doi:10.1016/j.bpj.2009.12.2050


380a Monday, February 22, 2010
mechanisms through which upregulation of Bcl2 affects earlier steps of BAX-
mediated apoptosis are not fully understood. We found that BAX insertion
into the MOM was the earliest apoptotic step inhibited by Bcl2 overexpression.
Paradoxically, we also found that BAX translocation to the mitochondria was not
inhibited but rather spontaneously increased in this same genetic context. This
increase in mitochondrial associated BAX required a physical interaction be-
tween BAX and Bcl2. We therefore propose that, at least when upregulated,
Bcl2 behaves as a ‘decoy receptor’ which sequestrates BAX at the mitochondria
but inhibits its insertion into the MOM, committing the cell to survive.
Supported by NYU Research challenge Funds to LD.

1961-Pos
Compartmentalization of BCL2 Family Proteins Mediated by Organelle
Lipid Membranes
Rebecca J. Boohaker, Ge Zhang, Adina Loosley, Kathleen Nemec,
Annette R. Khaled.
University of Central Florida, Orlando, FL, USA.
Cancer is defined by a pronounced inhibition of cell death. The BCL2 family of
proteins tightly regulates the delicate balance between life and death. One
method of regulation is the compartmentalization of antagonistic members.
For example, Bax, a pro-apoptotic member of this family, acts as the penulti-
mate factor in the apoptotic cascade by releasing apoptogenic factors such as
Cytochrome C from the mitochondrial lumen. The normally cytosolic protein
translocates from one internal compartment to another through an elusive
mechanism. Individual organelles are defined not only by function (mediated
by specific membrane bound proteins), but by the unique composition of their
phospholipid membranes. In this work, we have evaluated the contribution of
organelle lipids to the localization of of BCL2 proteins.

1962-Pos
In Search of the Structure of MAC in the Mitochondrial Outer Membrane
Pablo M. Peixoto1, Christian Renken2, Shefali Haldar3, Carmen Mannella2,
Kathleen W. Kinnally1.
1
New York University College of Dentistry, New York, NY, USA,

2Wadsworth Center, Albany, NY, USA, 3Academy of the Holy Names,
Albany, NY, USA.
Several groups have tried to determine the structure of the channel (MAC)
formed in mitochondrial outer membranes (OM) of apoptotic cells or in syn-
thetic membranes by Bax and related proteins/peptides, using electron micros-
copy (EM), atomic force microscopy and x-ray diffraction. Here, pore-like
structures ~3-10 nm were handpicked from transmission EM images of ura-
nyl-acetate-stained OMs isolated from control and apoptotic (IL3-deprived)
FL5.12 cells. These ‘‘candidate pores’’ were aligned by correlation procedures,
and class averages defined by principal component and K-means analyses.
Main differences in the class averages were (1) the presence of one or more
dark (stain-filled) pores, and (2) the nature of white (stain excluding) features
around the pores. A class average consisting of a single 3-nm pore with pro-
nounced white rim was rotationally averaged and used as a reference for
cross-correlation searches of 50 OM images from control and apoptotic cells.
Searches using this 3-nm ‘‘donut’’ motif and the same motif doubled in size
(6-nm ‘‘donut’’) yielded thousands of ‘‘hits’’ in both control and apoptotic
membranes, which were subsequently aligned and classified as before. The pre-
dominant stain-filled structures found with both motifs were not circular but
elongated (up to ~4x6 nm), extending away from stain-excluding crescent-
shaped features. The radial anisotropy ruled out reference bias and was in-
consistent with pores formed by rings of evenly spaced protein subunits. We
hypothesize that the different classes of structures detected represent stages
in formation of MAC as an increasingly large membrane bilayer defect (or
‘‘cleft’’) induced by successive aggregation or clustering of Bax/Bak molecules.
A progressive assembly mechanism for MAC has been recently suggested by
real-time monitoring of MAC conductances in isolated mitochondria by patch
clamping (Martinez-Caballero et al. J Biol Chem 284: 12235-45). Supported
by NIH grant GM57249.

