The International Journal of Biochemistry & Cell Biology 44 (2012) 718–721
Contents lists available at SciVerse ScienceDirect
The International Journal of Biochemistry
& Cell Biology
journal homepage: www.elsevier.com/locate/biocel
Organelles in focus
Golgi apparatus: Homotypic fusion maintains biochemical gradients within the
Golgi and improves the accuracy of protein maturation
Serge Dmitrieff, Pierre Sens∗
CNRS – UMR 7083 Gulliver ESPCI, 10 rue Vauquelin, 75231 Paris Cedex 05, France
a r t i c l e
i n f o
Article history:
Received 29 October 2011
Received in revised form 18 January 2012
Accepted 19 January 2012
Available online 28 January 2012
Keywords:
Golgi apparatus
Chemical identity
Homotypic fusion
Protein maturation
Dynamical phase transition
a b s t r a c t
Eukaryotic cells are compartmentalized into organelles which, although constantly exchanging proteins
and lipids with their environment, maintain a relatively well-defined biochemical identity. How can such
large heterogeneities of chemical composition between (and within) organelles be maintained if different organelles are in constant contact through mass transport? Generic nonlinearities in the transport
processes, as would result from specific molecular interactions, can cause the spontaneous chemical
differentiation of interacting organelles and compartments within organelles. For the Golgi apparatus,
the role of which is to process an incoming flux of lipids and proteins, this spontaneous differentiation
decreases inter-cisternal exchange and increases the protein transit time under conditions of high incoming flux, This mechanism enables the Golgi apparatus to spontaneously adjust the protein transit time to
the amount of protein requiring processing, thereby improving the processing accuracy of even a limited
amount of maturation enzymes.
© 2012 Elsevier Ltd. All rights reserved.
Organelle facts
• The Golgi apparatus is generally divided into stacked
sub-compartments, called cisternae, which have different
chemical identities.
• The Golgi apparatus is polarized, and many proteins synthesized in the Endoplasmic Reticulum (E.R.) enter the Golgi at
the cis face.
• Proteins mature during their progression through the Golgi
stack before exiting at the Golgi trans face.
• Identity of organelles enable a targeting of transport by specific molecular interactions, for instance by SNARE proteins.
• There is growing evidence for quality control throughout the
secretion pathway, including the Golgi apparatus.
• Defects in protein maturation, for instance of glycosylation,
are involved in many diseases, including various cancers.
1. Introduction
Compartmentalization was a crucial step towards the existence of complex living systems. The cell itself is a compartment,
∗ Corresponding author. Tel.: +33 1 40 79 45 52; fax: +33 1 40 79 47 31.
E-mail address: [email protected] (P. Sens).
1357-2725/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biocel.2012.01.016
exchanging molecules with its environment, and organelles within
the cell also constantly exchange material with one another
(Misteli, 2011). They must however retain a specific biochemical
and physical identity (in terms of the protein and lipid composition, pH, etc., Wilson and Ragnini-Wilson, 2010; Munro, 2005)
necessary for their biological function. At the center of the secretory pathway, the Golgi apparatus is an archetypical example of
a dynamical organelle, organized into compartments with welldefined biochemical identities, and constantly exposed to flux of
molecules (Lippincott-Schwartz et al., 2000). The Golgi imports
immature proteins and lipids from the E.R. and exports them after a
succession of chemical maturation steps (Rabouille et al., 1995), for
instance, the conversion of ceramids into sphyngolipids, or protein
glycosylation (Hirschberg and Snider, 1987). Maturation involves
successive chemical reactions in the different Golgi compartments,
each possessing specific enzymes (Dunphy et al., 1981).
The dominant mechanism for protein transport through the
Golgi apparatus is still controversial (Elsner et al., 2003). The vesicular transport model proposes that the different Golgi cisternae
(typically five to eight of them, usually regrouped into cis, medial
and trans Golgi regions) are fairly static structures between which
material is exchanged by vesicles. The cisternal maturation model
suggests that the cisternae themselves move through the stack
and change their chemical identity with time (Bonfanti et al.,
1998). Transport through the Golgi might be a combination of both
mechanisms. Here, we focus on vesicular exchange between compartments, a mechanism also relevant to the transport between
other cellular organelles; E.R. to Golgi transport for instance, or
S. Dmitrieff, P. Sens / The International Journal of Biochemistry & Cell Biology 44 (2012) 718–721
transport from the plasma membrane to endosomes along the
endocytic pathway.
