damage
is essential for mitochondrial transfer in vivo. In
mice, mesenchymal stem cells injected into the
trachea successfully transfer mitochondria to alveolar epithelial cells of the lung when it is injured
with endotoxin but not when it is uninjured (12).
When injected into the healthy lung, the mesenchymal stem cells quickly migrate to the interstitium without forming the physical attachments to
the alveolar epithelium that are necessary to facilitate
mitochondrial transfer. The failure to transfer mitochondria to healthy alveolar epithelium is probably
caused by a general lack of interaction between mesenchymal stem cells and alveolar epithelium in the
absence of damage signals, but it is possible that the
specific lack of a signal needed to spur mitochondrial
transfer is also a contributing factor. Identifying signals that facilitate mitochondrial transfer after mesenchymal stem cell attachment to the target tissue is
of potential importance to the goal of engineering
exogenous stem cells for clinical use.
Taken together, these results suggest that intercellular organelle transfer can occur in many cell types
under basal conditions. However, it seems that stem
cells possess mechanisms for upregulating organelle
donation in response to injury signals originating
from the recipient cell. Although organelle transfer
may proceed in a bi-directional manner between
participating cells, the stressed cell tends to be the
net recipient of organelles, as opposed to being the
net donor. That a mammalian cell under threat from
environmental stress becomes more likely to accept
organelles from a neighboring cell, as opposed to
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
stem cells are rendered mitochondrially dysfunctional by treatment with ethidium bromide (1). However, it remains unclear whether the reprogramming
was hindered by a lack of functional mitochondria in
adipose-derived cells or whether there was a failure
of mitochondrial transfer from adipose-derived cells
to cardiomyocytes.
Vallabhaneni et al. (29) show that co-culturing
mesenchymal stem cells with vascular smooth muscle cells increases the proliferation rate of the mesenchymal stem cells and is accompanied by
mitochondrial transfer from the vascular smooth
muscle cell to the mesenchymal stem cell. However,
the enhancement of mesenchymal stem cell proliferation fails to occur when mitochondrial transfer is
prevented by blocking nanotube formation or when
the transferred mitochondria are dysfunctional due
to ethidium bromide pretreatment (29).
Although most studies of intercellular organelle
transfer focus on the beneficial effects of the transfer, it is less clear whether intercellular organelle
transfer contributes to pathology. Kadiu and Gendelman demonstrate that macrophages in vitro
sprout nanotubes, which facilitate the exchange of
endosomes and lysosomes between cells. HIV-1
may package in these endosomes and spread from
one macrophage to the next (14). It has also been
shown that the nanotubes that are capable of serving as conduits for intercellular organelle transfer
can also be “hijacked” for spreading of Mycobacterium bovis between macrophages (21) or prions
between neurons (9). Lou et al. (18, 19) have shown
that mitochondrial exchange occurs between pleural mesothelioma cells in vitro. However, it is not
known whether selectively interfering with this mitochondrial exchange impacts cancer cell survival
or proliferation (18, 19).
417
REVIEWS
“shedding” its own organelles, is perhaps an important adaptation to higher multicellular organization
and cooperation. This stands in contrast to the behavior of unicellular organisms, such as the ciliated
protozoa, which perform vigorous extrusion of mitochondria in response to heat shock or attack by
surface-targeted antibodies (4).
The Mechanics of Intercellular
Organelle Transfer
The study of intercellular organelle transfer has
burgeoned in parallel with the study of “tunneling
F-actin
ezrin
FIGURE 1. Proposed model of organelle transfer from a stem cell to
a differentiated epithelial cell in culture
A: the epithelial cell releases unknown protein or metabolite “damage” signals into
the culture medium in response to stress in the extracellular environment (step 1).
The stem cell receives the epithelial cell damage signal, stimulating nanotube formation (step 2). B: the stem cell projects a nanotube toward the epithelial cell, creating a membrane-contiguous conduit between the cells (step 3). Step 3 can occur
in the absence of a damage signal, but occur with greater frequency when one is
present. The proteins ezrin and zona occludens-1 (ZO-1) anchor the nanotube at its
base on the stem cell, and the nanotube is supported by an F-actin cytoskeleton
attached to myosin. C: an organelle, in this case a mitochondrion, enters the nanotube from the stem cell, is attached to the actin-myosin cytoskeleton by unknown
attachment proteins, and is propelled by the myosin along the length of the nanotube until it is deposited in the cytosol of the epithelial cell (step 4). Note that the
nanotube is sprouted by the stem cell, which acts as the organelle donor, whereas
the distressed epithelial cell is the organelle recipient.