1963-Pos
Effects of MAC Formation on Mitochondrial Morphology
Pablo M. Peixoto, Shin-Young Ryu, Robert Range, Kathleen W. Kinnally.
New York University College of Dentistry, New York, NY, USA.
Accumulating literature associate mitochondrial dynamics with apoptosis,
since regulation of either process has reciprocal effects. These processes
seem to converge in formation of the mitochondrial apoptosis induced channel,
MAC, which releases cytochrome c and triggers the degradation phase of apo-
ptosis. While Bax and Bak, core components of MAC, were shown to interact
with fusion and fission proteins, some studies also suggest proteins from the in-
termembrane space could leak to the cytosol and further promote mitochondrial
fission during apoptosis. The temporal relationship between apoptosis induc-
tion, MAC formation and mitochondrial fragmentation was investigated by
time lapse microscopy. MAC function was induced through staurosporine treat-
ment and microinjection of tBid or cytochrome c. MAC formation and mito-
chondrial dynamics under these conditions were monitored in Hela cells (clone
10) that stably express low levels of GFP-Bax and were transiently transfected
with a pDsRed-2 plasmid. GFP-Bax relocation to mitochondria only during
apoptosis signals MAC formation, while pDsRed-2 expression shows mito-
chondrial structure as red fluorescence. Treatment with staurosporine and
microinjection with tBid or cytochrome c induced relocation of Bax and col-
lapse of the mitochondrial network. The temporal relationship between these
two events was further analyzed. Interestingly, pretreatment with iMAC2, a spe-
cific MAC blocker, protected against cell death and prevented mitochondrial
fragmentation after tBid injection. Our results suggest a link exists between
MAC formation and collapse of the mitochondrial network during apoptosis.
Supported by NIH grant GM57249.

1964-Pos
Mechanism of the Mitochondrial Cytochrome C Release Wave in
Bid-Induced Apoptosis
Soumya Sinha Roy, Cecilia Garcia-Perez, Erika Davies, Xuena Lin,
György Hajnóczky.
Thomas Jefferson University, Philadelphia, PA, USA.
Bid, a BH3-only Bcl2 family protein, plays a central role in apoptosis. Bid is
cleaved by caspase-8 and other enzymes forming tBid that induce mitochon-
drial outer membrane (OMM) permeabilization and cytochrome c (cyto-c) re-
lease. However a mystery remains how Bid synchronizes the function of a large
number of discrete organelles, particularly in mitochondria-rich liver or muscle
cells. Here we showed that tBid (0.5-50nM) elicited progressive OMM perme-
abilization and complete cyto-c release with a dose-dependent lag time and rate
in H9c2 cell populations. Once started, the OMM permeabilization was not pre-
vented by tBid washout. In contrast, the dose-response for digitonin-induced
OMM permeabilization displayed quantal behavior. In single cell imaging
studies, permeabilized H9c2 and primary human cardiac cells transfected
with cyto-c-GFP showed complete tBid-induced cyto-c-GFP release closely
followed by mitochondrial depolarization. The cyto-c-GFP release started at
discrete sites and propagated through the mitochondria with a constant velocity
and a relatively stable kinetic of release in each organelle. Similar tBid-induced
cyto-c-GFP release wave was documented in intact H9c2 cells transfected with
tBid. The waves were not dependent on Ca

2þ
, caspase activation or permeabil-

ity transition pore opening. However, treatment with MnTMPyP, a ROS scav-
enger or overexpression of mitochondrial superoxide dismutase suppressed the
coordinated cyto-c release and also inhibited tBid-induced cell death. On the
other hand, both superoxide and hydrogen peroxide sensitized mitochondria
to the tBid-induced permeabilization. Thus, tBid engages a ROS-dependent
inter-mitochondrial signaling mechanism for spatial amplification of the
apoptotic signal by mitochondrial waves.