If the rates of exchange between the Golgi compartments were
constant and uniform, the flux of a given molecular species between
two compartments would vary linearly with the concentration difference between the two compartments. This type of transport is
equivalent to Fick’s law: J = −DdC/dx which governs diffusive processes, and would naturally result in uniform protein and lipid
concentrations throughout the Golgi (J is the flux, dC/dx is the local
concentration gradient, and D is the diffusion coefficient). Generically, the linear relationship between concentration and fluxes
should break down at high protein concentration because of the
finite capacity of transport vesicles, or the finite availability of coat
proteins. This saturation mechanism (similar to the saturation of
enzymatic reaction in the Michaelis–Menten kinetics, Michaelis
and Menten, 1913) would however not prevent the homogenization of the Golgi compartments, but would merely slow down the
process. There is a growing interest in identifying and modeling the
molecular processes that bias transport and permit the existence
of stationary concentration gradients within and between cellular organelles (Heinrich and Rapoport, 2005; Binder et al., 2009).
They likely involve specific molecular interactions during vesicle
docking and fusion, for instance mediated by matching pairs of
SNARE proteins (McNew et al., 2000, Chen and Scheller, 2001).
Other types of specific interactions, e.g. involving coat proteins
(Bremser et al., 1999) or molecular motors (Jordens et al., 2001),
might influence transport and it appears important to go beyond
molecular details and to understand the implication of generic
transport non-linearities for intra-cellular transport.
We have shown (Dmitrieff and Sens, 2011) that the conjunction of two effects; (i) saturation of the vesicular carriers and (ii)
specific molecular interactions during vesicle transport or fusion,
creates a dynamical switch beyond a critical concentration, which
promotes the retrofusion of secreted vesicles with their compartment of origin. This apparently wasteful phenomenon permits the
existence of chemical gradients within the organelle. It also slows
down the transit of molecules through the Golgi apparatus, and
can strongly increase the organelle efficiency at processing a large
influx of proteins.
2. Organelle function
Our simplified model system only includes two compartments
(compartments 1 and 2). The first compartment, the “cis-Golgi”,
receives an incoming flux of material (e.g. secreted by the E.R.).
This material undergoes chemical maturation and is transported to
the second compartment, the “trans-Golgi”, before being exported
to the rest of the cell. In our model, the chemical identity of a compartment is characterized by the concentration C of a single species
(C1 in compartment 1 and in C2 compartment 2, Ctot = C1 + C2 the
total concentration in the system). Inter-compartment exchange
yields a set of kinetic equations for the two concentrations:
∂t C1 = Jin − (k1→2 C1 − k2→1 C2 )
!"#$
influx
! "# $
flux 1→2
! "# $
!
"#
net flux 1→2
$
! "# $
to the concentration in the emitting and receiving compartments.
Two generic types of non-linearities are introduced in the model
(Fig. 1b); (i) vesicular carriers have limited capacity so that the flux
of proteins secreted by a compartment saturates at high protein
content; Cs is the characteristic concentration for saturation, and
(ii) vesicle fusion involves specific molecular interactions, so that
the probability of retro and forward fusion depends on the concentrations in the two compartments; Cf is the concentration beyond
which specific molecular interactions has a sizable influence on
vesicle fusion (see Dmitrieff and Sens, 2011, for a mathematically
precise definition of Cs and Cf ).
For a closed system (without incoming and outgoing fluxes,
Jin = kout = 0), we asked whether compartments following nonlinear exchange rules could spontaneously differentiate into
different biochemical entities. The results are illustrated in Fig. 1. At
low concentration (Ctot $ Cs , Cf ) transport is nearly linear. Heterogeneities between the two compartments relax (diffusively) with
time and the two compartments tend to be chemically identical.
At higher concentration (Ctot larger than a critical concentration C*
depending upon Cs and Cf ), compartments spontaneously differentiate and adopt distinct chemical identities (Fig. 1c). This critical
behavior is due to the fact that a vesicle is more likely to undergo
fusion with a compartment of similar composition, a mechanism
called homotypic fusion. The composition of a vesicle reflects the
composition of the emitting compartment, and it has the tendency
to undergo retrofusion with this compartment. Beyond the critical
concentration C*, retrofusion becomes dominant, the symmetric
state becomes unstable and one compartment becomes enriched
in the chemical species of interest (Fig. 1d). In this regime, the flux
of material exchanged between the two compartments actually
decreases with an increasing total concentration.