418
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
ZO-1
nanotubes,” the principal conduit through which organelles travel from one cell to the other (for an
excellent review, see Ref. 20). Tunneling nanotubes
were initially defined as “ultrafine intercellular structures” with diameters of 50 –200 nm and lengths
spanning several cell diameters that connect cells
without contacting the substratum (26). Subsequent
reports confirm that these characteristics are shared
by nanotubes across diverse cell types and culture
conditions (16, 31). Those nanotubes, which serve as
conduits specifically for mitochondrial transfer,
might be larger in diameter than 200 nm to accommodate these larger organelles.
In multiple studies, organelle motion in nanotubes has been visualized during the process of
intercellular organelle transfer (5, 10, 26). Realtime studies by fluorescence microscopy indicate
that that the nanotube protrudes from the initiating cell as an extension of the plasma membrane.
The leading edge of the protrusion makes contact
and then fuses with the plasma membrane of the
recipient cell, creating a contiguous path formed
by cytosol, actin cytoskeleton, and the surrounding
plasma membrane. Organelles enter this conduit
from the organelle-donating cell, traverse the conduit along the actin cytoskeleton, and are deposited in the recipient cell’s cytosol (26).
Two lines of evidence suggest that nanotubes are
both sufficient and necessary as the structural
bridge for intercellular organelle transfer. They are
sufficient because, when cultured under cooled
conditions in which nanotube formation occurs,
but other transfer processes such as endocytosis,
exocytosis, and phagocytosis are blocked, intercellular organelle transfer is sustained between rat
pheocromocytoma cells and rat kidney cells (10,
26). However, the mechanism that preserves nanotube transport under cooled conditions remains
unclear. Nanotubes are necessary because, when
cultured under conditions that prevent nanotube
formation, such as continuous physical shaking of
the cells that breaks nanotubes, or chemical inhibition of nanotube formation, intercellular organelle
transfer is substantially reduced in proportion to the
reduction of nanotube formation (5, 10, 26).
The mechanisms of nanotube formation and
their composition are beginning to be understood.
Several reports indicate that nanotubes contain
F-actin but not microtubules, although Onfelt et al.
(21) report a special type of “thick nanotube” of
⬎700 nm that contains both F-actin and microtubules. Bukoreshtliev et al. (5) note that the majority
of nanotubes form by a process in which a cell
extends a filopodium to make contact with a
neighboring cell. The filopodium then retracts,
leaving a nanotube connection between the cells.
In a minority of cases, nanotubes form as the residual connection between two cells that were in
REVIEWS
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
direct contact and subsequently dislodged (5). Wittig et al. (32) showed that cells separated by a
distance of ⬎100 ␮m cannot form interconnecting
nanotubes, which likely explains the observed substantial reduction in intercellular organelle transfer
between two cell populations when they are physically separated by this distance or greater (23, 24).
Nanotubes impart directionality on intercellular
organelle transfer. For any one nanotube, organelles
travel in one direction, from the cell that sprouted the
nanotube to the cell that received the nanotube and
not vice versa (26). Organelles are transferred by myosin engagement, with the F-actin skeleton of the
nanotube. Blocking myosin involvement can impede
organelle transfer, or transfer can be enhanced by
specifically blocking retrograde movement powered
by myosin (10). These findings suggest that a complex interplay of mechanisms regulate anterograde
and retrograde organelle movement toward the recipient cell. These mechanisms remain to be elucidated. Fascin is expressed at the base of the
nanotube, and ezrin and zona occludens-1 occur at
the site of extrusion from the plasma membrane (19);
it is almost certain that Rho GTPases play some role
(3), although ␤-catenin and E-cadherin are notably
absent (19). Proteins that connect membrane-bound
organelles to myosin and support organelle movement along the nanotube remain undefined. However, the goal of defining these proteins may be
within reach since the “nanotube proteome” consisting of 275 proteins is now defined (15). FIGURES 1–3
propose an integrated model of the key structural
and signaling mediators of intercellular organelle
transfer along nanotubes.