1965-Pos
Role of Milton Domains in the Calcium-Dependent Regulation of Mito-
chondrial Motility
Sudipto Das, Gyorgy Hajnoczky.
Thomas Jefferson University, Philadelphia, PA, USA.
The mammalian GRIF1 and OIP106, and the drosophila Milton are kinesin-
binding proteins that form a complex with the Miro GTPase, an outer mitochon-
drial membrane EF-hand protein, to support the movement of mitochondria
along the microtubules. Our study demonstrates that in H9c2 cells overexpress-
ing OIP106 or GRIF1, the basal motility of mitochondria is increased, whereas
the sensitivity to calcium-induced movement inhibition is decreased. To dissect
the interaction between Milton, kinesin and Miro, three different Milton
constructs were tested: Milton (1-450), the soluble domain of Milton; Milton
(750-1116), lacking the kinesin heavy chain binding site and Milton
(847-1116) that lacks additional ~100 amino acid presumably containing part
of the Miro binding site. Immunohistochemistry revealed that the overexpressed
Milton (1-450) was cytoplasmic, whereas the other two Milton constructs
showed mitochondrial localization. The basal mitochondrial motility was in-
creased by Milton (750-1116) but was not altered by Milton (847-1116) or Mil-
ton (1-450). A plot of mitochondrial motility against slowly rising cytoplasmic
[Ca2þ] induced by thapsigargin (2mM), shows that overexpression of Milton
(750-1116) significantly reduced the calcium sensitivity of mitochondrial motil-
ity reminiscent of OIP106 and GRIF1. By contrast, Milton (847-1116) or Milton
(1-450) did not have any effect. The thapsigargin-induced cytoplasmic calcium
signal was not affected by any of the Milton constructs. These data indicate that



Monday, February 22, 2010 381a
the C terminus of Milton is an important regulator of the mitochondria associated
motors and is involved in conferring the calcium sensitivity from Miro to the
motors. Milton 847-1116 is sufficient for the mitochondrial binding, whereas
the 750-847 amino acids are critical for the control of calcium sensitivity of mi-
tochondrial motility.

1966-Pos
Biophysical Properties of Mitochondria Undergoing Fusion
Xingguo Liu1, Orian Shirihai2, Gyorgy Hajnoczky1.
1Thomas Jefferson University, Philadelphia, PA, USA, 2Boston University,
Boston, MA, USA.
Emerging evidence shows the importance of genes controlling mitochondrial
fusion in physiology and their deregulation in neurodegenerative and metabolic
disorders. However, apart from DJm, the biophysical properties of the fusion-
competent mitochondria remain elusive. To evaluate the conditions of contact
formation and fusion, we used organelle-targeted fluorescent proteins, includ-
ing photoactivatable GFP, which allow tracking of individual mitochondria. In
H9c2 cells, almost every mitochondrion is aligned with microtubules that pro-
vide the primary tracks for mitochondrial movement. Our results show that
~90% of fusion events involve moving mitochondria. However, fusion occurs
irrespective of mitochondrial speed. Furthermore, ~80% of fusion events in-
volve the tip portion of the mitochondrion, whereas only ~50% of these events
involve organelle side. Nocodazol, a microtubule disrupting agent that inhibits
mitochondrial movements decreases the fusion frequency and changes mito-
chondrial fusion sites. Strikingly, 80-90% of the physical contacts between ad-
jacent mitochondria did not result in fusion events. To evaluate whether the fu-
sion efficacy depends on the spacing between the outer and inner mitochondrial
membranes we used drugs that alter the matrix volume. Valinomycin, a K

þ
ion-

ophore that induces matrix swelling evoked a decrease in mitochondrial motil-
ity leading to fewer contacts among mitochondria but the number of fusion
events was maintained, indicating an increase in fusion efficacy. This change
occurred at a low valinomycin concentration (0.25nM) that did not affect
DJm or Opa1 cleavage. Nigericin (0.5mM), a K

þ
/H
þ

ionophore that induces
shrinkage of the matrix elicited fusion inhibition and mitochondrial aggrega-
tion. Importantly, no motility inhibition or Opa1 cleavage occurred at the
same time and the DJm was increased. These results suggest that mitochon-
drial fusion is facilitated by mitochondrial motility, the key determinant of
the inter-mitochondrial encounter numbers and preferentially involves the
front-tip of the moving organelle. In addition, fusion efficacy depends on the
mitochondrial matrix volume.