For an open system, in which material is imported in compartment 1 (with a flux Jin ) and exported from compartment 2, at a rate
kout (Fig. 2), the total concentration in the system results from the
balance between the incoming and outgoing fluxes (Ctot = Jin /kout ).
If the exchange between the two compartments follows the rules
outlined above, two qualitatively different regimes can be expected
with varying influx: (i) at low influx, the total concentration in the
system is smaller than C*, and exchange between compartments is
nearly linear, and (ii) at high flux, the concentration is above critical and the exchange between compartments is severely reduced.
This can be seen from the evolution of a short pulse of incoming
material (pulse-chase procedure, Fig. 2b). In the linear (low pulse)
regime, the initiation of the pulse leads to a steady increase of the
concentrations C1 and C2 in both compartments, and when the
pulse stops, both concentrations eventually decrease because of
material export. In the non-linear (high pulse) regime, the critical
concentration C* is reached, and exchange between compartments
is impaired. Consequently, the concentration C2 remains low at all
times, and the system takes a longer time to export all the imported
material.
3. Cell physiology
flux 2→1
∂t C2 = (k1→2 C1 − k2→1 C2 ) − kout C2
719
(1)
outflux
where ∂t represents the time derivative, Jin and kout are respectively the incoming flux (in compartment 1) and the rate of exit
(from compartment 2). k1→2 and k2→1 are the rates of transport (in
s−1 ) from compartment 1 to 2, and compartment 2 to 1, respectively. These rates include multiple steps, which can broadly be
grouped into a rate of secretion and a rate of fusion, as sketched
in Fig. 1. Linear transport corresponds to constant rates, insensitive
The increased transit time apparent in Fig. 2b may seem to be
a functional drawback for a transport organelle such as the Golgi
apparatus. However, transport through the Golgi apparatus is typically quite slow considering the small – micrometric – size of the
organelle (the typical protein transit time is of order 20 min in
mammalian cells, Bonfanti et al., 1998). Furthermore, longer transit
times under high pulse conditions seem to be observed experimentally (Trucco et al., 2004), although very few data are available at
this time. The role of the Golgi is to expose transit proteins to specific enzymes for chemical maturation. We argue that in this case,
the dynamical switch from fast to slow transport under conditions
720
S. Dmitrieff, P. Sens / The International Journal of Biochemistry & Cell Biology 44 (2012) 718–721
Fig. 1. Compartment differentiation in a closed system: (a) Sketch of the system, composed of a compartment 1 (with typical concentration C1 ) and a compartment 2 (with
typical concentration C2 ). Each compartment secretes transport vesicles, which may then fuse back with the emitting compartment (retrofusion), or fuse forward with
the other compartment. Specific molecular interactions are introduced by letting the fusion probability depend on the concentration of particular molecular species in the
membrane of both vesicle and receiving compartment. (b) The concentration-dependent flux J1→2 = k1→2 C1 from compartment 1 to 2 (see Eq. (1)) illustrates the non-linearities
discussed in the text. Linear exchange (light gray) corresponds to constant secretion of transport vesicles, and constant forward fusion probability. The limited capacity of
vesicular carriers causes the exchanged flux to saturate (dark gray) beyond a characteristic concentration Cs . Specific molecular interactions (specific fusion, black) increase
the probability of retrofusion and cause the exchanged flux to drop beyond the concentration Cf . (c) Phase showing the existence of a dynamical switch from identical to
differentiated compartments above a critical concentration C* “specific fusion” is Cf /Ctot and “saturation” is Cs /Ctot with Ctot = C1 + C2 . Increasing the concentration Ctot moves the
system along the blue arrow. The critical concentration is the intersection marked by a red arrow. (d) Dynamical switch: below C*, the concentrations in both compartments
are identical, whereas beyond C*, one compartment is enriched and the other depleted in the chemical species of interest.
of high incoming flux offers a definite functional advantage. One
could expect that a high influx of immature proteins would saturate
the available maturation enzymes, letting some immature proteins
go through the Golgi unprocessed. It was shown mathematically
in Dmitrieff and Sens, 2011 that the dynamical switch discussed
here can prevent this from happening. Indeed, although the maturation enzymes are saturated with substrate under conditions
of high incoming flux (for instance, following Michaelis–Menten
kinetics), inter-compartment exchange is yet more reduced. Maturation can take place during a much longer time, leading to a
comparatively higher processing efficiency (the fraction of processed to unprocessed proteins that leave the system increases with
the total protein concentration). Remarkably, homotypic fusion, the
mechanism that permits the maintenance of cisternal identity, also
improves maturation quality control.