A potential explanation for the finding that intercellular organelle transfer occurs at a basal rate that
can then be further stimulated by exposure to cell
stress is that this is a property of nanotube formation
itself. In cultured pleural mesothelioma cells, nanotube formation increases in acidified, hyperglycemic,
or low-serum media, as well as when the media is
supplemented with cytokines known to facilitate
epithelial-to-mesenchymal transition (EGF, FGF,
TGFB-3) (19). Similarly, treatment with the antibiotic
zeocin stimulates nanotube formation in human renal tubular epithelial cells (8).
Wang et al. (31) showed that, in rat hippocampal
neurons and astrocytes, nanotube formation is dependent on activation of the pro-apoptotic p53
protein, which through EGFR activates Akt/MTOR/
PI3K pathways to initiate nanotube formation.
Subsequently, they have demonstrated that the direction of nanotube formation depends on the establishment of a gradient in the concentration of
the protein S100A4, with the nanotube-initiating
cell having low levels of S100A4 relative to the
nanotube-receiving cell. They observed that the
activation of caspase-3, a downstream effector of
FIGURE 2. Proposed model of organelle transfer between two neurons in
cell culture
A: neuron A is damaged by exposure to stress in the extracellular environment. B: in
response, p53 is activated in neuron 1, initiating a signaling cascade through activation
of EGFR, which in turn activates the PI3K/Akt/mTOR pathway, which in turn activates
caspase 3, which cleaves S100A4, reducing the concentration of S100A4, allowing the
damaged neuron 1 to sprout a nanotube in the direction of the undamaged, relatively
S100A4-rich neuron 2. C: an organelle, in this case a mitochondrion, enters the nanotube from neuron 1, is attached to the actin-myosin cytoskeleton by unknown attachment proteins, and is propelled by the myosin along the length of the nanotube until it
is deposited in the cytosol of neuron 2. Note that the nanotube is sprouted by the
damaged neuron, which acts as the organelle donor, whereas the uninjured neuron 2 is
the organelle recipient.
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
419
REVIEWS
FIGURE 3. Proposed model of organelle transfer from a stem cell to a
differentiated epithelial cell in mice
A: the epithelial cell is damaged by release of TNF-␣ by leukocytes as part of the
systemic immune response to infection. B: in response to injury, the epithelial cell
increases expression of connexin-43 (Cx43). The exogenously administered stem
cell also expresses Cx43. C: the stem cell sprouts a nanotube toward the injured
epithelial cell dependent on Cx43 homotypic interaction. Once the nanotube connection is established, there is bidirectional exchange of unknown signaling molecules. D: in response to signals from the epithelial cell, the stem cell releases a
membrane-coated microvesicle that encases a membrane-bound organelle, in this
case a mitochondrion, onto the surface of the epithelial cell, which then engulfs
the microvesicle and its mitochondrion cargo through dynamin-dependent endocytosis. Note that the Cx43-dependent stabilizing interaction between stem cell
and epithelial cell occurs only in the presence of epithelial cell damage.
420
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
transfer via nanotubes. Perhaps a partial explanation is that signaling mechanisms that govern organelle transfer are essentially different for transfer
from progenitor to differentiated cells vs. transfer
among two differentiated cells, such as neurons.
Nanotubes have been identified in live tissues,
including between cardiomyocytes and fibroblasts
in the adult mouse heart (11), in the tumors of
patients with pleural mesothelioma and poorly differentiated lung adenocarcinoma (19), and in dendritic cells in the mouse cornea (6). However, there
are likely important differences in the composition
and characteristics of nanotubes in vivo and in
vitro. For example, nanotubes that withstand the
constant motion of the beating heart must in some
way differ from those that are highly fragile when
exposed to gentle shaking of the culture medium
(10, 26). Moreover, it has not been demonstrated
that nanotubes play as essential a role in intercellular organelle transfer in vivo. They are strikingly
absent in the one known example of endogenous
intercellular organelle transfer, that of the melanosome. Wu et al. (33) recently described this process,
and it involves adhesion between a melanocyte dendrite and keratinocyte, thinning of the dendrite behind its tip, abscission of the dendrite tip in contact
with the keratinocyte, and finally phagocytosis of the
extracellular package by the keratinocyte. In the case
of mitochondrial transfer from mesenchymal stem
cells to injured alveolar epithelial cells in the mouse
lung, the mesenchymal cells form processes that
contain mitochondria. However, the processes
seem not to be the specific mechanism for the
mitochondrial transfer. Rather, mesenchymal stem
cells release vesicles containing mitochondria, which
are engulfed by the epithelial cells in an endocytosisdependent manner, which can be blocked by the
dynamin inhibitor dynasore (12).