1967-Pos
Mitochondrial Fusion Dynamics in Human Skeletal Muscle-Derived Cells
Veronica Eisner, Gyorgy Hajnoczky.
Thomas Jefferson University, Philadelphia, PA, USA.
Mitochondria have a fundamental role in both muscle physiology and pathol-
ogy. Mitochondrial fusion and fission are important for energy metabolism,
calcium homeostasis and cell death. However, the mechanisms underlying
mitochondrial dynamics are poorly understood, especially in physiological
models such as skeletal muscle. Here we evaluated mitochondrial fusion dy-
namics in human skeletal muscle cells (HUSMC). Skeletal muscle satellite
cells were isolated from human muscle biopsies and were maintained and dif-
ferentiated in cell culture. Mitochondrial fusion events were evaluated by con-
focal imaging of cells expressing mitochondria matrix targeted DsRed and
matrix targeted or outer mitochondrial membrane (OMM) targeted photoacti-
vatable-GFP. When we tagged the mitochondria in ~20% of total cellular
area with photoactivated-GFP, we found those mitochondria undergoing matrix
fusion with a frequency of 1.3 5 0.1 events/min/cell (n=70). Among the fusion
events, 40% led to complete fusion and only 10% was followed by separation at
the apparent fusion site within 20 to 40 seconds. Both complete and transient
fusion events resulted mostly from longitudinal mergers, involving end to
end interaction or from mergers of adjacent mitochondria in side to side orien-
tation. Furthermore, we found that OMM and matrix fusion are sequential and
separable steps, displaying 5.8 5 1 seconds gap (n=10). Finally, we evaluated
the mitochondrial fusion dynamics in HUSMC derived from both normal and
malignant hyperthermia susceptible individuals. At resting state, no significant
differences were found in the number of events or in their characteristics. Thus,
mitochondrial fusion commonly occurs in HUSMC, and enables mixing of both
soluble and integral membrane factors. This process would help to maintain the
stability of mitochondrial metabolism. The relatively low frequency of the tran-
sient fusion is probably due to the parallel organization of the cytoskeletal
tracks for mitochondria and to the limited mitochondrial motility.
1968-Pos
Imaging Interorganelle Contacts and Local Calcium Dynamics at the
ER-Mitochondrial Interface
György Csordás1, Péter Várnai2, Tünde Golenár1, Swati Roy1,
George Purkins1, Tamás Balla3, György Hajnóczky1.
1
Thomas Jefferson University, Philadelphia, PA, USA,

2
Semmelweis

University, Budapest, Hungary, 3National Institutes of Child Health and
Human Development, Bethesda, MD, USA.
The local coupling between ER and mitochondria is essential for proper cell
function. A main role of the ER-mitochondrial junctions is to provide a local
calcium signaling domain that is important both for keeping energy production
in line with demand and for the control of apoptotic mechanisms. So far it has
not been possible to visualize the tiny ER-mitochondrial contact points in living
cells or to monitor the localized [Ca2þ] changes in the narrow space between
ER and mitochondria ([Ca2þ]ER-mt).
Here, we exploited rapamycin-mediated heterodimerization of FKBP12 and
FRB domains of fluorescent protein constructs respectively targeted to the outer
mitochondrial membrane and the ER as drug-inducible inter-organelle linkers
to identify ER-mitochondrial contacts and to measure the [Ca

2þ
]ER-mt. High-

resolution fluorescence imaging and 3D reconstruction revealed rapamycin-in-
duced clustering of the ER-targeted fluorescent linker-half to the contact areas
with the mitochondria without major changes in the spatial arrangement of the
ER. Essentially all mitochondria displayed contacts with the ER in both
RBL-2H3 and H9c2 cells. Plasma membrane-mitochondrial contacts were
less frequent with ER stacks being inserted between the two organelles. Single
mitochondria display discrete patches of ER contacts as well as continuous as-
sociations. Cytoplasmic and mitochondrial matrix [Ca2þ] showed robust
ER-mitochondrial Ca2þ transfer with considerable heterogeneity even among
adjacent mitochondria. Pericam-containing linkers revealed IP3-induced
[Ca