4. Organelle pathology
Dysfunction of the glycosylation machinery, caused for instance
by mutations of the conserved oligomeric Golgi (COG) complex,
has been shown to have dramatic consequences, including lethal
phenotypes (Wu et al., 2004). Some cancer cells have a fragmented
Golgi and exhibit defects in glycosylation (Kellokumpu et al., 2002).
The phenomenon described in this article could constitute an
important step for glycosylation quality control. A crucial test of
its influence would be to monitor the residency time of sensitive
proteins through the secretory pathway in general, and through the
Golgi in particular, to test whether, in certain pathologies, a higher
production of immature protein might be correlated with faster
protein transit through the secretory pathway. More generally,
there is a strong correlation between dysfunctional transport in the
secretory pathway and defects in protein folding and glycosylation,
though the causal chain remains controversial (Gonatas et al., 2006;
Cooper et al., 2006). The generality of the phenomenon described
here suggests that it could play an important role throughout the
secretory pathway.
5. Future outlook
Because identity maintenance imposes constraints on transport,
it plays a role in the function of trafficking organelles. We showed
that an apparent functional drawback (slowdown of intra-organelle
transport) can be beneficial (it increases maturation quality control). The interplay between identity maintenance, transport and
quality control have not been intensively studied as of yet. We
believe that studying this interplay, both experimentally and conceptually, can yield relevant information both in the understanding
of intra-organelle transport and of organelle dysfunction.
In our computations, we used a very simplified model with
only two compartments and one relevant chemical species, and
we made precise assumptions on the mathematical form of the
exchanged fluxes (Dmitrieff and Sens, 2011). However, our conclusions are very general: similar behaviors can be found for systems
with several compartments and several species, and with various
choices of transport functions. The critical fact is that transport
actually decreases when the total concentration in the system
increases, which is a consequence of the saturation of transport
and the increased probability of vesicle retrofusion at high concentration.
Transport through the Golgi remains controversial. Here, we
chose to focus on the so-called vesicular transport model, where
proteins are carried from one cisterna to the next by transport
vesicles. In the cisternal maturation model, newly synthesized proteins are transported through the Golgi without leaving a given
S. Dmitrieff, P. Sens / The International Journal of Biochemistry & Cell Biology 44 (2012) 718–721
721
interest. For instance, it has been shown that misfolded proteins
could undergo retrograde transport toward the E.R. for proper
refolding (Park et al., 2001). Specific fusion could also play a crucial
role in this process.
References
Fig. 2. Transport and maturation in an open system: (a) The same system as
sketched in Fig. 1a now interacts with its environment by receiving an influx Jin (in
compartment 1) and exporting molecules from compartment 2 at a rate kout . (b) Theoretical results of a pulse chase experiment, where a pulsed influx of molecules (in
red) enters the system via compartment 1. The concentration C1 quickly increases,
and exchange between the two compartments eventually causes the concentration C2 in compartment 2 to increase as well, but the dynamics of the process highly
depend on whether the system reaches the critical concentration C* (Fig. 1b), namely
on the intensity of the pulse. For low pulses, exchange between the two compartments remain roughly linear, C2 increases quickly to a level close to C1 , after what
both concentrations decreases with time at the export rate kout . For high pulses,
exchange is strongly suppressed by specific molecular interactions (which promotes
retrofusion), and C2 remains constant (and low) for a long time, because the weak
in-flux from compartment 1 is balanced by the export. Moreover, the transit time
of molecules through the system strongly increases under strong.
cisterna, which itself undergoes chemical maturation and physical
transport. However, this model must involve some kind of recycling, and hence some vesicular transport, and the consequences
of specific molecular interactions in this transport should be of
Binder B, Goede A, Berndt N, Holzhutter HG. A conceptual mathematical model
of the dynamic self-organisation of distinct cellular organelles. PLoS One
2009;4(12):e8295.