These observations raise the possibility that intercellular organelle transfer in vivo, to the extent it
occurs, involves somewhat different cellular machinery than that described in vitro. This is not
surprising, given a growing understanding that
living organisms use a variety of mechanisms to
communicate complex macromolecules over long
distances, even transporting miRNA through the
plasma via HDL (30).
Unanswered Questions and a
Research Agenda
The most pressing question in the field of intercellular organelle transfer is: To what extent does this
phenomenon occur in vivo? Visualizing intercellular
organelle transfer in real time in the intact living
organism remains a major methodological challenge,
although doing so in the ex vivo perfused organ is
feasible, particularly with the provision of exogenous
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
p53 under conditions of cell stress, cleaves S100A4,
leading to a decrease in S100A4 in the injured cell,
which induces that cell to initiate nanotube formation (28). Consistently, they observed nanotube
formation to be a property of the stressed (i.e.,
p53-activated) cell itself, not its healthy neighbors,
even demonstrating that exposing healthy cells to
the medium in which stressed cells were cultured
does not prompt them to form nanotubes (31).
A paradox arises from these observations. On the
one hand, as discussed in the previous section,
mitochondria and lysosomes tend to be transferred
with directional specificity from the uninjured to
the injured cell. Yet, in a pair of cells connected by
nanotubes, the cell that sprouts the nanotube is
the organelle donor, whereas the other cell is the
recipient. This is difficult to square with the finding
that the stressed cell, which would be expected to
be primarily an organelle recipient, is the cell that
sprouts the nanotube, since nanotube sprouting is
presumably the behavior primarily of an organelle
donor. We await further experiments to resolve
these conflicting models of intercellular organelle
REVIEWS
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: R.S.R. performed experiments; R.S.R.
analyzed data; R.S.R. and J.B. interpreted results of experiments; R.S.R. prepared figures; R.S.R. drafted manuscript;
J.B. conception and design of research; J.B. approved final
version of manuscript.
References
1.
Acquistapace A, Bru T, Lesault PF, Figeac F, Coudert AE, le
Coz O, Christov C, Baudin X, Auber F, Yiou R, Dubois-Randé
JL, Rodriguez AM. Human mesenchymal stem cells reprogram adult cardiomyocytes toward a progenitor-like state
through partial cell fusion and mitochondria transfer. Stem
Cells 29: 812– 824, 2011.
2.
Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Molecular Biology of the Cell. New York: Garland Science, 2008.
3.
Arkwright PD, Luchetti F, Tour J, Roberts C, Ayub R, Morales
AP, Rodríguez JJ, Gilmore A, Canonico B, Papa S, Esposti
MD. Fas stimulation of T lymphocytes promotes rapid intercellular exchange of death signals via membrane nanotubes.
Cell Res 20: 72– 88, 2010.
4.
Bisharyan Y, Clark TG. Calcium-dependent mitochondrial extrusion in ciliated protozoa. Mitochondrion 11: 909 –918, 2011.
5.
Bukoreshtliev N, Wang X, Hodneland E, Gurke S, Barroso JF,
Gerdes HH. Selective block of tunneling nanotube formation
inhibits intercellular organelle transfer between PC12 cells.
FEBS Lett 583: 1481–1488, 2009.
6.
Chinnery HR, Pearlman E, McMenamin PG. Cutting edge:
membrane nanotubes in vivo: a feature of MHC class II⫹ cells
in the mouse cornea. J Immunol 180: 5779 –5783, 2008.
7.
Cho YM, Kim JH, Kim M, Park SJ, Koh SH, Ahn HS, Kang GH,
Lee JB, Park KS, Lee HK. Mesenchymal stem cells transfer
mitochondria to the cells with virtually no mitochondrial function but not with pathogenic mtDNA mutations. PLos One 7:
32778 –32785, 2012.
8.
Domhan S, Ma L, Tai A. Intercellular communication by exchange of cytoplasmic material via tunneling nano-tube like
structures in primary human renal epithelial cells. PLos One 6:
21283–21290, 2011.