2þ
]ER-mt signals that were resistant to buffering bulk cytosolic [Ca

2þ
]c in-

creases and exceeded 9 uM. The largest [Ca
2þ

]ER-mt signals did not occur at
the tightest associations, indicating space requirements for the Ca

2þ
transfer

machinery and functional diversity among ER-mitochondrial junctions.
These studies provide direct evidence for the existence of high Ca2þ microdo-
mains between the ER and mitochondria in living cells, and open new possibil-
ities to probe the functional importance of this specialized compartment.

1969-Pos
Dependence of ER-Mitochondria Calcium Transfer on Different IP3
Receptor Isoforms
Tunde Golenar1, Szava Bansaghi1, Gyorgy Csordas1, David I. Yule2,
Suresh K. Joseph1, Gyorgy Hajnoczky1.
1Thomas Jefferson University, Philadelphia, PA, USA, 2University of
Rochester Medical Center, Rochester, NY, USA.
IP3 receptors (IP3Rs) release Ca2þ from the ER, which is locally relayed to the
mitochondria to control several aspects of mitochondrial function. Recent stud-
ies have suggested that type 3 IP3Rs (IP3R3) are particularly important for me-
diating the Ca2þ transfer at the ER-mitochondrial interface. We set out to sys-
tematically evaluate the respective role of each IP3R isoform in chicken DT40
cell lines that express only one IP3R isoform (double-knockout, DKO1, DKO2
and DKO3) or provide a null-background (triple-knockout, TKO) for analysis
of mammalian IP3Rs and their mutants.
Simultaneous imaging of cytoplasmic and mitochondrial matrix [Ca2þ]
([Ca2þ]c and [Ca2þ]m) was performed in either permeabilized cells chal-
lenged with IP3 or in muscarinic receptor overexpressing intact cells stimulated
with carbachol (Cch), an IP3-linked agonist. Saturating IP3 evoked complete
discharge of ER calcium and resulted in comparable [Ca2þ]m increases in
each DKO. Furthermore, the Cch-induced [Ca2þ]c spike was closely followed
by a [Ca2þ]m rise in each DKO. When TKO cells were rescued with rat IP3R1
or IP3R3, the latter mediated a larger [Ca2þ]c transient but the [Ca2þ]m
increases were very similar for both isoforms. The relationship between the
[Ca2þ]c peak and the corresponding [Ca2þ]m response was also very similar
for IP3R1 and IP3R3. To assess the impact of the release kinetic through IP3R1
in the mitochondrial Ca2þ transfer we used two point mutants of the IP3R1,
which display either enhanced inhibition by Ca2þ (D426N) or are relatively
insensitive to Ca2þ inhibition (D442N). D426N showed dampened [Ca2þ]m
signal and [Ca2þ]m vs. [Ca2þ]c relationship, whereas D442N displayed
a steeper [Ca2þ]m vs. [Ca2þ]c relationship than the wild type IP3R1. Thus,
each IP3R isoform can support local ER-mitochondrial Ca2þ transfer and their
competence to activate mitochondrial Ca2þ uptake depends on their deactiva-
tion kinetic.


	Compartmentalization of BCL2 Family Proteins Mediated by Organelle Lipid Membranes
	In Search of the Structure of MAC in the Mitochondrial Outer Membrane
	Effects of MAC Formation on Mitochondrial Morphology
	Mechanism of the Mitochondrial Cytochrome C Release Wave in Bid-Induced Apoptosis
	Role of Milton Domains in the Calcium-Dependent Regulation of Mitochondrial Motility
	Biophysical Properties of Mitochondria Undergoing Fusion
	Mitochondrial Fusion Dynamics in Human Skeletal Muscle-Derived Cells
	Imaging Interorganelle Contacts and Local Calcium Dynamics at the ER-Mitochondrial Interface
	Dependence of ER-Mitochondria Calcium Transfer on Different IP3 Receptor Isoforms