Bonfanti L, Mironov Jr AA, Martínez-Menárguez JA, Martella O, Fusella A, Baldassarre
M, et al. Procollagen traverses the Golgi stack without leaving the lumen of
cisternae: evidence for cisternal maturation. Cell 1998;95(7):993–1003.
Bremser M, Nickel W, Schweikert M, Ravazzola M, Amherdt M, Hughes CA, et al.
Coupling of coat assembly and vesicle budding to packaging of putative cargo
receptors. Cell 1999;96(4):495–506.
Chen YA, Scheller RH. SNARE-mediated membrane fusion. Nat Rev Mol Cell Biol
2001;2(2):98–106.
Cooper AA, Gitler AD, Cashikar A, Haynes CM, Hill KJ, Bhullar B, et al. !-Synuclein
blocks ER–Golgi traffic and Rab1 rescues neuron loss in Parkinson’s models.
Science 2006;313(5785):324–8.
Dunphy WG, Fries E, Urbani LJ, Rothman JE. Early and late functions associated
with the Golgi apparatus reside in distinct compartments. PNAS 1981;78(12):
7453–7.
Dmitrieff S, Sens P. Cooperative transport in cellular organelles. Phys Rev E
2011;83(4):041923.
Elsner M, Hashimoto H, NilssonT. Cisternal maturation and vesicle transport: join
the band wagon! Mol Membr Biol 2003;20(3):221–9 [review].
Gonatas MK, Stieber A, Gonatas JO. Fragmentation of the Golgi apparatus in neurodegenerative diseases and cell death. J Neurol Sci 2006;246(1-2):21–30.
Heinrich R, Rapoport TA. Generation of nonidentical compartments in vesicular
transport systems. J Cell Biol 2005;168:271–80.
Hirschberg CB, Snider MD. Topography of glycosylation in the rough endoplasmic
reticulum and golgi apparatus. Annu Rev Biochem 1987;56(1):63–87.
Jordens I, Fernandez-Borja M, Marsman M, Dusseljee S, Janssen L, Calafat J, et al.
The Rab7 effector protein RILP controls lysosomal transport by inducing the
recruitment of dynein–dynactin motors. Curr Biol 2001;11(21):1680–5.
Kellokumpu S, Sormunen R, Kellokumpu I. Abnormal glycosylation an altered
Golgi structure in colorectal cancer: dependence on intra-Golgi pH. FEBS Lett
2002;516:217–24.
Lippincott-Schwartz J, Roberts TH, Hirschberg K. Secretory protein trafficking
and organelle dynamics in living cells. Annu Rev Cell Dev Biol 2000;16:
557–89.
McNew JA, Parlati F, Fukuda R, Johnston RJ, Paz K, Paumet F, et al. Compartmental specificity of cellular membrane fusion encoded in snare proteins. Nature
2000;407(6801):153–9.
Michaelis L, Menten ML. Die kinetik der invertinwirkung. Biochem Z 1913;49(333369):352.
Misteli T. The concept of self-organization in cellular architecture. J Cell Biol
2011;155:181–5.
Munro S. The golgi apparatus: defining the identity of golgi membranes. Curr Opin
Cell Biol 2005;17(4):395–401.
Park B, Lee S, Kim E, Chang S, Jin M, Ahn K. The truncated cytoplasmic tail of
HLA-G serves a quality-control function in post-ER compartments. Immunity
2001;15(2):213–24.
Rabouille C, Hui N, Hunte F, Kieckbusch R, Berger EG, Warren G, et al. Mapping the
distribution of golgi enzymes involved in the construction of complex oligosaccharides. J Cell Sci 1995;108(4):1617.
Trucco A, Polishchuk RS, Martella O, Di Pentima A, Fusella A, Di Giandomenico D,
et al. Secretory traffic triggers the formation of tubular continuities across golgi
sub-compartments. Nat Cell Biol 2004;6(11):1071–81.
Wilson C, Ragnini-Wilson A. Conserved molecular mechanisms underlying homeostasis of the golgi complex. Int J Cell Biol 2010:758230.
Wu X, Steet RA, Bohorov O, Bakker J, Newell J, Krieger M, et al. Mutation of the
COG complex subunit gene COG7 causes a lethal congenital disorder. Nat Med
2004;10:518–23.