9.
Gousset K, Schiff E, Langevin C, Marijanovic Z, Caputo A,
Browman DT, Chenouard N, de Chaumont F, Martino A,
Enninga J, Olivo-Marin JC, Männel D, Zurzolo C. Prions hijack
tunneling nanotubes for intercellular spread. Nat Cell Biol 11:
328 –336, 2009.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
organelle donor cells, which can be fluorescently
labeled (12). However, these studies do not clarify
whether intercellular organelle transfer occurs endogenously. For example, do mesenchymal stem
cells transfer mitochondria to hematopoietic cells in
their natural bone marrow niche? And, if so, under
what circumstances? Future studies might address
this question in the context of specific cell types such
as multipotent adult stem cells, mesenchymal stromal cells surrounding epithelial parenchyma in various organs, fibroblasts, and hematopoietic stem
cells that are likely to be organelle donors.
Future research also needs to address the mechanisms that couple organelle transfer and nanotube formation. Importantly, it is not clear whether
the two processes are interdependent, and, if so,
how the interdependence is set up. Gurke et al. (10)
showed that the myosin inhibitor BDM increases
nanotube formation while decreasing organelle
transfer, suggesting that the processes might be
mechanistically distinct. However, it is not clear
whether organelle transfer independently evolved
as an adaptive mechanism.
Finally, future therapeutic implications of this research need to consider strategies to pharmacologically augment intercellular organelle transfer when
desirable (i.e., to replenish dysfunctional mitochondrial and lysosomes in the cell under stress) or block
its occurrence when it is deleterious, such as in the
spread of infection or the maintenance of malignant
cells. The reports reviewed here suggest that cell
stress is a powerful stimulus for intercellular organelle transfer, and much work has already been done
to elucidate its key functional regulators and associated cytoskeletal machinery. Yet, we are far from a
full, detailed delineation of the protein interactions
and signaling cascades involved in a transfer event. 䡲
10. Gurke S, Barroso JF, Hodneland E, Bukoreshtliev NV,
Schlicker O, Gerdes HH. Tunneling nanotube-like structures
facilitate a constitutive, actomyosin-dependent exchange of
endocytic organelles between normal rat kidney cells. Exp
Cell Res 314: 3669 –3683, 2008.
11. He K, Shi X, Zhang X. Long-distance intercellular connectivity
between cardiomyocytes and cardiofibroblasts mediated by
membrane nanotubes. Cardiovasc Res 92: 39 – 47, 2011.
12. Islam MN, Das SR, Amin MT, Wei M, Sun L, Westphalen K,
Rowlands DJ, Quadri SK, Bhattacharya S, Bhattacharya J.
Mitochondrial transfer from bone-marrow-derived stromal
cells to pulmonary alveoli protects against acute lung injury.
Nat Med 18: 759 –766, 2012.
14. Kadiu I, Gendelman H. Human immunodeficiency virus type 1
endocytic trafficking through macrophage bridging conduits
facilitates spread of infection. J Neuroimmune Pharmacol 6:
658 – 675, 2011.
15. Kadiu I, Gendelman HE. Macrophage bridging conduit trafficking of HIV-1 through the endoplasmic reticulum and golgi
network. J Proteome Res 10: 3225–3238, 2011.
16. Koyanagi M, Brandes R, Haendeler J. Cell-to-cell connection
of endothelial progenitor cells with cardiac myocytes by
nanotubes: a novel mechanism for cell fate changes? Circ Res
96: 1039 –1041, 2005.
17. Linch C, Whiting D, Holland M. Human hair histogenesis for
the mitochondrial DNA forensic scientist. J Forensic Sci 46:
844 – 853, 2001.
18. Lou E, Fujisawa S, Barlas A, Romin Y, Manova-Todorova K,
Moore MA, Subramanian S. Tunneling nanotubes: a new
paradigm for studying intercellular communication and therapeutics in cancer. Commun Integr Biol 5: 399 – 403, 2012.
19. Lou E, Fujisawa S, Morozov A, Barlas A, Romin Y, Dogan Y,
Gholami S, Moreira AL, Manova-Todorova K, Moore MA.
Tunneling nanotubes provide a unique conduit for intercellular transfer of cellular contents in human malignant pleural
effusion. PLos One 7: 33093–33103, 2012.
20. Marzo L, Gousset K, Zurzolo C. Multifaceted roles of tunneling nanotubes in intercellular communication. Front Physiol
3: 1–14, 2012.
21. Onfelt B, Nedvetzki S, Benninger RK, Purbhoo MA, Sowinski
S, Hume AN, Seabra MC, Neil MA, French PM, Davis DM.
Structurally distinct membrane nanotubes between human
macrophages support long-distance vesicular traffic or surfing of bacteria. J Immunol 177: 8476 – 8483, 2006.
22. Otsu K, Das S, Houser SD, Quadri SK, Bhattacharya S, Bhattacharya J. Concentration-dependent inhibition of angiogenesis by mesenchymal stem cells. Blood 113: 4197– 4205,
2009.
23. Plotnikov EY, Khryapenkova TG, Galkina SI, Sukhikh GT, Zorov DB. Cytoplasm and organelle transfer between mesenchymal multipotent stromal cells and renal tubular cells in
co-culture. Exp Cell Res 316: 2447–2455, 2010.
24. Plotnikov EY, Khyrapenkova TG, Vasileva AK, Marey MV.
Cell-to-cell cross-talk between mesenchymal stem cells and
cardiomyocytes in co-culture. J Cell Mol Med 12: 1622–1631,
2008.
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org
421
REVIEWS
25. Rebbeck C, Leroi A, Burt A. Mitochondrial capture
by a transmissible cancer. Science 331: 303, 2011.
26. Rustom A, Saffrich R, Markovic I, Walther P,
Gerdes HH. Nanotubular highways for intercellular organelle transport. Science 303: 1007–1010,
(2004).
27. Singh SK, Kurfurst R, Nizard C, Schnebert S, Perrier E, Tobin DJ. Melanin transfer in human skin
cells is mediated by filopodia - a model for homotypic and heterotypic lysosome-related organelle transfer. FASEB J 24: 3756 –3769, 2010.
27a. Spees JL, Olson SD, Whitney MJ, Prockop DJ.
Mitochondrial transfer between cells can rescue
aerobic respiration. Proc Natl Acad Sci USA 103:
1283–1288, 2006.
28. Sun X, Wang Y, Zhang J, Tu J, Wang XJ, Su XD,
Wang L, Zhang Y. Tunneling-nanotube direction
determination in neurons and astrocytes. Cell
Death Dis 3: 438 – 452, 2012.
29. Vallabhaneni K, Haller H, Dumler I. Vascular
smooth muscle cells initiate proliferation of mesenchymal stem cells by mitochondrial transfer via
tunneling nanotubes. Stem Cells Develop 21:
3104 –3113, 2012.
33. Wu XS, Masedunskas A, Weigert R, Copeland NG,
Jenkins NA, Hammer JA. Melanoregulin regulates
a shedding mechanism that drives melanosome
transfer from melanocytes to keratinocytes. Proc
Natl Acad Sci USA 109: 2101–2109, 2012.
30. Vickers KC, Palmisano BT, Shoucri BM, Shamburek RD, Remaley AT. MicroRNAs are transported
in plasma and delivered to recipient cells by highdensity lipoproteins. Nat Cell Biol 13: 423– 433,
2011.
34. Yasuda K, Khandare A, Burianovskyy L, Maruyama
S, Zhang F, Nasjletti A, Goligorsky MS. Tunneling
nanotubes mediate rescure of prematurely senescent endothelial cells by endothelial progenitors:
exchange of lysosomal pools. Aging (Milano) 3:
597– 608, 2011.
31. Wang Y, Cui J, Sun X, Zhang Y. Tunnelingnanotube development in astrocytes depends on
p53 activation. Cell Death Differ 18: 732–742,
2011.
32. Wittig D, Wang X, Walter C. Multi-level communication of human retinal pigment epithelial cells
via tunneling nanotubes. PLos One 7: 33195–
33203, 2012.
35. Yasuda K, Park HC, Ratliff B, Addabbo F, Hatzopoulos AK, Chander P, Goligorsky MS. Adriamycin
nephropathy: a failure of endothelial progenitor
cell-induced repair. Am J Pathol 176: 1685–1695,
2010.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.3 on June 15, 2017
422
PHYSIOLOGY • Volume 28 • November 2013 • www.physiologyonline.org